V/ AS 8332- Volume 101 Number 1 Spring 2015 Journal of the WASHINGTON ACADEMY OF SCIENCES MCZ LIBRARY NOV 3 0 2015 HARVARD UNIVERSITY Board of Discipline Editors Editor’s Comments S. Rood iii Intellectual Washington Today S. Umpleby 1 Does Speed Matter? The Employment Impact of Increasing Access to Fiber Internet P. Lapointe 9 Benjamin Banneker and Celestial Navigation: Just How Did They Know Where They Were, Then? S. Howard 29 Washington Academy of Sciences Awards Program 2015 43 Addendum to Washington Academy of Sciences 2014 Membership Directory 53 In Memoriam: Burton G. Hurdle (1918-2015) 57 Membership Application 59 Instructions to Authors 60 Affiliated Institutions 61 Affiliated Societies and Delegates 62 ISSN 0043-0439 Issued Quarterly at Washington DC Washington Academy of Sciences Founded in 1898 Board of Managers Elected Officers: President Mina Izadjoo President Elect Mike Coble Treasurer Ronald Hietala Secretary John Kaufhold Vice President, Administration Nick Tran Vice President, Membership Sue Cross Vice President, Junior Academy Vice President, Affiliated Societies Gene Williams Members at Large Paul Arveson Michael P. Cohen Frank Haig, S.J. Neal Schmeidler Mary Snieckus The Journal of the Washington Academy of Sciences The Journals the official organ of the Academy. It publishes articles on science policy, the history of science, critical reviews, original science research, proceedings of scholarly meetings of its Affiliated Societies, and other items of interest to its members. It is published quarterly. The last issue of the year contains a directory of the current membership of the Academy. Subscription Rates Members, fellows, and life members in good standing receive the Journal free of charge. Subscriptions are available on a calendar year basis, payable in advance. Payment must be made in U.S. currency at the following rates. U.S. and Canada $30.00 Other Countries $35.00 Single Copies (when available) $15.00 Claims for Missing Issues Claims must be received within 65 days of mailing. Claims will not be allowed if non- delivery was the result of failure to notify the Academy of a change of address. Past President Terrell Erickson Affiliated Society Delegates Shown on back cover Editor of the Journal Sally A. Rood Notification of Change of Address Address changes should be sent promptly to the Academy office. Notification should contain both old and new addresses and zip codes. POSTMASTER: Send address changes to Washington Academy of Sciences, Room 113, 1200 New York Ave. NW, Washington, DC 20005 Academy Office Washington Academy of Sciences Room 113 1 200 New York Ave. NW Washington, DC 20005 Phone: (202) 326-8975 Journal of the Washington Academy of Sciences (ISSN 0043-0439) Published by the Washington Academy of Sciences (202) 326-8975 Email: iournal@washacadsci.ora Website: www.washacadsci.ora 1 Journal of the 1200 New York Ave. Suite 113 Washington DC 20005 www.washacadsci. org MCZ LIBRARY WASHINGTON NOV 3 0 2015 ACADEMY OF SCIENCES HARVARD UNIVERSITY Volume 101 Number 1 Spring 2015 Contents Board of Discipline Editors ii Editor’s Comments S. Rood iii Intellectual Washington Today S. Umpleby 1 Does Speed Matter? The Employment Impact of Increasing Access to Fiber Internet P. Lapointe 9 Benjamin Banneker and Celestial Navigation: Just How Did They Know Where They Were, Then? S. Howard 29 Washington Academy of Sciences Awards Program 2015 43 Addendum to Washington Academy of Sciences 2014 Membership Directory ....53 In Memoriam: Burton G. Hurdle (1918 - 2015) 57 Membership Application 59 Instructions to Authors 60 Affiliated Institutions 61 Affiliated Societies and Delegates 62 ISSN 0043-0439 Issued Quarterly at Washington DC Spring 2015 ii Journal of the Washington Academy of Sciences Editor Sally A. Rood, PhD sally.rood2@gmail.com Board of Discipline Editors The Journal of the Washington Academy of Sciences has a 12- member Board of Discipline Editors representing many scientific and technical fields. The members of the Board of Discipline Editors are affiliated with a variety of scientific institutions in the Washington area and beyond — government agencies such as the National Institute of Standards and Technology (NIST); universities such as George Mason University (GMU); and professional associations such as the Institute of Electrical and Electronics Engineers (IEEE). Anthropology Astronomy Biology/Biophysics Botany Chemistry Environmental Natural Sciences Health History of Medicine Operations Research Physics Science Education Systems Science Emanuela Appetiti eappetiti@hotmail.com Sethanne Howard sethanneh@msn.com Eugenie Mielczarek mielczar@, physics. gmu.edu Mark Holland maholland@salisbury.edu Deana Jaber diaber@marvmount.edu Terrell Erickson terrell.ericksonl@wdc.nsda.gov Robin Stombler rstombler@auburnstrat.com Alain Touwaide atouwaide@hotmail.com Michael Katehakis mnk@rci.rutgers.edu Katharine Gebbie katharine.gebbie@nist.gov Jim Egenrieder iim@deepwater.org Elizabeth Corona elizabethcorona@gmail.com Washington Academy of Sciences Ill Editor’s Comments In this issue of the Journal of the Washington Academy of Sciences we are celebrating the Washington, D.C., region and its science and technology presence! Back in 1985, Amitai Etzioni’s Washington Post editorial, “The World-Class University that Our City Has Become,” was his personal statement as a new resident of the Washington, D.C., area in the mid- 1980s. It provided an interesting view of the city’s aspirations in science and technology and policy circles at that time. Stuart Umpleby rediscovered this editorial and provides an updated perspective in “Intellectual Washington Today.” While Etzioni’s emphasis was on the policy community — he called it “Washington Metropolitan University” or W.M.U. — Umpleby’ s emphasis is on the more recent growth of information-related activities in the Washington, D.C. area. Regardless, these dual perspectives highlight the important role of the science and technology community and academic and policy institutions in the affairs of the Washington, D.C., metropolitan region. In line with this celebration of the Washington area’s science presence, it is fitting that this issue documents the Academy’s annual Awards Program and ceremony. Sethanne Howard presented the keynote at the banquet — about the scientist, Benjamin Banneker, who lived in the Baltimore area from 1731 to 1806. The geographical boundary for Washington, D.C., was surveyed back in the late 1700s using the eclipses of the Galilean satellites to determine longitude. As part of the survey team, Banneker timed the eclipses of the Galilean satellites by Jupiter and he kept the survey clocks running at the right time. Thus, the title of the presentation at the awards ceremony was “Benjamin Banneker and Celestial Navigation: Just How Did They Know Where They Were, Then?”. Banneker’s story was quite interesting at the banquet ... and is now equally interesting in this issue of the Journal ! As an introduction to the Academy’s 2015 Awards Program, we provide a “backgrounder” on the Awards Program. It includes an early history of the program, some traditions, and a primer on the nominations process. Congratulations to these distinguished scientists and educators in Washington, D.C., area scientific institutions whom the Academy honored with awards in their fields this year: Ronald Colie, Ram D. Sriram, Marcus Cicerone, Paul Peterson, Robert Gover, Gregory Strouse, and Spring 2015 IV MaryBeth Petrasek. Details on each of their awards are presented along with photos from the ceremony and banquet. A quantitative study of the policy implications of broadband improvements across the country is presented in the paper by Paul Lapointe entitled “Does Speed Matter? The Employment Impact of Increasing Access to Fiber Internet.” The study finds a positive association between increasing access to fiber cable and increases in employment and the number of firms at the county level which, in turn, offers evidence that promoting access to fiber internet is a viable approach to economic development. This Journal issue also includes an Addendum to the Academy’s 2014 Membership Directory which appeared in the Winter 2014 issue of the Journal of the Washington Academy of Sciences. The names of twenty new members who were inadvertently omitted from the 2014 Membership Directory are printed in this issue instead of waiting for the next Directory. My sincere apologies for their omission from our annual Membership Directory this past year. In addition, in this Journal issue we honor the life of Burton Hurdle, a long-time member of the Washington science community who passed away this Spring. Finally, this Journal issue marks my last issue as editor. I’ve edited the Journal for three years, and at this time I am handing over the editorship to Sethanne Howard. Please send Sethanne your manuscripts and other input going forward. Eve enjoyed working with all the authors, reviewers, proofreaders, and members of the Board of Discipline Editors who have contributed their time so that we can maintain high standards for the Journal. I’ve been blessed by the large number of talented people interested in supporting this unique peer-reviewed interdisciplinary Journal. It’s been an honor working with all of you ... too many to mention individually over that period of time ... please know that I appreciate and thank each of you from my heart. Sally A. Rood, PhD, Outgoing Editor Journal of the Washington Academy of Sciences Washington Academy of Sciences 1 Guest Editorial Intellectual Washington Today Stuart A. Umpleby The George Washington University, Washington, D.C. Abstract In a Washington Post editorial thirty years ago, Amitai Etzioni described how Washington, D.C. was becoming an intellectual city. Previously, Washington was viewed as the home of the national government, journalism, lawyers, and lobbyists but not as an academic or intellectual city. However, Etzioni claimed that Washington had become a policy and scientific powerhouse as a result not only of its growing and improving universities and their research institutes, but also because of its federal agencies, think tanks, and policy research organizations. This article reviews the points made by Etzioni and examines the situation today. Introduction Washington, D.C., is a city with many ironic descriptions. It is often described as the Northern-most Southern city. John F. Kennedy said it was a city of Northern charm and Southern efficiency. It has been called a city full of former student body presidents, and a city consisting of residents who come from somewhere else. Currently Washington may be known as a city of politicians, interns, diplomats, and bloggers. It is not often thought of as a scientific city or an intellectual city. However, Washington has been growing and changing. As the nation becomes increasingly a post-industrial society, Washington is becoming a leader in new types of organizations and new kinds of jobs. A Description of Washington 30 Years Ago To explain the purpose of an editorial he contributed to the Washington Post in the Spring of 1985, Amitai Etzioni said that people sometimes asked him why he had moved from Columbia University to Washington, D.C., which previously had not had a reputation as a source Spring 2015 2 of innovative ideas. He wrote that the Washington Metropolitan University — the combination of universities, policy research institutes, and government agencies — “easily matches the intellectual vigor of contemporary London,” and that it had “almost as many little magazines (where intellectuals float new ideas) and writers-in-residence as the Left Bank of Paris.” Etzioni pointed out that several new research organizations had been added to the D.C. area prior to 1985: the Roosevelt Center, the Center for National Policy, and the Cato Institute. He also noted that the National Institutes of Health (NIH) did more research in biology and related disciplines than was conducted at Harvard, Yale, Princeton, Columbia and Brown combined. Major research centers in economics could be found in the World Bank, the Federal Reserve Board, and the Congressional Budget Office. The natural sciences were strong in the Carnegie Institution of Washington, the Smithsonian Institution, the National Institute of Standards and Technology in Gaithersburg, Maryland, and the Department of Defense (DOD). Etzioni made a distinction between academics who were deep scholars of narrow topics and intellectuals who took a broader perspective on the direction of American society and trends in literature and the arts. He claimed that many intellectuals had moved to D.C. because they found the academic abundance congenial. Etzioni also noted that academics and intellectuals communicated with each other not only in seminars, but also in magazines that stimulated new ideas. As just a few of these published in D.C., he listed: • The Wilson Quarterly , • the American Enterprise Institute’s Public Opinion , • Regulation , which reports on the effects of government intervention, • The Cato Journal, and • Foreign Policy magazine, then a new competitor to Foreign Affairs, published in New York. Etzioni further noted that Science magazine was the nation’s leading journal of science, and that the National Academy of Sciences published Washington Academy of Sciences 3 Issues if 7 Science and Technology > (both products of D.C.) Finally, he noted that Washington provides numerous television news and discussion programs to the nation. As an academic and intellectual city, how has Washington progressed since 1985? Many Universities in Washington There has been continued growth and improvement in universities, particularly the growth of George Mason University since it became independent in 1972. The familiar Washington, D.C., universities — American University, Catholic University, Georgetown University, George Washington University, Howard University, Johns Hopkins University’s School of Advanced International Studies, the University of Maryland, Marymount University, and the University of the District of Columbia — are all prospering. Several well-established universities, for example George Washington University and the University of Maryland, now have buildings in several parts of the city. These locations provide classes more conveniently to students but also conduct research, as does George Washington University’s Virginia Science and Technology Campus in close-by Ashbum, Virginia. Universities based in other cities also have a presence in the Washington area. For example, Cornell University, New York University, Syracuse University, Pepperdine University, and Virginia Tech are all here. Clearly universities find it desirable to have a connection to Washington, D.C. The Growth of Information-Based Activities The information intensive activities of the federal government have also expanded greatly since 30 years ago. A few examples of such activities in the Washington area are the following: • The National Security Agency at Fort Meade, Maryland, has become the center of a “cyber valley” in the Baltimore - Washington corridor. [Schiff, 2013] Spring 2015 4 • The Dulles access toll road in northern Virginia contains the expanded “beltway” contracting firms and information services firms such as AOL. • The Route 270 corridor in Maryland just north of D.C. continues to be the home for biological research, with key institutions being NIH, the Walter Reed Army Medical Center, and Bethesda Naval Medical Center. • Research programs, administered at NASA Headquarters and the Goddard Space Flight Center in nearby Greenbelt, Maryland, have made fundamental contributions to improving weather forecasting, to earth science, and to our understanding of climate change. NASA’s Hubble Space Telescope has dramatically advanced our understanding of the cosmos. • The number of patents issued by the U.S. Patent and Trademark Office (USPTO), headquartered in Alexandria, Virginia, in the past twenty years has more than tripled, from 113,268 in 1994 to 329,613 in 2014. [USPTO, 2015] The USPTO now has not only a new building but a new campus in Alexandria, just south of D.C. Washington is definitely a leader in applications of information technology. The Internet, an outgrowth of an earlier DOD research project, has transformed business, government and personal communication in recent years. The D.C. area’s knowledge workers now spend hours each day in “cyberspace” and the contents of filing cabinets are now “in the cloud” with both positive and negative consequences. Cybersecurity is a leading domestic and international concern and “identity theft” is a new worry for private citizens. The Washington Post is now owned by Amazon.com. Many newspapers have gone out of business. There are now numerous blogs written by former journalists. Improving Management in Government and Business In the years since Etzioni wrote his article, there have been many changes in the federal government which have transformed both the practice of government in Washington and also influenced the management of corporations and state and local governments. In 1987, Congress created the Malcolm Baldrige National Quality Improvement Program aimed at improving the productivity of U.S. firms, Washington Academy of Sciences 5 which in the 1970s were having difficulty competing with Japanese manufacturers. The Baldrige National Quality Award was expanded to include education and health care organizations in 1 999, and a government and non-profit category was added in 2007. As an example of Washington’s growing influence, a 1991 General Accounting Office report showed how the Baldrige Program companies increased their market share an annual average of 13.7 percent. [Garvin, 1991] Such a high growth rate meant that companies using quality improvement methods in just a few years bought or replaced companies not using these methods. A more recent study said that participating companies had an 820:1 ratio of benefits for the U.S. economy to program costs. [Link and Scott, 2012] To arrive at this ratio, they compared the benefits received by the 273 Malcolm Baldrige National Quality Award applicants from 2007 to 2010 with the cost of operating the Baldrige Program. The 820-to-l ratio represents only the benefits for the surveyed applicants, but it represents all of the Baldrige Program’s costs. Link and Scott note that the benefit-to-cost ratio would be much higher if program costs were compared with benefits for the entire U.S. economy. Quality Improvement Methods were taken seriously by President Bill Clinton who appointed Vice President A1 Gore to head the National Performance Review in 1993. This initiative had the goal of dramatically improving the performance of the federal government through a combination of process improvement methods and increased contracting as an alternative to larger government agencies. In March 1998, the National Performance Review pointed to a number of important achievements, later presented in Kamensky [1999]: • The size of the federal civilian workforce was cut by 351,000 — the smallest since President Kennedy held office and, as a percentage of the national workforce, the smallest since 1931. • Action was recommended on about 1,500 issues in 1993 and 1995. Agencies completed about 58 percent. Of the original recommendations, 66 percent were reported as completed. For those requiring Presidential or Congressional action, President Clinton signed 46 directives and Congress passed and the President signed over 85 laws. • About $177 billion in savings were recommended over a 5-year period. Agencies locked into place about $137 billion. In addition, as of March 1998, the process improvement award winners Spring 2015 6 estimated savings or cost avoidances of about $3 1 billion because of their actions. • Agencies eliminated about 640,000 pages of internal rules, about 16,000 pages of Federal Regulations, and rewrote 31,000 additional pages into plain language. • Agencies sponsored 850 labor-management partnerships. A 1998 survey of employees showed those in organizations that actively promoted reinvention were twice as satisfied with their jobs. • Over 570 federal organizations had committed to more than 4,000 customer service standards. Kamensky [1999] also reported that public trust in the federal government was increasing after a 30-year decline. While it was not clear whether this improvement was directly linked to the work of the National Performance Review, the Review made an important contribution. The Use of Information in Policy-Making Who analyzes information and writes reports in the D.C. area has also been changing. Since Etzioni wrote his article, the Office of Technology Assessment was closed and the number of Congressional staff was significantly reduced during the time that Newt Gingrich was Speaker of the U.S. House of Representatives. As a result, it can be said that the task of providing background information for legislation has been taken up by lobbying firms on K Street which often draft new legislation, a task previously performed by Congressional staff members. [Benen, 2011] Also, political activity has moved from public demonstrations on the mall to campaign contributions and lobbying behind closed doors. It is harder now, in 2015, to know what changes in laws are occurring. Hedrick Smith [2012] in his recent book notes that when he was head of the Washington bureau of The New York Times, he did not realize that a series of laws and court decisions were fundamentally changing taxes and entitlement programs beginning in the late 1970s. Over time, these changes have led to a dramatic increase in inequality in the United States, which has affected all U.S. citizens. Conclusion In his article 30 years ago, Amitai Etzioni focused primarily on the many policy research organizations in Washington. During his years as a professor at the George Washington University, Etzioni himself has made Washington Academy of Sciences 7 notable contributions to policy research and discussions. He created the Society for the Advancement of Socio-Economics and an academic journal, the Journal for the Advancement of Socio-Economics. He founded and leads the Communitarian Network, a non-profit, non-partisan organization dedicated to supporting the moral, social and political foundations of society. He is currently Director of the Institute for Communitarian Policy Studies at George Washington University. Of course not all of the information-related activities in the Washington area involve transformative policy analyses. Much of the work — for example, at the Patent Office and the National Security Agency — requires careful attention to detail. In the past 30 years, the number of information-related jobs in the Washington, D.C., area has increased dramatically. However, large organizations that conduct these information processing activities create a demand for educated workers and, just as importantly, for additional innovations in handling information-related tasks. For this reason, several local universities have recently started degree programs in big data, data analytics, and cyber security. The “post-industrial” society which has exploded in the D.C. area in the past 30 years has also been growing globally. Around the world, new universities are being established and are improving. In any event, Washington, D.C., is well-positioned to be a leading city in this post- industrial era. Overall, the city-wide university that Etzioni described 30 years ago is a key player in defining and creating the nation and the world in the 21st Century. References Benen, Steve. 2011. Lobbyists Go Back to Writing Laws, Washington Monthly , March 18. Etzioni, Amitai. 1985. The World-Class University That Our City Has Become. The Washington Post. April 28. Garvin, David. 1991. How the Baldrige Award Really Works. Harvard Business Review. November-December, pp. 80-94. Spring 2015 8 Kamensky, John. 1999. National Partnership for Reinventing Government (formerly the National Performance Review): A Brief History. Washington, D.C. 20006. http://govinfo.hbrarv.unt.edu/npr/whoweare/history2.html Link, Albert N. and Scott, John T. 2012. On the Social Value of Quality: An Economic Evaluation of the Baldrige Performance Excellence Program. Science & Public Policy , 39, 5: 680-689. Schiff, Philip. 2013. Commentary: How to build a regional ‘Cyber Valley’ in Capital Business, The Washington Post Co. Smith, Hedrick. 2012. Who Stole the American Dream? Random House, 2012. U.S. Patent and Trademark Office. 2015. Performance and Accountability Report, http://www.uspto.gov/about-us/perfonuance-and- planning/uspto-annual-reports. Bio Stuart A. Umpleby is Professor Emeritus in the School of Business at The George Washington University in Washington, D.C. He may be contacted at umplebv@gwu.edu. www.gwu.edu/-umplebv. Washington Academy of Sciences 9 Does Speed Matter? The Employment Impact of Increasing Access to Fiber Internet Paul Lapointe Georgetown University, Washington, D.C. Abstract As internet technology continues to improve at a rapid pace, there is constant debate over the relative value of various internet connection technologies. In recent years, policymakers have debated over several important questions regarding broadband. What speed qualifies as high- speed broadband? How much public funding should be spent increasing access to broadband? And, what regulations to impose on internet providers? While the potential and proven benefits of high-speed internet are diverse, the economic impacts are often at the forefront of policy discussions. To date, most research into the economic impact of internet has focused on increasing access to and adoption of broadband internet in general, without emphasizing the speed of the broadband connections. This paper utilizes new data available as a result of the American Recovery and Reinvestment Act to examine the relationship between employment growth and access to fiber internet, currently seen as the gold standard of internet connections in terms of speed and reliability. Using data from the National Broadband Map, this study finds a positive association, within the United States, between increasing access to fiber and increases in employment and number of firms at the county level. This positive relationship provides evidence to policymakers that promoting access to fiber internet is a viable economic development approach. Introduction Although there is a strong consensus that high-speed internet is related to economic growth, many questions remain about what speed is optimal. As the internet becomes ubiquitous in the United States, attention has shifted from expanding access to the internet towards improving the connections that Americans have access to. Table 1 shows the percent of U.S. households that have access to different types of internet technologies. Almost all households have access to some form of internet connection, whether it is a fixed line connection, wireless internet, or satellite. Additionally, 95 percent of households have access to fixed line internet, including 87 percent that have access to a cable internet connection. The opportunity that remains is in expanding access to state of the art technologies such as optical fiber, where access is expanding in Spring 2015 10 recent years, but still remains out of reach for most American households, as only one in four households has access to it. Table 1. Percent of U.S. Households with Access to Internet Technologies in June 201 1 and December 2013 June 2011 December 2013 Change Any internet 99% 99% 0% Any fixed line internet 95% 95% 0% Cable internet 83% 87% 5% Optical fiber internet 17% 24% 7% Source: National 3roadband Map Over the past two decades, the policy focus has been on increasing broadband access and adoption. In the aggregate, these efforts have largely been successful. Broadband access (using the Organisation for Economic Co-operation and Development (OECD) definition of 256 Kbit/sec) in the United States has increased from 4.4 percent of households having access in 2000 to 19.9 percent of households in 2003 and 68.2 percent in 2010 (OECD, 2014). Now that broadband access is wider, many policymakers have shifted away from increasing access towards increasing speed. The demand for high speed is clear; when Google announced plans to pilot its Google Fiber networks, more than 1,100 communities across the country applied (Kelly, 2010). Absent private investment, some municipalities have dedicated vast tax payer resources to construct fiber networks of their own. Clearly, effort is being put into improving internet connections, yet there is little empirical evidence as to whether these ultra-high-speed networks promote growth beyond the benefits of more common speeds. The purpose of this paper is to examine the economic impact of fiber internet availability in the United States. Now, thanks to recent enhancements to the Federal Communication Commission’s (FCC) data collection strategies, researchers have access to data which allows the examination of the impact of fiber networks for the first time. By evaluating the economic impact of fiber internet, information can be provided to policymakers to help guide them in determining the amount of resources to invest in the technology. Washington Academy of Sciences 11 Literature Review There is a growing body of literature on the economic impact of high speed internet both in the United States and around the world; the general consensus is that high speed internet leads to economic growth (Qiang, 2009; Van Gaasbeck, 2008; Whitacre et al 2013; Kolko, 2012). The literature can be divided into research on differences in broadband technology across countries and differences in broadband technology within a single country. While this paper will focus on the effect of broadband differences in the United States, it is important to examine the literature in both areas to build a cohesive picture of the state of research on economic effects of broadband. International Literature As a whole, the literature on country-level effects of broadband technology shows that countries with higher levels of broadband penetration have generally higher levels of GDP growth. Czemich et al. (2009) used data from a panel of 25 OECD countries between 1 996 and 2007 to create a model — using pre-existing telephone and TV networks to predict maximum broadband penetration rates — to examine economic impact. They (2009) found a statistically significant positive relationship; a 10 percent increase in broadband penetration raised GDP per capita by 0.9- 1.5 percent. Similarly, Qiang and Rossotto (2009) used Information Communications and Technologies Development (ICTD) and World Bank data for 120 countries between 1980 and 2006 to understand how differing broadband penetration rates are related to GDP per capita growth. They estimated that a 10 percent increase in broadband adoption is associated with a 1.21 percent increase in GDP per capita for developing countries and a 1.38 percent increase for developed countries. However, Qiang and Rossotto (2009) caution that causality is not abundantly clear; that is, there could be a back and forth effect as increased wealth also increases the demand for broadband services. Koutroumpis (2009) attempted to account for the fact that broadband can both influence and be influenced by economic factors using a simultaneous equation model to identify the macro impact of broadband in 15 European Union countries between 2003 and 2006. He separated the increased demand for broadband caused by increased wealth from the economic growth caused by increased broadband usage with models that predict the supply and demand for broadband growth. After separating out these influences, there was still a Spring 2015 12 significant, positive relationship between broadband penetration and GDP per capita. National Literature Research within the United States has been building over time, with researchers using a variety of datasets and approaches to building models. While the approaches vary, there is a consensus that improvements in broadband technology are related to higher levels of employment, although findings are mixed on other economic indicators such as number of firms and income. The early literature in the United States focused on building cross- sectional panel models that take advantage of varying levels of technological development across regions of the country. Lehr et al. (2005) used data from the FCC form 477 and the Population Censuses and Establishments Surveys to investigate the effect of broadband presence (as a binary measure) on economic indicators such as employment, wages, and industry mix. Their model, which used data from 1998-2002, showed that in zip codes with mass-market broadband availability there was higher employment, more firms overall, and more firms in the IT sector. The broadband speed studied was 200 kilobits per second, speeds that would now be considered slow. Moreover, the Lehr et al. (2005) study showed the tradeoffs associated when choosing to study broadband at the state or community level in the United States. Crandall et al. (2007) built on this model with data from 2003-2005 to examine state level GDP growth associated with increased broadband penetration. While they found that higher levels of broadband penetration were associated with higher levels of GDP growth, the results were not statistically significant, which reinforces the notion that state-level data are too broad to study broadband in America. While several dependent variables of interest, such as GDP, are not available at smaller geographical units than the state, there is generally not sufficient variation between states in broadband availability to draw meaningful conclusions. While much of the research in the United States uses FCC data, two studies in 2007 corroborate the larger national studies using different data sources. Van Gaasbeck (2008) used cross-sectional panel household survey data from Scarborough Research to examine the potential employment effects of expanding broadband adoption in California. They found that increased broadband adoption was associated with higher employment but fewer establishments. Similarly, Shideler et al. (2007) Washington Academy of Sciences 13 looked at county-level effects for a single state, Kentucky. They focused on increased broadband availability, instead of broadband adoption. Using infrastructure data from providers collected through ConnectKentucky, they examined county-level employment growth and sector employment growth relative to broadband availability, controlling for past growth, education, unemployment, and road density. They found a positive, statistically significant relationship between broadband availability and total employment. While the limited scope of these studies restricts the applicability to broader national policies, they help to validate the general association between broadband and employment. In a qualitative analysis, Ezell et al. (2009) made the case for facilitating the development of internet with speeds of at least 20 Mbit/sec downloading and preferably 50 Mbit/sec or greater. While most policy efforts focus on increasing broadband adoption and availability, the authors encourage policymakers to consider efforts to increase speeds as well. They count fiber to the home, fiber to the node, and DOCSIS 3.0 cable as the most desirable fixed-line broadband delivery methods and 4G as the most desirable wireless delivery method. They point out that countries such as Japan, Singapore, South Korea, and Sweden are far ahead of the United States in terms of high speed internet, giving them an advantage in developing innovative web-based applications. In order for the United States to remain the global leader in internet based innovation, they contend that there needs to be a greater focus on increasing broadband speed. There has been a current focus on the impact of broadband expansion in rural communities in particular. While high speed internet has become standard in most urban and suburban communities, lower population density makes it much more costly for providers to expand into rural areas. Therefore, many policy initiatives have focused on how the government can play a role in expanding access in rural communities. Stenberg et al. (2009) match rural counties that had broadband by 2000 with those that did not based on a variety of characteristics in order to test a causal relationship between broadband and economic growth in rural counties. They aggregated FCC Form 477 data to measure broadband availability and found faster employment growth in counties with more availability. There is also evidence that counties which had early adoption of broadband experienced relative income growth, but this faded over time as broadband became more profuse. Whitacre et al. (2013) used data newly available from the National Broadband Map combined with adoption rates from FCC Form 477 to examine economic impacts of Spring 20 1 5 14 broadband expansion into rural communities. They used three different techniques to examine the relationship between broadband and economic health. The collective results indicated a positive relationship between rural economic indicators and broadband availability. They concluded that adoption thresholds had more of an impact than availability thresholds (Whitacre et at., 2013). In regards to the debate over whether to use adoption or availability as the key indicator of broadband penetration, Kolko (2012) made the case for availability. He pointed out that adoption rates can be influenced by economic growth more so than availability. Additionally, increasing availability is a more feasible approach for policymakers than increasing adoption rates. Using cross-sectional panel data from the FCC between 1999 and 2006, Kolko built a model to identify the impact of availability on local level employment and county-level labor market outcomes. He found a statistically significant, positive relationship between broadband expansion and local employment, but cautions that the increased employment is accompanied by increased population growth, resulting in no impact to employment rates. In 2013, NC Broadband hosted a research roundtable to discuss the state of research on the economic and community impact of broadband expansion (Feser et al., 2013). The final report suggested that there is a need for more research on specific broadband policies and investments at the margin; including increases in broadband speeds and reliability and use of new technology. This paper will attempt to fill some of that gap. It benefits from the requirement in the National Broadband Plan that states collect more detailed information on different technologies and speeds available at local levels. Using this new dataset, it is now possible to start evaluating whether or not incremental expansion of the presence of fiber technology is associated with increased economic growth. Study Hypothesis The central hypothesis being tested in this study is that, within the United States, increasing access to fiber internet connections is related to increased levels of economic growth, as measured by employment levels, number of firms, and income. Broadband, in general, can lead to economic growth in several ways. By connecting individuals and companies across the globe, the internet can make it easier for small and medium sized firms to do business with suppliers and customers that they otherwise would not have interactions with. Further, individuals are able to use the internet to Washington Academy of Sciences 15 connect with employers and potential work remotely for companies anywhere in the world, opening up more employment opportunities and facilitating virtual talent mobility. Lastly, we would expect a short-term increase in employment due to the fact that creating the connections requires the hiring of employees to dig up cables, install new lines, and provide on-going maintenance services. Because fiber internet provides a faster, more reliable connection that allows the almost-instantaneous transfer of large amounts of data, it is likely that these effects are enhanced beyond what would be expected with more common speed levels. Data Much of the prior literature in the United States used FCC form 477 data to understand where broadband technology was available. While this dataset provided a relatively complete picture, it did not offer insight into different speeds within each geographical region. As part of the American Recovery and Reinvestment Act of 2009, the National Broadband Map was commissioned. The National Broadband Map provided funding for each state to gather more detailed internet data. The methodology used by each state to obtain these data differs slightly, but there are set data fields that each state is required to provide. This semi- annual data release is what allows the examination being conducted in this study. The data are made available in several formats, such as the analyze tables that aggregate internet statistics by region with accompanying descriptive data that can be used in modeling efforts. The first analyze table to be released was in 201 1 and it has been released every six months subsequently. Dependent variables for this model will come from the Quarterly Census of Employment and Wages (QCEW) survey conducted by the Bureau of Labor Statistics (BLS). The QCEW provides county-level summaries of a variety of economic indicators, including employment, number of firms, and average annual pay, broken down into industry and sector. In order to match up with the National Broadband Map data, annual average survey data released between 2011 and 2013 will be used. Combining the National Broadband Map data with the QCEW data results in a dataset that contains 3,142 counties with 6 observations per county. As shown in Table 2, between the first and last time period, roughly two-thirds of the counties experienced an increase in access to fiber internet. Counties with an increase in access to fiber experienced substantially more employment growth than counties that did not, and also Spring 2015 16 had greater changes in the number of total firms and total average weekly pay. This provides some initial evidence of a positive relationship between access to fiber internet and employment growth; however, a simple difference in means comparison is not sufficient to draw policy conclusions from. There could be a variety of factors that contribute to both job growth and improved internet infrastructure. Further, different counties saw drastically different changes in internet access and employment growth. Table 3 breaks up the counties that experienced an increase in access into quartiles (based on percent of households with access to fiber). The relationship between the magnitude of the increase in access and the change in economic indicators is more complex than the binary comparison, although there are some indicators where there is clearly a positive correlation, such as number of total firms. This provides evidence for using a continuous rather than discrete or binary variable for access to fiber internet. Table 2. Comparing Economic Indicator Changes by Changes in Access to Fiber between June 201 1 and December 2013 Negative or No Change in Access to Fiber Positive Change in Access to Fiber Number of Counties 1,008 2,134 Average % Change in Employment 0.66% 2.49% Average % Change in Firms 1.31% 1.67% Average % Change in Average Weekly Pay 10.13% 10.98% Average % Change in Private Employment 1.89% 3.51% Average % Change in Private Firms 2.35% 2.31% Average % Change in Private Average Weekly Pay 15.59% 15.51% Source: National Broadband Map Methodology This study uses a two-way fixed effects regression1 to evaluate the relationship between access to fiber internet connections and economic growth. The model has fixed effects for county and for time. A fixed effects regression is superior to a simple cross-sectional model or a pooled ordinary least squares model in these circumstances because it allows the model to control for unmeasured characteristics of counties that may be Washington Academy of Sciences 17 correlated with access to fiber technology and influence measures of economic growth in addition to factors that were common across all counties for any given time period. If the hypothesis holds, counties that experience increases in access to fiber internet will have greater increases in employment than counties that have no change in high speed internet. While the fixed effects model will not definitively prove causality, it does provide a stronger case for causality than a cross-sectional model (Whitacre et al., 2013). Table 3. Comparing Economic Indicator Changes between 2011 and 2013 for Counties that Increased Fiber Access < 25% Change in Access 25% - 50% Change in Access 50% - 75% Change in Access >75% Change in Access Number of Counties 1,803 205 96 30 Average % Change in Employment 2.41% 1.78% 4.75% 4.67% Average % Change in Firms 1.48% 2.08% 3.61% 3.55% Average % Change in Average Annual Pay 11.02% 10.94% 10.20% 11.80% Average % Change in Private Employment 3.52% 2.05% 5.58% 6.32% Average % Change in Private Firms 2.07% 3.09% 4.66% 3.87% Average % Change in Private Average Annual Pay 15.55% 15.82% 13.62% 17.02% Source: National Broadband Map The independent variable of interest will be percent of households within a region that have access to fiber internet technology. Due to data limitations, the percent of households having access serves as a proxy for both individuals and businesses having access in that region. Because GDP is not available at the county level, the primary dependent variable will be employment, which is available. Additionally, the number of firms and average annual wages will be used in order to provide a more comprehensive overview of the economic impact. For both employment and firms, natural logs will be used so that the results are meaningful across counties of drastically different sizes. By running each model for both the private sector and total economy, fiber internet’s impact on the Spring 2015 18 private sector and the public and non-profit sector can be contrasted. Control variables for county demographics and access to cable internet are included to isolate the relationship between fiber internet and employment. Exhibit 1 shows the model and variables that are the main focus of this paper. Exhibit 1. Model Variables and Predicted Relationships Yit ~ Po + PiXut + P2X2U + P3X3U + P4X4U + Ps^sit + ai + at where: Variable Variable Name Definition Predicted Relationship Study Y Ln(total employment) The natural log of total employment Crandall, Lehr, Litan Y (Alternate) Ln(private employment) The natural log of private employment Crandall, Lehr, Litan Y (Alternate) Ln(total firms) The natural log of total firms Whitacre, Gallardo, Strover Y (Alternate) Ln(private firms) The natural log of private firms Whitacre, Gallardo, Strover Y (Alternate) Ln(total annual average wage) The natural log of total average weekly wages Whitacre, Gallardo, Strover Y (Alternate) Ln(private annual average wage) The natural log of private average weekly wages Whitacre, Gallardo, Strover XI Households with access to optical fiber The percent of households within a county that have access to optical fiber internet Positive X2 Ln(population) The natural log of the county population Positive Whitacre, Gallardo, Strover X3 Adults with bachelors or greater The percent of the population with a bachelor's degree or greater Positive Crandall, Lehr, Litan X4 Median Household Income The median income for households within a county Positive Whitacre, Gallardo, Strover X5 Households with access to cable The percent of households within a county that have access to cable internet Positive Whitacre, Gallardo, Strover “l County fixed effects Controls for unmeasurable and constant differences between counties Time fixed effects Controls for unmeasurable and constant differences between time periods Washington Academy of Sciences 19 The model will examine the relationship at a county level; Kolko (2011) showed that a state-level model is too aggregated to show statistically significant differences in access to broadband. While there is substantial variation in change in access to fiber internet at the state-level, the small sample size and fact that most states are clumped at the lower end of the spectrum would likely lead to a similar finding in this dataset. Table 4 presents state level fiber optic data. Figures 1 and 2 illustrate that there is much variation, at a county-level, in the level of access to fiber technology, providing a robust dataset on which to conduct analysis. Further, there is very little geographical concentration to where fiber is being deployed, which will allow the results of this model to be applied across all of the United States. The National Broadband Map began data collection in 2010; however, there were concerns over the quality of the first year’s data collection methodology; the data were cleaned up for subsequent years (Whitacre et al., 2013). Therefore, this study will examine data from each of the releases in 2011, 2012 and 2013. While a larger dataset would be ideal in order to understand the lasting effect of increasing access to fiber internet, available data are sufficient to provide early evidence on the relationship between access to fiber internet and economic indicators. Policymakers will not delay actions for the next few years in order to collect more data; neither should researchers. Results The results for the primary dependent variable, total employment, are displayed in Table 5. Column (1) shows a simple one-way fixed effects model with no control variables; the coefficient on access to fiber internet is highly statistically significant, with a t-statistic of over ten. When the natural log of population is controlled for in column (2), the coefficient and its significance do not change substantially; the R squared value rises from .017 to .949, though. This is as expected, as the overwhelming determinant of how many employed people are in a county will be population. In column (3), controls for changes in demographic characteristics are added in. While the inherent wealth and education of each county are absorbed by the unit fixed effects, adding these variables accounts for any changes in income and education level that may have occurred over the time period studied. We see that both of these controls are statistically significant, as we would expect since wealth and education are traditionally positively correlated with employment. Spring 2015 20 Table 4. Percent of Households with Access to Fiber by State, Ranked by Access to Fiber in December 2013, in both June 201 1 and December 2013 State Name Counties Access to Fiber June 2011 Access to Fiber Dec. 2013 Change in Access to Fiber Change in Employment Rhode Island 5 78.81% 97.05% 18.24% 2.26% Oregon 36 37.32% 73.98% 36.66% 5.30% South Dakota 66 64.78% 70.33% 5.55% 2.69% Montana 56 1.57% 65.30% 63.73% 3.42% New Jersey 21 54.71% 59.50% 4.79% 2.53% North Dakota 53 15.53% 59.40% 43.87% 16.23% New York 62 47.52% 57.76% 10.24% 4.34% Maryland 24 52.51% 55.89% 3.38% 2.41% Delaware 3 48.90% 50.00% 1.10% 3.66% Pennsylvania 67 45.67% 48.22% 2.55% 1.27% Utah 29 12.36% 46.10% 33.74% 8.77% Indiana 92 3.46% 44.65% 41.19% 4.58% Connecticut 8 6.01% 44.52% 38.51% 2.14% Virginia 134 38.90% 42.88% 3.98% 1.90% D.C. 1 20.13% 40.16% 20.03% 3.01% Florida 67 22.54% 37.80% 15.26% 6.50% Massachusetts 14 36.53% 37.06% 0.53% 4.23% Iowa 99 27.96% 28.19% 0.23% 3.35% Nebraska 93 2.10% 27.01% 24.91% 4.17% Tennessee 95 12.75% 22.91% 10.16% 5.10% Washington 39 22.08% 20.16% -1.92% 5.36% South Carolina 46 10.95% 19.38% 8.43% 4.17% Kansas 105 6.67% 18.68% 12.01% 3.36% Mississippi 82 15.55% 18.02% 2.47% 2.10% Georgia 159 9.21% 16.41% 7.20% 4.43% Vermont 14 14.95% 15.41% 0.46% 4.28% Minnesota 87 5.53% 15.38% 9.85% 4.14% California 58 13.23% 14.81% 1.58% 7.11% Illinois 102 0.24% 14.59% 14.35% 2.92% Nevada 17 1.82% 11.71% 9.89% 5.35% Wyoming 23 6.24% 10.74% 4.50% 2.01% Ohio 88 6.11% 10.52% 4.41% 3.84% Texas 254 6.73% 9.87% 3.14% 7.37% Missouri 115 5.68% 9.41% 3.73% 2.32% Louisiana 64 8.02% 9.16% 1.14% 3.46% Colorado 64 1.88% 9.08% 7.20% 7.12% Kentucky 120 4.87% 8.99% 4.12% 3.77% North Carolina 100 2.50% 8.84% 6.34% 4.81% New Mexico 33 2.02% 8.19% 6.17% 1.58% Idaho 44 3.42% 8.02% 4.60% 5.06% Oklahoma 77 1.92% 7.35% 5.43% 4.46% Hawaii 4 4.47% 6.41% 1.94% 6.31% Alabama 67 6.75% 6.12% -0.63% 2.07% Arkansas 75 3.28% 5.63% 2.35% 0.93% Wisconsin 72 1.95% 4.99% 3.04% 2.85% New Hampshire 10 1.12% 2.60% 1.48% 3.06% Michigan 83 1.29% 2.46% 1.17% 5.58% Arizona 15 6.64% 2.38% -4.26% 7.68% Alaska 29 1.82% 1.99% 0.17% -1.53% West Virginia 55 0.30% 1.93% 1.63% 1.51% Maine 16 0.27% 0.73% 0.46% 2.00% Washington Academy of Sciences 21 Figure 1. Access to 1 gig/sec Download Speed by County (2011) Source: National Broadband Map Figure 2. Access to 1 gig/sec Download Speed by County (201S) Source: National Broadband Map Spring 2015 22 Column (4) adds in a control for access to cable internet. This ensures that any association between access to fiber internet and employment growth is not actually due to the relationship between employment and increased access to internet in general. The coefficient on cable is surprisingly not statistically significant. Based on the body of literature, a positive and statistically significant coefficient on access to cable internet was anticipated. A possible explanation for this could be that during the time period in question, roughly $5 billion in stimulus funding was spent on expanding broadband access, much of which was spent on expanding access to cable internet in rural areas of the country. These areas that did not already have access to broadband likely were some of the hardest hit and last to recover from the recession, explaining why they lag behind in employment growth while experiencing an increase in access to cable internet. Columns (5) and (6) add in time-fixed effects. This is particularly important as the country was recovering from the Great Recession during this time, so employment growth could be the result of a generally positive economic trend. The time fixed effects may account for part of the coefficient for access to fiber internet, yet this coefficient is still statistically significant at a 99 percent confidence level. Finally, column (6) adds in robust standard errors to control for potential heteroscedasticity. A control variable for state level stimulus spending delivered through the National Telecommunications and Information Association (NTIA) was also used, although not shown. Adding in the control for NTIA stimulus spending had almost no impact to any of the other coefficients, perhaps because the only available data are not at the county level or accurate enough in terms of timing of implementation. Similar models were run for the other dependent variables of interest: total establishment count, total average weekly wages, private sector employment, private establishment count, and private average weekly wages; the results for model (6) are shown in Table 6. For the wage models, median household income is replaced by the log of total employment. Statistically significant coefficients are found for total and private employment and total and private establishment count. The coefficients on average weekly wages were significant until time-fixed effects were added in, which soaked up most of the coefficient and significance. Since the wages are in nominal values, the relationship depicted prior to adding time-fixed effects was likely due to inflation.2 Washington Academy of Sciences 23 Table 5. Regression Results for Total Employment Variables Mean (1) (2) (3) (4) (5) (6) Log Total Employment 9.134 %of Households w/ Access to Fiber 0.124 0.0369*** 0.0358*** 0.0263*** 0.0263*** 0.0132*** 0.0134*** (10.49) (10.20) (7.496) (7.489) (3.696) (2.80) Log (Population) 10.27 0.163*** 0.115*** 0.115*** 0.164*** 0.164*** (9.232) (6.463) (6.457) (9.142) (3.72) % of Pop w/ Bachelors or Higher 0.168 0.430*** 0.430*** 0.358** 0.356 (2.596) (2.595) (2.181) (1.01) Median Household Income 45,883 2.64e-06*** 2.64e-06*** 8.72e-07*** 8.74e-07* (17.70) (17.65) (4.710) (1.82) % of Households w/ Access to Cable 0.566 0.000447 -0.00197 -0.00202 (0.119) (-0.525) (-0.36) Constant 9.129*** 7.450*** 7.755*** 7.755*** 7.337*** 7.345*** (17,419) (40.97) (42.84) (42.83) (40.35) (16.64) Observations 18,848 18,848 18,848 18,848 18,848 18,848 R-squared 0.0170 0.9491 0.9258 0.9258 0.9550 0.9549 F stat 131.68 97.95 131.68 105.34 86.12 46.62 Number of counties 3,142 3,142 3,142 3,142 3,142 3,142 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1 Overall, the results show evidence of a strong positive correlation between the percent of households that have access to optical fiber internet in a county and the number of employed individuals and number of firms. Specifically, a 10 percent increase in the percent of households with access to fiber internet is associated with a 0.13 percent increase in total employment and a 0.1 percent increase in the number of firms. There is no evidence of a relationship between access to fiber internet and average weekly wages within a county. Spring 2015 24 Table 6. Coefficient on Percent of Households with Access to Fiber Internet for each Dependent Variable* Coefficient T-stat R-squared F stat Log Total Employment 0.01340 2.80*** 0.9549 46.62 Log Total Establishment Count 0.01030 2.26** 0.9315 30.25 Log Total Average Weekly Wages 0.00061 0.17 0.0165 540.63 Log Private Employment 0.01580 2.65*** 0.9389 66.82 Log Private Establishment Count 0.01068 2.09** 0.9188 41.43 Log Private Average Weekly Wages 0.00293 0.68 0.0535 847.74 *Each model controls for 2- way fixed effects (county and date) and demographics. Without a controlled or quasi-experiment, a causal relationship between access to fiber internet and employment growth cannot be claimed, but the results shown do support the theory that installing fiber internet can help job growth. While controlling for time and unit fixed effects and other controls helps to isolate the relationship between access to fiber and employment growth, there is still the possibility that there are unmeasured factors that influence both access to fiber and job growth. While state level NTIA stimulus spending is controlled for, county level spending cannot be controlled for due to data limitations. This creates a slight problem; while the source of funding for the increase in access to fiber is not the topic of this paper, stimulus funds had a specific goal of creating jobs and contractors typically had to lay out a plan for hiring additional employees as a part of their bid for stimulus funding. Therefore, if some of the infrastructure that led to the increase in fiber access was because of stimulus spending, it may have created more jobs than private investment, which does not have to meet any job creation criteria. While a better control for this would be ideal, it is unlikely that this is the primary cause of the positive relationship. As mentioned previously, most of the broadband stimulus spending went to expanding access to cable technologies, not optical fiber. Additionally, there is a possibility that job growth is driving access to fiber internet, rather than the other way around. The positive relationship could be due to internet service providers expanding into growing areas. While it is likely that some of the positive relationship can be attributed to this, it is unlikely to be the primary reason. Most of America still is without access to fiber internet, so service providers would be more likely to invest in areas where they already see demand rather than trying to predict where employment growth will be. Additionally, Washington Academy of Sciences 25 laying the infrastructure for fiber internet takes time and planning; since this model looks at six month intervals, it is unlikely that service providers saw employment growth in an area and were able to move in and offer fiber service within six months. A final critique of the model could be the relatively short-term time frame used. Policymakers are not concerned with much longer time frames than two years when investing heavily in internet technologies. Unfortunately, the relative newness of the National Broadband Map data set, and limitations of previous data collection efforts, limit the years that can be examined. As data collection efforts continue, researchers should continue to evaluate this relationship to test whether or not better internet leads to sustained growth, or if growth is merely temporary. Conclusion: Policy Relevance This study provides evidence that increasing access to state of the art internet like optical fiber and employment growth are related. Policymakers considering investments in improving internet technologies might consider these results when debating whether or not the cost of the investment is appropriate. This information is useful to policymakers at all levels of government, who have taken a variety of approaches to improving access to ultra-high speed internet networks. In January of 2015, the FCC changed its definition of broadband internet from offering download speeds of 4 Mbit/sec or greater to offering much faster download speeds of 25 Mbit/sec or greater. This was a highly contentious shift in policy that will impact how data are collected and what networks qualify for future public investments. Additionally, it may change how the FCC views the state of competition within the telecommunications industry, which could lead to other legislative, executive or even judicial actions (Brodkin, 2014). While this study does not address whether or not 25 Mbit/sec internet fosters more economic growth than 4 Mbit/sec internet, it does provide preliminary evidence that there could be a public interest in promoting faster internet speeds. This contradicts what many of the detractors of the FCC’s change in definition have argued; that the internet is fast enough and people do not benefit any more from speeds over 25 Mbit/sec than they would at lower levels. Another contentious policy area has been the recent development of local (partially or fully) tax-payer funded high speed fiber networks which offer internet speeds of up to 1 Gbit/sec (O’Toole, 2014). In response to these networks, some states have considered blocking these Spring 2015 26 efforts in order to prevent municipalities from crowding out private expansion into high-speed internet markets. This paper does not provide a cost benefit analysis of publicly-owned fiber optic networks, but does provide evidence that policymakers should consider when deciding whether or not these municipal fiber networks are wise uses of taxpayer funds. On the other hand, though, the non-significant coefficient on access to cable internet may provide evidence that pushing internet technologies into underserved regions may not unilaterally lead to economic growth. A more thorough examination of the specific investments made during the stimulus act would provide better insight into this, though, as that was not the primary focus of this paper. As part of the American Recovery and Reinvestment Act, the FCC developed the National Broadband Plan which outlines goals for internet infrastructure in America. In the plan, the FCC set ambitious long-term goals including providing affordable access to internet with speeds of 100 Mbit/sec or greater to at least 100 million homes and eventually ensuring that every American has access to affordable fiber internet. The results of this paper show that these are not unfounded goals, and there may be a public, economic interest in achieving the goals outlined in the National Broadband Plan. Endnotes A two-way fixed effects model controls for unmeasured variables that remain constant throughout the time period for each county, as well as variables that are common across all units for a single time period. These variables could potentially cause bias if left uncontrolled for. 2 Diagnostic tests indicated that fixed effects are preferable to random effects and suggested that robust standard errors are needed due to potential heteroscedasticity. References Brodkin, J. (2014, September 8). AT&T and Verizon say 10Mbps is too fast for ‘Broadband’. Crandall, R. W., Lehr, W., and Litan, R. E. (2007). The effects of broadband deployment on output and employment: a cross- sectional analysis of U.S. data. Issues in Economic Policy , 6 Washington, D.C.: Brookings Institution. Washington Academy of Sciences 27 Czemich, N., Falck, O., Kretschmer, T., and Woessmann, L. (201 1). Broadband Infrastructure and Economic Growth. The Economic Journal, 727(552), 505-532. Ezell, S. J., Atkinson, R. D., Castro, D., and Ou, G. (2009). The need for speed: the importance of next-generation broadband networks. Available at SSRN 1354032. Feser, E., Horrigan, J., and Lehr, W. (2013, March). Symposium Report: Findings from the Research Roundtable on the Economic and Community Impact of Broadband. NC Broadband. Fixed and wireless broadband subscriptions per 100 inhabitants. (Dec. 2013). OECD Broadband Portal. Retrieved on October 13, 2014 from http ://www . oecd. or g/ sti/broadband/ oecdbroadb andportal . htm . Kelly, J. (2010, April 16). Next steps for our experimental fiber network. Official Google Blog. Retrieved October 13, 2014. Kolko, J. (2010). A new measure of U.S. residential broadband availability. Telecommunications Policy, 34(3), 132-143. Kolko, J. (2012). Broadband and local growth. Journal of Urban Economics, 77(1), 100-113. Koutroumpis, P. (2009). The economic impact of broadband on growth: A simultaneous approach. Telecommunications Policy, 33(9), 471 - 485. Lehr, W. H., Osorio, C. A., Gillett, S. E., and Sirbu, M. A. (2005). Measuring broadband’s economic impact. In Broadband Properties, December 2005, 12-24. National Broadband Plan. (2010, March 17). O’Toole, J. (2014, May 20). Chattanooga’s super-fast, publicly-owned Internet. http://monev.cnn.com/2014/05/20/technology/innovation/chattano oga-internet/index.html Spring 2015 28 Qiang, C. and Rossotto, C. (2009). Economic Impacts of Broadband. In 2009 Information and Communications for Development: Extending Reach and Increasing Impact, (pp. 35-50). Washington, D.C.: World Bank. Selyukh, A. (2014, August 5). U.S. FCC asks if broadband should mean faster Internet speeds. http://www.reuters.com/article/2014/08/05/usa-intemet-speed-fcc- idUSL2N0QB15S20 140805 Shideler, D., Badasyan, N., and Taylor, L. (2007, August). The economic impact of broadband deployment in Kentucky. Federal Resewe Bank of St. Louis Regional Economic Development \ 3(2), 88-1 18. Stenberg, P., Morehart, M., and Cromartie, J. (2009). Broadband internet service helping create a rural digital economy. Amber Waves, 7(3), 22-26. Van Gaasbeck, K. A. (2008). A rising tide: Measuring the economic effects of broadband use across California. The Social Science Journal, 45(4), 691-699. Whitacre, B., Gallardo, R., and Strover, S. (2013). Rural broadband availability and adoption: Evidence, policy challenges, and options. Webinar conducted for the National Agricultural and Rural Development Policy Center, March 18, 2013. Bio Paul Lapointe is a recent graduate of the Masters of Public Policy program at Georgetown University’s McCourt School of Public Policy, where he focused his studies on domestic economic policy. Prior to Georgetown, he worked in distribution and logistics analytics in the private sector. Paul received his bachelor’s from the Robert H. Smith School of Business at the University of Maryland. The author may be contacted at plapoint50@gmail.com. The diagnostics for this study are available to readers upon request. Washington Academy of Sciences 29 Benjamin Banneker and Celestial Navigation: Just How Did They Know Where They Were, Then?1 Sethanne Howard USNO, Retired Abstract Benjamin Banneker was an American scientist of the late eighteenth century. He was a self-educated free black and became an expert in astronomy, mathematics, and surveying. Major Andrew Ellicott asked him to join the team surveying the original boundaries that became Washington D.C. This paper presents Banneker’s story — which is inspiring for all those who struggle against strong odds — and also discusses the techniques used in those days to determine latitude and longitude for surveying. Introduction Benjamin Banneker was bom in 1731 in Baltimore County, Maryland. He died in 1 806 in Baltimore County, Maryland. He lived his entire life on the family’s 100-acre tobacco farm near Oella, Maryland, a small hamlet which is near Catonsville, Maryland. His mother, Mary, was a free black, his father, Robert, was a freed slave from Guinea. Figure 1 shows a reconstmction of his log cabin. Figure 1. A reconstruction of Banneker’s cabin. 1 Presented at the Washington Academy of Sciences 2015 Annual Meeting and Awards Banquet, May 14, 2015. Spring 2015 30 One might think that free blacks were extremely rare. That is almost true. The state of Maryland had the largest number of free blacks of any of the states according to the 1830 census. There were over 52,000 free blacks in Maryland at that time. Figure 2 shows a woodcut of Banneker. It probably is somewhat idealized. It appeared on the cover of one of his publications, and in those days, publishers felt free to embellish their publications. Figure 2. Woodcut of Benjamin Banneker. Over the years the name Banneker has had various spellings, and currently it is Banneker. Washington Academy of Sciences 31 Banneker may have attended a few years of a nearby Quaker school; however, the vast majority of his learning was self-taught. He borrowed books from neighbors and studied them thoroughly. He loved mathematics and astronomy and became more than proficient in them; he was a skilled researcher, the equal of any of his contemporaries. Later in life he expressed that, “ The colour of the skin is in no way connected with strength of the mind or intellectual powers .” At age 22, Banneker built a wooden clock that struck the hours throughout his life. Clocks were not common items in the late 1700s. Local people came to marvel at his remarkable clock. He was not the first person to build a clock in the colonies, but he was one of the rare few who succeeded in doing so. Banneker believed in seeking peaceful resolutions to conflicts. Along with Dr. Benjamin Rush in 1792, he wrote a proposal to the Federal Government asking the government to establish a Peace Office with equal status to the Department of War. Almost 200 years later, the government set up the United States Institute of Peace. The U. S. Institute of Peace (USIP) works to prevent, mitigate, and resolve violent conflict around the world. USIP does this by engaging directly in conflict zones and by providing analysis, education, and resources to those working for peace. Created by Congress in 1984 as an independent, nonpartisan, federally- funded organization, USIP has more than 300 staff working at the Institute’s Washington, D.C. headquarters and on the ground in the world’s most dangerous regions. A Bit of Mathematics Benjamin Banneker loved math. He taught himself algebra, geometry, trigonometry, and spherical trigonometry. He drew great delight from creating math puzzles and solving them. One of his puzzles is: Divide 60 into four such parts that the first being increased by 4, the second decreased by 4, the third multiplied by 4, the fourth divided by 4 such that the sum, the difference, the product, and the quotient shall be one and the same number.2 2 The answer to the puzzle appears at the end of this paper. Spring 2015 32 A Bit of Astronomy Banneker liked to lie outside his log cabin all night watching the sky. So far as we know he did not own a telescope, although he knew how to use one (his neighbors, the Ellicotts, had telescopes). Those math skills he had learned he applied to astronomy. In 1788, he accurately predicted the solar eclipse of 1789. Predicting eclipses had long been the province of professional astronomers. The techniques for doing so were not commonly taught. He timed the eclipses of the Galilean satellites by Jupiter. Figure 3 shows those satellites with Jupiter. Callisto Ganymede Europa lo Jupiter Figure 3. The Galilean satellites of Jupiter. Surveying President George Washington asked Major Andrew Ellicott to survey the land we now call Washington, D.C. It became the District of Columbia in 1801. A Commission was set up to oversee the project. In 1791, Ellicott asked Banneker to be part of the survey team. Banneker was 59 when he joined Ellicott’s team. He was hired for his astronomical and mathematical knowledge. So he spent several months slogging through the swampy land putting down milestone markers. Washington Academy of Sciences 33 The original plan for Washington, D.C., called for 10 miles on a side square, equaling 100 square miles of land. For every mile of the perimeter, the survey team laid down boundary markers. Figure 4 shows the original boundary marker at 6980 Maple Street, N.W. Many of the boundary markers have disappeared over the years, but a few remain. Some of those have Banneker’s name on them. Figure 4. Boundary Marker on Maple Street, N.W. In 1846, Washington, D.C., gave the Virginia portion of the District of Columbia back to Virginia, leaving the District we have today. George Washington asked l’Enfant to design the new city. F Enfant drew up a plan for the city (see Figure 5), but ran afoul of the Commission overseeing the project. He left the project. Major Ellicott then drew up a city plan (see Figure 6) that was used to construct the original city. Spring 2015 34 Figure 5. l’Enfant’s plan for the city. I. \M v i 111! jlltljill r* _• w « 5 ji 4 jj ii |1| Hi i |}]Li =! h -m Its! i Washington Academy of Sciences jV.I/,S OA /‘o/.K.V 35 Note that Major Ellicott set 0° longitude at the Capitol Building. The 0° longitude meridian was set by many nations at many places over the centuries. In most cases, however, the meridian at Greenwich, England, was used as 0° longitude because England ruled the seas in the eighteenth century. It was not until 1884 by international treaty that Greenwich was chosen as the permanent 0° longitude — the Prime Meridian. VI. AN .' rL ifiiq ■* *Wdlh»iMl0n - in i hr Trrrilory <4 CpIubiIma. — .w-« V M ■ -i __ I '//(G/.s l. / --T/ . »/. i/ti'L. f \ fl llnurt itmrruc MB V. tVr. rft.h . V/;. 'tiRXMHXT. .rrF K“ r^EEEF ZeEMF'EFJL'.F- — rrr r.-srs [■ — n ■ i — p^- FTIRFEWEV -■ t /UFERFFFFFI rZF ,, FFFFr^-F &■*< r^'~F1FFF0PfiFFP.r'!^r CiKORGR Town. .^njp^pri^BkEB FIF .^nUFRPEFPrj.r- F~ ■7nF«^3iHFH~iE>np p’fB^crFF.^-r— .. ( \ -Efc igBBEJBEEte.'t—JsSfTgBBCgEEEEfe ‘ ’ ^ ✓ ■■ i rr v4Ck n-rr-r^ raui rrr k i / ■ trctEferr-RRFr ^ «■ Ere^liBewHrwi f*i t>. v\ : i^EEEFf.: i afig^v r- ^CRFEFaSRSfera, jObfmuuiono '>■ ocpliiniilorv of llir plan. A\V- JjrraStli .if tin $»irrris. Figure 6. Major Ellicott’s city plan. The Nitty-Gritty of 1791 Surveying To survey is to measure the latitude and longitude of the perimeter of the land under consideration. Today surveyors use GPS, but in 1791 GPS did not exist. Surveyors turned to events in the sky to determine latitude and longitude. This means astronomy. Spring 2015 36 It takes two coordinates to find a location on a sphere [we do all agree that we live on the surface of a sphere, the Earth]: a left/right one and an up/down one. Figure 7 illustrates the concept. Figure 7. The concept of latitude and longitude. The up/down is latitude which runs from 0° at the equator to ±90° at the poles. The left/right is longitude which runs from 0° to 180° East and 0° to 180° West. We call the latitudes circles of latitude ; we call the longitudes meridians of longitude. See Figure 8 for an illustration. Latitude North 90 (+) 30 South (-) Longitude 0 Prime meridian Figure 8. Illustration of the circles of latitude and meridians of longitude. Washington Academy of Sciences 37 Let us start with latitude. People knew how to find their latitude by the 2nd century BCE. Start with a sextant and sight along your horizon. In the northern hemisphere then find Polaris, the North Star — no telescope needed. Measure the altitude of Polaris (how far you tilt upward with your sextant) with the sextant. Your latitude is 90° minus the altitude of Polaris. It is a simple procedure. Figure 9 illustrates the concept. Latitude uses angles ranging from 0° at the equator to ±90° at the poles. zenith S N altitude nadir Figure 9. How to measure the altitude of Polaris. Spring 201 5 38 Longitude, on the other hand, is not so simple. Someone must define a starting point for 0° longitude. The British colonies (and the new United States) used Greenwich as the Prime Meridian. The time at Greenwich is known as Greenwich Mean Time (GMT) or Universal Time (UT). Longitude uses both time and angle. That begs the question of how longitude relates time to angle. The Earth is not sitting still; it is spinning on its axis. This does not matter for measuring latitude. It does matter for longitude. The Earth rotates once around each day, or 360°/day. A day has 24hrs. Work your way down from this to the Earth rotates 15° in lhr. So if you are at the equator, you are spinning with the Earth at 1675 km/hr. If you can measure your local time and the time at Greenwich, you can get your longitude. For example, if it is 1800hrs at Greenwich when it is 9h20m on the same day, then your local time is 8h and 40m behind Greenwich Time. Convert that value to angles, and you have your longitude: Long = - (8x15°) + ( 40 — xl5° Uo y = -[120° + 10°] = -130° = 130° West However in the 1700s, one could not call Greenwich from Washington, D.C., and ask the time. So the survey team needed to use events in the sky. They did have an ephemeris. An ephemeris is a time- ordered list of future positions of the Sun, Moon, planets, satellites, and stars (at 0hrs GMT). Astronomers were paid to develop ephemerides for various places. This took a great deal of mathematical skill. Most travelers took an ephemeris with them on their travels. An Almanac contains an ephemeris along with other important information such as holidays, eclipses, sunrise, and sunset. The Federal Government still publishes the Astronomical Almanac and the Nautical Almanac each year. An example of an ephemeris for Mars for the month of June 2015 looks like this: Target body Start time Stop time Step-size Date (UT) name: Mars : 2015-May-31 00:00:00.0000 UT : 2015-Jun-30 00:00:00.0000 UT : 1440 minutes HR:MN R.A. DEC Washington Academy of Sciences 39 ^ •X' *1* *1* *1* <1* «L* si# si* vl/ si* si* si* si* si* si* si* si* sis si* si,* si* si* si* si* si* si* i* si* si* si* si* si* ^L* si* si* 'I' *x* *T* 'T* 'T' /T' -T* *T* 'T* 'T* 'T' ^ #ys #ys #ys *Js *Js *Js *Js ^ *ys *Js *Js *ys *Js *Js *Js *Js *Js *|S *Js *Js *[S *ys #*jv *|S *js *[S 'p 'P 2015-May-31 00:00 2015-Jun-01 00:00 2015-Jun-02 00:00 2015-Jun-03 00:00 2015-Jun-04 00:00 2015-Jun-05 00:00 2015-Jun-06 00:00 2015-Jun-07 00:00 2015-Jun-08 00:00 2015-Jun-09 00:00 2015-Jun-10 00:00 2015-Jun-l 1 00:00 2015-Jun-12 00:00 2015-Jun-13 00:00 2015-Jun- 14 00:00 2015-Jun-15 00:00 2015-Jun-16 00:00 2015-Jun-17 00:00 2015-Jun- 18 00:00 2015-Jun-19 00:00 2015-Jun-20 00:00 2015-Jun-21 00:00 2015-Jun-22 00:00 2015-Jun-23 00:00 2015-Jun-24 00:00 2015-Jun-25 00:00 2015-Jun-26 00:00 2015-Jun-27 00:00 2015-Jun-28 00:00 2015-Jun-29 00:00 2015-Jun-30 00:00 04 46 . 02.43 +22 , 50 i42.4 04 49 00.96 +22 56 22.6 04 51 59.52 +23 01 50.0 04 54 58.13 +23 07 04.7 04 57 56.76 +23 12 06.7 05 00 55.42 +23 16 55.8 05 03 54.10 +23 21 32.1 05 06 52.80 +23 25 55.6 05 09 51.50 +23 30 06.2 05 12 50.20 +23 34 04.0 05 15 48.89 +23 37 49.0 05 18 47.57 +23 41 21.1 05 21 46.22 +23 44 40.3 05 24 44.84 +23 47 46.7 05 27 43.41 +23 50 40.2 05 30 41.92 +23 53 20.9 05 33 40.37 +23 55 48.8 05 36 38.75 +23 58 03.8 05 39 37.03 +24 00 06.1 05 42 35.22 +24 01 55.5 05 45 33.29 +24 03 32.1 05 48 31.25 +24 04 56.0 05 51 29.08 +24 06 07.2 05 54 26.78 +24 07 05.7 05 57 24.32 +24 07 51.5 06 00 21.71 +24 08 24.6 06 03 18.94 +24 08 45.1 06 06 15.99 +24 08 53.1 06 09 12.86 +24 08 48.5 06 12 09.54 +24 08 31.4 06 15 06.03 +24 08 01.9 The coordinates are right ascension and declination — two common astronomical coordinates. There are two important clocks one needs for surveying in the 1700s: a clock keeping GMT and a clock keeping local time. If the GMT clock stops, it is very difficult to retrieve the correct GMT (one cannot call Spring 2015 40 home). If the local clock stops, it is not as difficult to retrieve the local time but it does take some effort. Banneker had the responsibility for keeping the clocks wound. This was a vital position to have. Now they needed a celestial event to time. There were a few schemes used in the 1700s — most were not very precise. One celestial event that showed promise was the planet Jupiter eclipsing the four Galilean satellites (the four brightest moons of Jupiter, discovered by Galileo), see Figure 3. They had an ephemeris for the eclipse times for these satellites. Today you can use a smart phone app to get the ephemeris for the Galilean satellites. Both the Mason-Dixon Line and the boundary for Washington, D.C., were set using the eclipses of the Galilean satellites. Banneker worked on the survey team for just a few months. Ill health drove him back to his farm. He continued to compute ephemerides and, beginning in 1792, published a series of six Almanacs. They sold in six cities in four states for the years 1792 through 1797: Baltimore; Philadelphia, Pennsylvania; Wilmington, Delaware; Alexandria, Virginia; Petersburg, Virginia; and Richmond, Virginia. They were best sellers at the time. Today, very few exist. The Maryland Historical Society has one. The cover for the 1792 edition is shown in Figure 10. People who did not have clocks depended on an Almanac to give them sunrise and sunset times so they could tell the time of day. Summary Benjamin Banneker was an American scientist of repute. As a testament to his reputation, the Federal Gazette wrote the following obituary: “Mr. Banneker is a prominent instance to prove that a descendant of Africa is susceptible of as great mental improvement and deep knowledge into the mysteries of nature as that of any other nation.” There are some who say that his intellect matched that of Ben Franklin. There are many schools named after Banneker, and the Benjamin Banneker Museum and Park is maintained by Baltimore Recreation and Parks. Its address is 300 Oella Avenue, Catonsville, MD 21228. Washington Academy of Sciences 'MMHI 41 r Benjamin Banneker’s PENNSYLVANIA, DELAWARE, MARYLAND and VIRGINIA EPHEMERIS, For the YEAR of our LORD, 1792; Being BISSEXTILE, or LEAP-YEAR, and the Six- teenth Year of AMERICAN INDEPENDENCE, which commenced July 4, 1776. Containing, the Motions of the Sun and Moon, thetrut Places and AfpcCts of the Planets, the Riling and Setting of the Sun, and the Riling, Setting and Southing, Place and Age of the Moon, See. — The Lunations, Conjunctions, Eclipfes, Judgment of the Weather, Feftivals, and other remarkable Days ; Days for holding the Supreme and Circuit Courts of tin United States , as alfo the ufual Courts in Pennfylvania , Dela- ware, Maryland, and Virginia. — Also, feveral ufcful Tables, and valuable Receipts. — Various Selections from the Com- monplace-Book ot the Kentucky P Lilcf ] her , an American Sage j with interefting and entertaining Elfays, in Profe and Verfc — the whole comprifing a greater, more pleafmg, and ufeful Va riety, than any Work of the Kind and Price in North- America. BALTIMORE: Printed and Sold, Wholefale and Retail, b> William Goddar d and James Angell, at their Print- ing-Office, in Market- Street. — Sold, a!fo, by Mr. Joseph Crukshank, Printer, in Market-Street , and Mr. Daniei' Humphreys, Printer, in Soutb-Frcnt-Street, Philadelphia and by MefTrs. Hanson and Bond, Printers, in Alexandria I Figure 10. 1792 cover for the Banneker Almanac. Spring 2015 42 Bio Sethanne Howard is an astronomer and retired Chief of the Nautical Almanac Office at the U.S. Naval Observatory. She maintains her research field of interacting galaxies. As the first woman to receive a bachelor’s degree in physics from the University of California, Davis, she went on to get a master’s degree in nuclear physics from Rensselaer Polytechnic Institute, and a PhD in astrophysics from Georgia State University. She worked at several astronomical observatories, at NASA managing operational astrophysical satellites, at NSF as Program Officer for Extragalactic Astronomy and Cosmology, and finally at the U.S. Naval Observatory. The answer to the math puzzle presented earlier in this paper is: W = 5.6 is the first part X = 13.6 is the second part Y = 2.4 is the third part Z = 38.4 is the fourth part W + X + Y + Z = 60 W + 4 = X- 4 = Y*4 = Z/4 = 9.6 One needs to solve the set of simultaneous equations to get the solution. Washington Academy of Sciences 43 Washington Academy of Sciences Awards Program 2015 Background The purpose of the Washington Academy of Sciences, which was founded more than a century ago in 1898, is to encourage the advancement of science and “to conduct, endow, or assist investigation in any department of science.” To recognize scientific work of distinction, the Academy gives awards annually to scientists who work in the greater Washington, D.C., area. The awards are presented by colleagues at the academy’s annual business meeting and awards ceremony.1 The public is invited to help celebrate and recognize the extraordinary achievements of the honored scientists and engineers, so the Academy hosts a formal Business and Awards Banquet in the Washington area. At this ceremony, the nominating colleague gives a short 3 -minute introduction describing the awardee, and the awardee must be present to accept the award, but the tradition of requiring formal acceptance speeches ended back in 1955. Photo: A1 Teich Washington Academy of Sciences annual Awards Banquet at the conference center of the National Rural Electric Cooperative Association (NRECA) in Arlington, Virginia, May 14, 2015. 1 Per the Academy’s by-laws, the annual business meeting takes place by the third Thursday in May, and usually consists of brief reports by the outgoing and incoming presidents and an audit report. Spring 2015 44 Awards Program Early History While the Academy passed its centennial year in 1998, the Academy’s Awards Program has featured 75 years of achievement at this point in time. It’s interesting to recount the history of the program which began in 1940. The Academy’s Bylaws had been amended the previous year to permit the Academy to award “medals and prizes . . . [for] scientific work of high merit.” At that time, 1939, the Academy’s Board of Managers established awards for noteworthy accomplishments during the year by young scientists — no more than 40 years old — in the biological, physical, and engineering sciences. A proposal to raise the age limit to 45 for the Biological, Engineering, and Physical Sciences categories was rejected in 1953. The requirement that “candidates shall not have passed their 41st birthday” was dropped later on in 1982, and 1983 was the first year in which an award for a Distinguished Career in Science was given. Some Award Traditions The year 1956 marked the first year that more than one award was presented in a given category. In 1961, the Board of Managers officially encouraged granting more than one award in any given category should multiple qualified candidates exist. Traditionally, the Academy’s awards have been given for work done in the Washington, D.C., area. Since the Washington Academy of Sciences was incorporated in 1898, the year 1998 marked the Academy’s centennial year and the D.C.-area tradition was waived during that year — as some awards were given to individuals affiliated with organizations outside the D.C. area." In 1998, the Academy gave fourteen Centennial Awards for Lifetime Achievement in Science, including awards in Science Policy, Technology Policy, and History of Science. The next year, 1999, was the first non-centennial year in which the award for Science Policy was given. The History of Science award was not given again until 2012, which was the same year Service to Science was first awarded. An award for Lifetime Achievement in the Public Understanding of Science was first made in 2014. 2 The Washington Academy of Sciences founders included Alexander Graham Bell and Samuel Langley, Secretary of the Smithsonian Institution from 1887 to 1906. Washington Academy of Sciences 45 Establishment of the Education and Teaching Awards The first year an award was given for the Teaching of Science was 1952. For this award category, and for this category only, the age limitation of 40 years was waived. This “special award” was given in 1952 and 1953. The award category for the Teaching of Science was officially established in 1956. In 1976, the Berenice Lamberton Award for the Teaching of Science in High Schools was established. Lamberton was a professor at Georgetown University with a long-time interest in education. The Washington Academy of Sciences’ Junior Academy was initially set up by Lamberton and others at Georgetown. The Leo Schubert Award for College Teaching was established in 1979. The year 2000 was the first year in which awards were given for Achievement in Education and Teaching of Science in Middle Schools. In 2002, the Board of Managers, acting on the recommendation of the Awards Committee, established the Marilyn Krupshaw Award for Non-Traditional Education/Teaching. The award was named in honor of the long-time leader of George Washington University’s Science and Engineering Apprentice Program (SEAP) for high school students, sponsored by the U.S. Department of Defense (DoD). The award was presented for the first time in 2004. And in 2005, a special award was given for Service to Science Education. Nomination Process The Academy welcomes nominations for its Awards Program each year. The following is a complete list of the award categories as established by the Academy’s Board of Managers: • Distinguished Career in Science • Biological Sciences • Engineering Sciences • Physical Sciences • Health Sciences • Behavioral and Social Sciences • Mathematics and Computer Science • Krupsaw Award for Non-Traditional Teaching • Lamberton Award for Teaching of Science in High School • Leo Schubert Award for Teaching of Science in College Spring 2015 46 • Special Award ( e.g ., science policy, lifetime achievement in education) To carry out the Awards Program each year, the Academy’s Vice President for Membership appoints an Awards Committee which sets the submission dates for the year. Please watch the Academy’s website, www.washacadsci.org, for those deadlines, typically in early Spring. To nominate an individual, print and complete the Nomination Form that is available at the website, and mail it directly to the Awards Committee as indicated on the form. The Awards Committee typically uses the standard categories that appear on the nomination form, but when necessary, the Special Award category may be used to include other categories. For example, an award for Mathematics was first given in 1960. This other award category was expanded to Mathematics and Computer Sciences in 1979. The Academy first made an award for Behavioral Sciences in 1976; this other award category was changed to Behavioral and Social Sciences in 1987. 4 Awards were first given for two other categories — Health Sciences and Environmental Science — in 1997, and later for Public Health. The year 2000 was the first year awards were given for the categories of Anthropology and Astronomy. Back in 1961, the Board rejected a proposal that an Earth Sciences award category be instituted; it was not until 2014 that an award for Lifetime Achievement in Natural Resources Sciences was made. 2015 Annual Banquet At the Washington Academy of Sciences Annual Business and Awards Banquet on May 14, 2015, the Academy’s ceremony honored an illustrious group of individuals for their work in physical, biological, and engineering sciences and other areas. Ronald Colie received the 2015 award for Distinguished Career in Science in recognition of his “lifetime work and major contributions in radionuclidic metrology. Within the world of radioactivity measurements, it is almost impossible to hear the words ‘radon,’ ‘uncertainty’ or 3 Since 1 990, additional awards have also been given at the annual awards ceremony for special recognition of Service, or Meritorious Service, to the Washington Academy of Sciences; however, these awards go through a different process. 4 Ainitai Etzioni, noted earlier in this issue of the Journal of the Washington Academy of Sciences, was a recipient of this award in 1988. Washington Academy of Sciences 47 ‘metrologisf without thinking of the name Dr. Colie.” He is a specialist in nuclear radiochemistry and the development of standards, and he and his collaborators developed methods to analyze and standardize brachytherapy sources, pellets of radioactive material designed to be implanted in the body at sites requiring direct radiation exposure. The Academy presented its 2015 award for Distinguished Career in Engineering Sciences to Dr. Ram Duvvuru Sriram in recognition of “contribution and technical leadership in developing computational tools and techniques for engineering design and for enabling interoperability of CAD/CAM/CAE systems.” A 2015 award for Physical and Biological Sciences was presented to Dr. Marcus Cicerone in recognition of “establishing and pioneering the use of Broad Band Coherent Anti-Stokes Raman Spectroscopy imaging and establishing exquisite optical techniques for examining the dynamics of proteins and other biological molecules in the glassy sugar matrices commonly used for their preservation.” The Academy’s 2015 award for Biological Sciences was presented to Dr. Paul M. Peterson in recognition of being a “tireless and prolific taxonomist, collector, and publisher who has extensively revised the classification of the large grass subfamily Chloridoideae and its genera, and is leading the effort to prepare a DNA database for the grasses of North America and noxious weeds for the Bar Code of Life.” The Academy presented its 2015 award for Engineering Sciences to Dr. Robert Gover in recognition of “work at the Naval Research Laboratory on the development, implementation, and application of high- fidelity physics-based digital models for the development of optimized Electronic Warfare countermeasures against modem anti-shipping cruise missiles.” A 2015 award for Physical Sciences went to Mr. Gregory Strouse in recognition of “international leadership in high-precision temperature metrology, and innovative contributions to next-generation temperature sensors.” The Krupsaw Award for Non-Traditional Teaching was presented in 2015 to Ms. MaryBeth Petrasek in recognition of her “teachings in the techniques of medicolegal death investigation and forensic pathology to young people.” A 2015 award was also presented to Dr. Sally Rood in special recognition of Service to the Academy for “momentous work as editor of Spring 2015 48 the Journal of the Washington Academy of Sciences and coordination of the agreement to begin the process of digitally preserving more than 100 years of the Journal’s published works.” Photo: A1 Teich Lisa Karan presenting the award for Distinguished Career in Science to Ronald Colie. Photo: A1 Teich Award for Distinguished Career in Engineering Sciences presented to Ram D. Sriram (right) by Steven Fenves. Washington Academy of Sciences 49 Photo: A1 Teich Award for Physical and Biological Sciences, presented to Marcus Cicerone by Laurie Locascio. Photo: A1 Teich Award for Biological Sciences, presented to Paul Peterson (left) by Chris Puttock. Spring 2015 50 Photo: A1 Teich Award for Engineering Sciences, presented to Robert Gover (left) by Douglas Fraedrich. Award for Physical Sciences, presented to Gregory Strouse (right) by Gerald Fraser. Washington Academy of Sciences 51 Photo: A1 Teich Krupsaw Award for Non-Traditional Teaching, presented to MaryBeth Petrasek by Anne Cupero (left). Photo: A1 Teich Special recognition for Service to the Washington Academy of Sciences was presented to Sally Rood (right) by Master of Ceremonies Terrell Erickson. Spring 2015 52 Washington Academy of Sciences 53 1 200 New York Ave. Suite 113 Washington DC 20005 wvwv. wa shacadsci.org Addendum* to Washington Academy of Sciences 2014 Membership Directory M=Member; F=Fellow; LF=Life Fellow; LM=Life Member; EM=Emeritus Member; EF=Emeritus Fellow Adkins, Michael K. (Mr.) 4143 Elizabeth Lane, Annandale VA 22003 (M) Arif, Muhammad (Dr.) National Institute of Standards and Technology (NIST), 100 Bureau Drive, MS 8460, Gaithersburg MD 20899-8460 (M) Berry, Jesse F. (Mr.) 2601 Oakenshield Drive, Rockville MD 20854 (M) Boisvert, Ronald F. (Dr.) National Institute of Standards and Technology (NIST), 100 Bureau Drive, MS 8910, Gaithersburg, MD 20899-8910 (F) Brown, Elise A. B. (Dr.) 681 1 Nesbitt Place, Mclean VA 22101-2133 (LF) * These twenty names were inadvertently omitted from the Academy’s 2014 Membership Directory in the Winter 2014 issue of the Journal of the Washington Academy of Sciences, so we are printing them here instead of waiting to include them in the 2015 membership listing. As indicated in the inside cover of each quarterly Journal, the last issue of the year contains a directory of the current membership of the Academy. Spring 2015 54 Buford, Marilyn (Dr.) 3073 White Birch Court, Fairfax VA 22031 (F) Caws, Peter J (Dr.) 2475 Virginia Avenue, NW, Apt. 230, Washington DC 20037 (M) Cupero, Jerri Anne (Dr.) 2860 Graham Road, Falls Church VA 22042 (F) Danner, David L. (Dr.) 1364 Beverly Road, Suite 101, McLean VA 22101 (M) Elster, Eric Andrew (Dr.) 3223 Geiger Avenue, Kensington MD 20895 (F) Hollinshead, Ariel (Mrs.) 23465 Harbor View Road, #622, Punta Gorda FL 33980-2162 (F) Jayarao, Arundhati (Dr.) 881 1 Trafalgar Court, Springfield VA 22151 (M) Kaufhold, John (Dr.) 4601 N. Fairfax Drive, Suite 1200, Arlington VA 22203 (M) Martin, Charles R. (Dr.) P.O. Box 98521, M/S NLV085, Las Vegas NV 89193 (F) Mittleman, Don (Dr.) 4650 54th Avenue S., Apt. 57B, St. Petersburg FL 33711-4638 (F) O’Hare, John J. (Dr.) 108 Rutland Boulevard, West Palm Beach FL 33405-5057 (EF) Sozer, Amanda (Dr.) 4707B Eisenhower Avenue, Alexandria VA 22304 (M) Snieckus, Mary (Ms.) 1700 Dublin Drive, Silver Spring MD 20902 (M) Washington Academy of Sciences 55 Williams, Jack (Dr.) 6022 Hardwick Place, Falls Church VA 22041 (F) Williams, Tenisha (Ms.) 1209 7th Street, NW, Washington DC 20001 (M) Wu, Keli (Mr.) 360 Swift Avenue, Suite 48, South San Francisco CA 94080 (M) Spring 2015 56 Washington Academy of Sciences 57 In Memoriam Burton G. Hurdle (1918-2015) Burton Garrison Hurdle, a Fellow of the Washington Academy of Sciences, passed away peacefully on March 4, 2015. He was a research physicist with the Naval Research Laboratory’s Acoustics Division beginning in 1943 for 50 years. Hurdle was bom in 1918 in Roanoke, Virginia, the son of Grover Cleveland Hurdle and Bronna Rene (Garrison) Hurdle. He was raised during the Great Depression and graduated from Jefferson Senior High School in Roanoke in June 1936. After graduation from high school, he went to work for the Norfolk and Western Railway that had its headquarters in Roanoke. In that period, he started taking night classes and then switched to becoming a full time undergraduate student at Roanoke College. In 1941, he received his B.S. degree in physics, with a minor in mathematics, and then enrolled at Virginia Polytechnic Institute for graduate studies. He intended to major in mechanical engineering at Virginia Tech, but after only two weeks decided to major once again in physics. While he was studying for his Master’s Degree in physics he taught some classes in the university’s Mathematics Department to supplement his income. He also had an industrial fellowship with the Standard Register Company of Dayton, Ohio. Although he was within about a year and a half from receiving a Doctorate in Physics, he left the university to join the U.S. Navy for the war effort, and so was awarded a M.S. degree in General Physics at that time. While at Virginia Tech, he had interviewed with recruiters from the Naval Research Laboratory (NRL). After considering several other potential job Spring 2015 58 opportunities, he accepted a position in the NRL Sound Division as a Research Physicist and started work there in 1943. NRL was doing much applied research then in support of the War effort. His doctoral thesis topic was on the subject of acoustic interference fields in the ocean. Hurdle worked under all five Superintendents of the Acoustics Division. His first supervisor at NRL was Dr. Raymond Steinberger, and his early senior NRL colleagues included Harvey Hayes, Raymond Steinberger, and Prescott Arnold who were all Harvard-educated scientists. Hurdle briefly left NRL during the period 1947-1949 to work at Engineering Research Associates’ Physics and Chemistry Division in Arlington, Virginia. During this period, he worked on several research projects including investigations of the sound speed and absorption in liquids using an interferometer; development of methods for calibration of accelerometers using free-free bars; and development of methods for calibrating acoustic pressure gauges and impulse gauges for use in measuring the propagation of elastic energy in soil and rock. Hurdle completed his Ph.D. at a later time, during work in the United Kingdom. In addition to being a Fellow of the Washington Academy of Sciences, Dr. Hurdle was also a Fellow of the Acoustical Society of America (ASA). He served the ASA in various capacities including the Membership Committee, the Underwater Acoustics Technical Committee, the Nominating Committee, and the Publications Policy Committee. Dr. Hurdle was also a member of Sigma Xi. He served as Associate Editor of the U.S. Navy’s Journal of Unden\>ater Acoustics (1979-2004). He also served as General Chairman and Session Chairman at meetings of the U.S. Navy Symposia on Underwater Acoustics. Dr. Hurdle received numerous awards and commendations including the Alan Berman Research Publication Award for “The Nordic Seas” in 1985 and the Navy Superior Civilian Service Award in 1987. In 1998, he was the recipient of the Distinguished Technical Achievement Award from the Oceanic Engineering Society (OES) of the Institute of Electrical and Electronic Engineers (IEEE). He was cited for his outstanding contributions to understanding the oceanography and acoustics of the Nordic Seas. Washington Academy of Sciences 59 Washington Academy of Sciences 1200 New York Avenue, NW Room 1 1 3 Washington, DC 20005 Membership Application Please fill in the blanks and send your application to the Washington Academy of Sciences at the address above. We will contact you as soon as your application has been reviewed by the Membership Committee. Thank you for your interest in the Washington Academy of Sciences. (Dr. Mrs. Mr. Ms.) Business Address Home Address Email Phone Cell Please indicate: Preferred mailing address Type of membership Business Regular Home Student Schools of Higher Education Attended Degree Dates Present Occupation or Professional Position Please list memberships in scientific societies - and include office held: Spring 2015 60 Journal of the Washington Academy of Sciences Instructions to Authors The Washington Academy of Sciences publishes its interdisciplinary peer- reviewed Journal of the Washington Academy of Sciences four times a year — Spring, Summer, Fall, and Winter. 1 . Deadlines for quarterly submissions are: Spring - February 1 Fall - August 1 Summer - May 1 Winter - November 1 2. Draft manuscripts using a word processing program (such as MSWord), not PDF. 3. Papers should be 6,000 words or fewer. With 7 or more graphics, reduce the number of words by 500 for each graphic. 4. Include an abstract of 150-200 words. 5. Graphics must be in black & white or greytone. They must be referenced in the text. 6. Use endnotes, not footnotes. The bibliography may be in a style considered standard for the discipline or professional field represented by the paper. 7. Submit papers as email attachments to the editor. 8. Include the author’s name, affiliation, and contact information — including postal address. Membership in an Academy-affiliated society may also be noted. 9. Manuscripts are peer reviewed and become the property of the Washington Academy of Sciences. 10. There are no page charges. Please see the Academy’s web site, www.washacadsci.org, for the library subscription rate, listing of articles dating to 1899, and information on accessing them. Washington Academy of Sciences 61 Washington Academy of Sciences Affiliated Institutions National Institute for Standards & Technology (NIST) Meadowlark Botanical Gardens The John W. Kluge Center of the Library of Congress Potomac Overlook Regional Park Koshland Science Museum American Registry of Pathology Living Oceans Foundation National Rural Electric Cooperative Association (NRECA) Spring 20 1 5 62 Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies Acoustical Society of America, Washington Chapter American Association of Physics Teachers, Chesapeake Section American Astronomical Society American Ceramics Society American Fisheries Society American Institute of Aeronautics and Astronautics American Meteorological Society, Washington, DC Chapter American Nuclear Society, Washington DC Section American Phytopathological Society, Potomac Division American Society for Cybernetics American Society for Metals, Washington Chapter American Society of Civil Engineers, National Capital Section American Society of Mechanical Engineers, Washington Section American Society of Microbiology, Washington Branch American Society of Plant Biologists, Mid-Atlantic Anthropological Society of Washington ASM International Association for Women in Science, DC Metropolitan Chapter Association for Computing Machinery, DC Area Chapter Association for Science, Technology, and Innovation Association of Information Technology Professionals Biological Society of Washington Botanical Society of Washington Capital Area Food Protection Association Chemical Society of Washington District of Columbia Institute of Chemists District of Columbia Psychological Association Eastern Sociological Society Electrochemical Society, National Capital Section Entomological Society of Washington Geological Society of Washington Historical Society of Washington DC Human Factors and Ergonomics Society, Potomac Chapter Institute of Electrical and Electronics Engineers, Northern Virginia Section Paul Arveson Frank R. Haig, S. J. Sethanne Howard Vacant Lee Benaka David W. Brandt Vacant Charles Martin Vacant Stuart Umpleby Vacant Vacant Daniel J. Vavrick Vacant Mark Holland Vacant Toni Marechaux Jodi Wesemann Alan Ford F. Douglas Witherspoon Chuck Lowe Stephen Gardiner Chris Puttock Keith Lempel Elise Ann Brown Vacant Tony Jimenez Ronald W. Mandersheid Vacant Vacant Jeffrey B. Plescia Jurate Landwehr Vacant Gerald P. Krueger Murty Polavarapu (continued on next page) Washington Academy of Sciences Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies (continued from previous page) Institute of Electrical and Electronics Engineers, Washington Section Institute of Food Technologies, Washington DC Section Institute of Industrial Engineers, National Capital Chapter International Association for Dental Research, American Section International Society for the Systems Sciences International Society of Automation, Baltimore Washington Section Instrument Society of America Marine Technology Society Maryland Native Plant Society Mathematical Association of America, Maryland-District of Columbia-Virginia Section Medical Society of the District of Columbia National Capital Area Skeptics National Capital Astronomers National Geographic Society Optical Society of America, National Capital Section Pest Science Society of America Philosophical Society of Washington Society for Experimental Biology and Medicine Society of American Foresters, National Capital Society Society of American Military Engineers, Washington DC Post Society of Manufacturing Engineers, Washington DC Chapter Society of Mining, Metallurgy, and Exploration, Inc., Washington DC Section Soil and Water Conservation Society, National Capital Chapter Technology Transfer Society, Washington Area Chapter Virginia Native Plant Society, Potowmack Chapter Washington DC Chapter of the Institute for Operations Research and the Management Sciences (WINFORMS) Washington Evolutionary Systems Society Washington History of Science Club Washington Paint Technology Group Washington Society of Engineers Washington Society for the History of Medicine Washington Statistical Society World Future Society, National Capital Region Chapter Richard Hill Vacant Neal F. Schmeidler J. Terrell Hoffeld Vacant Vacant Hank Hegner Jake Sobin Vacant D. S. Joseph Vacant Vacant Jay H. Miller Vacant James Cole Vacant Eugenie Mielczarek Vacant Daina Apple Vacant Vacant E. Lee Bray Terrell Erickson Richard Leshuk Vacant Russell Wooten Vacant Albert G. Gluckman Vacant Alvin Reiner Alain Touwaide Michael P. Cohen Jim Honig Washington Academy of Sciences Room 113 1200 New York Ave. NW Washington, DC 20005 NONPROFIT ORG US POSTAGE PAID MERRIFIELD VA 22081 PERMIT# 888 Return Postage Guaranteed w AS 03$^ Volume 101 Number 2 Summer 2015 Journal of the WASHINGTON ACADEMY OF SCIENCES MCZ LIBRARY NOV 30 206 HARVARD UNIVERSITY Editor’s Comments S. Howard ii Board of Discipline Editors iii The Curious Case of Schmidt’s Star T Lipscombe 1 Docosahexaenoic Acid Induces Death in Murine Leukemia Cells by Activating the Extrinsic Pathway of Apoptosis E E Williams 1 3 A Nineteenth Century Historical Analysis of Game Warden Efforts: Focus on Rabbits and Hares K. GUcrease 39 Uranus and Neptune Revisited S. Howard 57 Membership Application 77 Instruction to Authors 78 Affiliated Institutions 79 Affiliated Societies and Delegates 80 ISSN 0043-0439 Issued Quarterly at Washington DC Washington Academy of Sciences Founded in 1898 Board of Managers Elected Officers President Mina Izadjoo President Elect Mike Coble Treasurer Ronald Hietala Secretary John Kaufhold Vice President, Administration Nick Tran Vice President, Membership Sue Cross Vice President, Junior Academy Vice President, Affiliated Societies Gene Williams Members at Large Paul Arveson Michael Cohen Frank Haig, S.J. Neal Schmeidler Mary Snieckus Past President Terrell Erickson Affiliated Society Delegates Shown on back cover Editor of the Journal Sethanne Howard Journal of the Washington Academy of Sciences (ISSN 0043-0439) The Journal of the Washington Academy of Sciences The Journal is the official organ of the Academy. It publishes articles on science policy, the history of science, critical reviews, original science research, proceedings of scholarly meetings of its Affiliated Societies, and other items of interest to its members. It is published quarterly. The last issue of the year contains a directory of the current membership of the Academy. Subscription Rates Members, fellows, and life members in good standing receive the Journal free of charge. Subscriptions are available on a calendar year basis, payable in advance. Payment must be made in US currency at the following rates. US and Canada $30.00 Other Countries $35.00 Single Copies (when available) $15.00 Claims for Missing Issues Claims must be received within 65 days of mailing. Claims will not be allowed if non- delivery was the result of failure to notify the Academy of a change of address. Notification of Change of Address Address changes should be sent promptly to the Academy Office. Notification should contain both old and new addresses and zip codes. POSTMASTER: Send address changes to WAS, Rm 113, 1200 New York Ave. NW Washington, DC 20005 Published by the Washington Academy of Sciences email: iournal@washacadsci.org website: www.washacadsci.org Academy Office Washington Academy of Sciences Room 113 1 200 New York Ave. NW Washington, DC 20005 Phone:(202) 326-8925 Journal of the 1200 New York Ave Suite 113 Washington DC 20005 www. wa sh acad scl. org WASHINGTON ACADEMY OF SCIENCES Volume 101 Number 2 Summer 2015 Contents Editor’s Comments S. Howard ii Board of Discipline Editors iii The Curious Case of Schmidt’s Star T. Lipscombe 1 Docosahexaenoic Acid Induces Death in Murine Leukemia Cells by Activating the Extrinsic Pathway of Apoptosis E. E. Williams 13 A Nineteenth Century Historical Analysis of Game Warden Efforts: Focus on Rabbits and Hares K. Gilcrease 39 Uranus and Neptune Revisited S. Howard 57 Membership Application 77 Instructions to Authors 78 Affiliated Institutions 79 Affiliated Societies and Delegates 80 ISSN 0043-0439 Issued Quarterly at Washington DC Y1CZ library NOV 3 0 2015 harvard UNIVERSITY II Editor’s Comments In the summer of 2015 we celebrate the diversity of topics presented in our Journal. The papers in this issue range from stars to rabbits. Before I briefly describe them, let me say it is an honor to serve as the new editor of the Journal of the Washington Academy of Sciences. Our Journal has a long and distinguished history. It is almost unique in its breadth. Each of you can help continue this history as we go forward. Please celebrate with me and continue to submit manuscripts on all sorts of topics: on the many sciences, on technical subjects, engineering, and mathematics. Expand the field to include the history, sociology, and psychology of these subjects. I welcome letters to the editor and book reviews. This is an exciting and challenging task. Join with me as we go forward. First up we have a paper by Trevor Lipscombe that reviews the curious case of Schmidt’s star (first mentioned in 1 891 and then relegated to the history texts). Trevor resurrects the star to glean possible new information. To follow it we have a paper by Gene Williams who discusses fatty acids and cancer. He explains one of the things fish oil is likely doing for you should you have any cancer cells wandering around that have not been killed off by the immune system. Third we have Kelsey Gilcrease writing about the efforts of 19th century game wardens (chasing those rabbits) in New Jersey and Massachusetts. To complete this issue I include a history paper on how the rotational periods (the lengths of their day) of Uranus and Neptune were determined before the space mission Voyager traveled by them. In the 2007 Spring issue, Vol 93, we published an article by Y. Said (then at George Mason University): “On the Eras in the History of Statistics and Data Analysis”. We have since retracted this article because of suggested controversy over its uniqueness. Sethanne Howard Editor Washington Academy of Sciences Ill Journal of the Washington Academy of Sciences Editor Sethanne Howard sethanneh@msn.com Board of Discipline Editors The Journal of the Washington Academy of Sciences has a 12-member Board of Discipline Editors representing many scientific and technical fields. The members of the Board of Discipline Editors are affiliated with a variety of scientific institutions in the Washington area and beyond — government agencies such as the National Institute of Standards and Technology (NIST); universities such as Georgetown; and professional associations such as the Institute of Electrical and Electronics Engineers (IEEE). Anthropology Emanuela Appetiti Astronomy Sethanne Howard Biology/Biophysics Eugenie Mielczarek Botany Mark Holland Chemistry Deana Jaber eappetiti@hotmail.com sethanneh@msn.com mielczar@physics.gmu.edu maholland@salisbury.edu diaber@mai~vmount.edu Environmental Natural Sciences Terrell Erickson Health Robin Stombler History of Medicine Alain Touwaide Operations Research Michael Katehakis Physics Katharine Gebbie Science Education Jim Egenrieder Systems Science Elizabeth Corona terrell.ericksonl@wdc.nsda.gov rstombler@aubumstrat.com atouwaide@hotmail.com mnk@rci.rutgers.edu katharine.gebbie@nist.gov i im@deepwater.org elizabethcorona@gmail.com Summer 2015 Washington Academy of Sciences 1 The Curious Case of Schmidt’s Star Trevor Lipscombe Catholic University of America Press, Washington, DC. Abstract This article discusses the internal structure of a type of star first proposed by August Schmidt in 1891, one that causes any light to enter it to move in a circle. An exact analytical solution of the equation of hydrostatic equilibrium is thus obtained. The solution is physically realistic, in the sense that the central density, central pressure, and total mass are all finite, while both density and pressure drop to zero at the outer radius of the star. In the core of the star, the pressure depends only weakly on density. The outer layers of the star can be well- approximated as isothermal. Schmidt’s star, then, is a physical system of historical, pedagogical, and mathematical interest. Introduction In the late 1800s astrophysicists faced a conundrum. The age of the Earth had been reliably determined and thus the minimum age of the Sun was also known; but given that age, and the known laws of physics, stars should have burnt out long before.1 We now know that the stars shine because of nuclear processes that take place in their core, processes unknown in the nineteenth century. In 1891 a German scientist, August Schmidt (1840-1929), proposed a radical solution to resolve the paradox. What if stars didn't shine but were, in a sense, mirages? That is to say, suppose stars acted as giant lenses, with a refractive index that varies as a function of radius and that causes any ray of light to enter it to move with a circular motion, thereby never leaving the star? This would remove any need for a mechanism by which stars had to bum fuel to generate energy, and so resolve the conundrum. Schmidt proposed a model in which the Sun’s outer surface was such an optical illusion2. This idea caught the attention of “a student of astronomy” E.J. Wilczynski, who wrote the first English-language article on Schmidt’s theory in 1895. It appeared in the first-ever volume of the Astrophysical Journal3 (ApJ), followed only a few pages later by a note Summer 2015 2 from James Edward Keeler, the co-founder and co-editor of the ApJ, who commented: “...The theory is apt to be more favorably regarded by mathematicians than by observers' ,4 a sentiment echoed by George Ellery Hale— the other co-founder and co- editor of the ApJ — who, in the very next volume, wrote : “As a theoretical discussion the theory is interesting and valuable, but few observers of the Sun wall consider it capable of accounting for the varying phenomena encountered in their investigations"5 . Here, though, we follow Michael Faraday’s dictum that “Nothing is too wonderful to be true, if it be consistent with the laws of nature" and investigate Schmidt’s theory, to see whether such an astrophysical object could actually exist. Heretofore, studies have only paid attention to the optical properties of Schmidt’s star. (The exact nature of the outer visible layer of the Sun, which was in large part what Schmidt dealt with, still generates controversy.6) In this article, we determine the physical properties of Schmidt’s star, which appears not to have been done before. The basic mechanism proposed by Schmidt leads to an exact solution for the equation of hydrostatic equilibrium, which governs self-gravitating stationary spheres of fluids. Density Determination In Waves and Grains , Mark Silverman analyzes the optical properties of Schmidt’s star7. If a spherical lens has a refractive index n(r) and possesses spherical symmetry, then Silverman shows that a point on a light ray r is given by: dr He (1) where R is the radius of the star. For the light rays to move in a circular path, we require that r = constant, and the refractive index in the medium must therefore vary as: n{r ) = —■ (2) r Washington Academy of Sciences 3 Note that when r — R we have n - 1, which is the refractive index of the vacuum. The refractive index in a medium depends, among other things, on the density of that medium, which is the basic mechanism behind the formation of mirages. The Clausius-Mosotti (or Lorentz-Lorenz) relation can be used to relate the refractive index of a substance to its density8: rr - 1 n2+ 2 Kp (3) where K is a constant that depends on the particular gas of which the star consists. Studies of the Lorentz-Lorenz relation for gaseous and liquid hydrogen show that K remains approximately constant (-1.03 cm”3 / g) for a broad range of temperatures (15-298 Kelvin) and pressures (9-200 atm)9, though it may well not hold at the high densities typically found at the centers of stars. Substituting in from Eq. (3) above: R2-r 2 R2+2r2 - Kp. (4) When r = R, the density falls to zero, as it should at the outer radius of the star. Note also that when r = 0, Kp{ 0) = 1, so that K is the inverse of the central density pc. Thus, using the scaled dimensionless radius x = r / R, we can write: P = Pc ^1-x2^ 1 +2x" (5) As a consequence of knowing how the density varies within this astrophysical object, we can calculate its mass, M: i M = 47rpcR2\ x 2 (l-x2 ) 1 + 2x" +- 9V2 {Vlx x2 -51n(2x2 +l) + 61nx + -4x4+6x2-6 Eq. (16) simplifies to: tan 6x + const. (16) 1 7 2 9 4 27 7x — x 2 2 2 + 9V2 ln(2x2 +l) f n x tan 1 V2x V x) (17) +const. At the outer radius of the star, the pressure drops to zero and thus P(x = 1 ) = 0. This boundary condition allows for the calculation of the integration constant, since we must have: P{x = 1) = 0 = KGRjpl 1 9 27 7-- In 3 2 2 2 + const. (18) Hence: 7cGR2 p2 const = — 1 9 27, 0 — + 7 H 1 In 3 2 2 2 (19) The complete solution for the pressure distribution within Schmidt's star is thus: . 7rGR 2 p2. P(x) = ^23 + 27 In 3^ „ , 9 4 - lx" — x 2 J 27 f 1 A In ^2x2 +l) + 9v2 x — tan”1 V2x v X ) (20) Note that: f lim x— >0 tan 1 >[2x x (21) so that the central pressure is indeed finite and has the value: Summer 2015 6 p _ ”GR2Pc rc ~ ^ 27 In 3 — 1 3 ^ v 2 , (22) Hence: Pc m=- ^23 + 27 ln3N 2 , |x4-yta(2xJ+l)+9\/2 7 271n3— 13V l 2 J (23) A graph of the pressure and density within the star, as a function of stellar radius, is shown in Figure 1. Fig.l Density and Pressure as a Function of Radius Density Pressure from Eq. (20) The pressure and density of the star are such that a good approximation for their relation is1 1 : P = \.0\\\2PC 1 - exp ^-4,1 A) V Pc y (24) The exact relation of pressure with density [from Eqs. (5) and (20)] is compared with the approximate relationship of Eq. (24) in Figure 2. Washington Academy of Sciences 7 Hence, by starting with a simple requirement — light rays travel in circles within the astrophysical object — we can determine that such a star has a finite central density and pressure and determine the variation of the pressure and density as a function of the radius. Comparison with Poiytropic Models The usual approach to stellar astrophysics is to explore the polytrope equation12. That is to say, one seeks solutions of the form: P = KpMln (25) in the equation for hydrostatic equilibrium. This model generates the Lane-Emden equation, named after Jonathan Homer-Tane, an astrophysicist who spent many years in Washington DC, and Swiss astrophysicist Robert Emden. Here n is not the refractive index, but the so-called polytropic index of the star. It is complicated to compare Schmidt’s star with standard polytrope solutions. For example, as seen from Eq. (10), the central density is related to the average density by: pc = 4.348/? (26) Summer 2015 8 which, by use of the Polytrope Tool13, is equivalent to the central pressure of a polytrope whose index is n — 1.31. The standard model for the Sun (the Eddington model, for which n = 3) has the value pc =54.18 p, just over a factor of ten larger14. The central pressure in Schmidt’s star is given by numerically evaluating Eq. (22): P.. =8.331 26 P' . (27) 6 The mass, though, is given in Eq. (8), as: M ~ 4xpcRi (0.0767) (28) and so by substituting in for the central density, we obtain: 8.33126;rGfl2 ^ w v P, c M 4/rR3 (0.0767) (29) or: R. * 4.7 c GM 2 R4 (30) Again, by means of the Polytrope Tool, this is an expression equivalent to the central pressure of a polytrope of index n = 2.595, almost double the index obtained from consideration of the central density. The Eddington model for the Sun has the numerical factor 1 1 .05 rather than 4.7, a central pressure some 2.35 times higher than the Schmidt star.1'' For further comparison, note that for a polytrope: P_ T (jl| V Pc J 1+1/ n Consequently: P_p£ = fP_\n Pc P 'Pc' (31) (32) Washington Academy of Sciences 9 We can thereby define an effective pointwise polytropic index n by: n = (33) A curve of n as a function of radius is shown in Figure 3. Fig. 3 Effective polytropic index as function of radius Index, n The best fits have n = —1.22 from x = 0 to 0.5, half way out through the star. This is a negative polytrope of varying index. Approaching x = 1, the index becomes large and negative, so that P ~ const p , which is the equation describing an isothermal outer layer (infinite polytropic index) to Schmidt’s star. Discussion Negative-index polytropes were first discussed by Eddington in 1931 16. In the same paper, he used the phrase “Incomplete poly tropes” to describe a structure similar to Schmidt’s star, wherein the inner core might best be modeled by one value of the polytropic index and the outer layers by another. Viala and Horedt17 showed that astrophysical objects with negative polytropic indexes are good models for, among other things, interstellar clouds. In addition, they showed that sufficiently negative Summer 2015 10 indexes ( n < 1), can be stable. Chaplygin gases, which have negative indexes, are currently of great interest in cosmology, as they are candidates for dark matter and can form stable gravitational structures (both in classical Newtonian and general relativistic gravity)18. In Schmidt's star, both the pressure and the density decrease when moving radially outwards. In the outer layers, the pressure varies almost linearly with density, which suggest an isothermal envelope for the star. However, in the stellar core it is similar to a negative-index polytrope of index n < H, in that the temperature must increase radially outwards. This creates a significant problem for Schmidt's star, in that to be physically realistic, the model must represent an astrophysical object whose core is being heated externally, either by particles or by radiation, in a spherically symmetric manner, but whose outer layers are isothermal. Keeler and Hale's original criticisms of Schmidt’s proposal were that it was of importance only mathematically. Regrettably that may indeed be the case. Schmidt’s star, though, remains of interest. Such interest is not just historic; Schmidt’s star also is of pedagogical value19. Undergraduate physics students, as an exercise in physical modeling, could be presented with the Schmidt-Silverman equation for the refractive index of Schmidt’s star and then asked to solve for the pressure and density of this object. This requires knowledge of various disciplines within physics. They could also be asked to comment on whether such an object could indeed exist, which requires them to recognize that the temperature in an astrophysical object should, to be realistic, fall off with increasing radius. Conclusions In this paper, we have explored the structure of an astrophysical object whose physics has not previously been determined completely. By requiring such a star to act as a graded refractive index lens that causes all light entering in to it to move in a circular path, we have been able to determine the density of the star and its pressure. Such a star has a pointwise negative polytropic index, but its pressure and density both decrease as the radius increases, and the value of the index is such that Schmidt’s star is likely to be stable. While the study was motivated by Schmidt’s suggestion in 1891, the density profile and pressure profile here represent an exact solution of the equation of hydrostatic equilibrium. Washington Academy of Sciences II whether one gives credence to Schmidt’s belief or not. This is one of the few physically motivated stellar models, other than polytropes, that is not singular at the origin nor infinite in extent. While the temperature profile makes Schmidt's star likely to be physically unrealistic, it remains of historical, mathematical, and pedagogical value. This paper is dedicated to Kelsey Schmidt and Tom LaCour on the occasion of their marriage. REFERENCES 1 Frank D. Stacey ‘Kelvin’s Age of the Earth paradox revisited,” Journal of Geophysical Research 105(B6), pp 13155-13158 (2000). 2 August Schmidt “Die Strahlenbrechung auf der Sonne: ein geometrisches Beitrag zur Sonnenphysik” (Stuttgart: Metzlerscher Verlag, 1891). 3 Ernest J Wilczynski “Schmidt’s Theory of the Sun”, Astrophysical Journal vol. 1, pp 112-126 (1895). 4 James E Keeler “Schmidt’s Theory of the Sun”, Astrophysical Journal vol. 1, pp 178- 179 (1895). 5 George E Hale “Notes on Schmidt’s Theory of the Sun”, Astrophysical Journal vol. 2, pp 69-74 (1895). 6 See, for example, Pierre-Marie Robitaille “Commentary on the Radius of the Sun: Optical Illusion or Manifestation of a Real Surface”, Progress in Physics Vol. 2, L5-L6 (2013). 7 Mark P. Silverman Waves and Grains (Princeton N.J.: Princeton University Press, 1998), pp 27-30. 8 See, for example, Charles Kittel Introduction to Solid State Physics (8,h edition) (New York: John Wiley & Sons, 1990). 9 Dwain E. Diller “Refractive Index of Gaseous and Liquid Hydrogen” J. Chem Phys. 49(7) 3096-3105 (1968). 10 See, for example, S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago: University of Chicago Press, 1939), page 63, equation (6). 11 By inspection, a trial solution is P — const[ 1 — exp(— ap)]. The requirement that P = 1 when p ~ 1 determines the constant in terms of a. At low densities, we have P = [a/[ 1 — exp(— a)]p and so use of the data at low densities in a linear regression estimator (such as LINEST in Microsoft Excel) leads to the best numerical fit. 12 S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago: University of Chicago Press, 1 939), pp. 84-182. Summer 2015 12 L' http://www.webnucleo.Org/home/online_tools/polytrope/0.8/ 14 S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago: University of Chicago Press, 1939), equation 56, chapter 6, page 230. 15 S. Chandrasekhar An Introduction to the Study of Stellar Structure (Chicago: University of Chicago Press, 1939), chapter 6, page 230, equation 57. 16 Arthur S. Eddington “A Theorem Concerning Incomplete Polytropes” Mon. Not. Roy. Ast. Soc. 91 pp 440-444 (1931). 17 Yves P. Viala and Georg P. Horedt “Polytropic Sheets, Cylinders, and Spheres with Negative Index”. Astron. & Astrophys. 33 pp 195-202 (1974). 18 *i Trevor C. Lipscombe “Self-gravitating clouds of generalized Chaplygin and anti- Chaplygin gases,” Physica Scripta 83(3) ID = 035901 (201 1). 19 Another example of historically motivated physics of pedagogical value might be the modeling of Newton’ s-bucket experiment in Carl E. Mungan and Trevor C. Lipscombe “Newton’s Rotating Water Bucket: A Simple Model,” Journal of the Washington Academy of Sciences 99(2), pp 15-24 (2013). Bio Trevor Lipscombe is the director of the Catholic University of America Press. He holds a doctorate in theoretical physics from Oxford, is a Fellow of the Royal Astronomical Society, and tries to do theoretical physics in his spare time. He is the author of “The Physics of Rugby’' (Nottingham University Press, 2009); coauthor, with Alice Calaprice, of “Albert Einstein: A Biography” (Greenwood, 2005); and editor of a critical edition of Blessed John Henry Newman's novel “Loss and Gain: The Story of a Convert” (Ignatius Press, 2012). Washington Academy of Sciences 13 Docosahexaenoic Acid Induces Death in Murine Leukemia Cells by Activating the Extrinsic Pathway of Apoptosis. E. Eugene Williams Salisbury University Abstract Docosahexaenoic acid (DHA) is a unique fatty acid that is found predominantly in the phospholipids of cell membranes. It has wide- ranging therapeutic effects that are broadly appreciated but poorly understood. Its principal location in the membranes of cells suggests that these myriad effects are manifest there. When cultured in DHA-enriched medium, cells of the murine leukemia cell line T27A took up the fatty acid and incorporated it into cellular phospholipids, particularly those of the plasma membrane. Culture in DHA-enriched media also caused significant dose-dependent cell death accompanied by increased plasma membrane bleb formation. Cysteine-dependent aspartate-directed proteases (caspases)-3 -8 and -9 were also activated, establishing apoptosis as the mechanism of DHA-induced cell death. Inhibition of any one of these caspases rescued the cells from apoptotic death. Caspase inhibition experiments identified T27A cells as belonging to the type II group of apoptotic cells and showed that apoptosis was initiated via the extrinsic pathway. Together these and previous data support the hypothesis that DHA causes cell death in leukemic cells by inducing alterations in the structure of lipid rafts that lead to the ligand-independent activation of death receptors and apoptosis. Introduction Docosahexaenoic acid (DHA, 22:6n-3) is a unique fatty acid that is found in the cells of a wide range of organisms from bacteria to humans. It is the longest and most unsaturated of the commonly occurring n-3 (omega-3, co-3) fatty acids (Salem et al. 1986). DHA has diverse therapeutic properties that are acclaimed in both the scientific and lay communities (Stillwell and Wassail 2003; Siddiqui et al. 2004; Chapkin et al. 2009). A remarkable number of conditions and diseases have been demonstrated to be prevented, mitigated, counteracted or improved by DHA. These include maladies as disparate as cancer, heart disease, cystic fibrosis, diabetes, immune function and even psychiatric disorders (Stillwell and Wassail 2003; Siddiqui et al. 2004; Calder 2012; Mischoulon and Freeman 2013). While the relationship between DHA Summer 2015 14 and improved health is widely appreciated, the basic molecular mechanism underlying this relationship remains unclear. As noted by Stillwell (2008), the assortment of seemingly unrelated biochemical and physiological processes underlying the diseases and conditions that are influenced by DHA suggests that this fatty acid influences a fundamental cellular function or property. DHA has been shown to have powerful anti-cancer effects in animals and cultured tumor cells (Siddiqui et al. 2004). For example, it is effective at reducing the accumulation of leukemic cells in vitro and in slowing the rate of progression of leukemia in animals ( e.g . Jenski et al. 1993; Jenski et al. 1995; Zerouga et al. 1996). DHA has been shown to induce cell death in human and mouse leukemia cells in a dose dependent manner (Kafrawy et al. 1998; Yamagami et al. 2009) and it has been suggested the anti-leukemia properties of DHA are in general founded on the ability of DHA to induce cell death in tumor cells (Serini et al. 2009). Despite continuing efforts, it is currently unclear precisely how DHA triggers cell death. DHA can be converted into reactive oxygen species that can influence cell survival (Siddiqui et al. 2008), and into powerful anti-inflammatory and pro-resolving mediators (resolvins, protectins and maresins) that can influence cell survival and disease etiology (Serhan et al. 2014; Colas et al. 2014; Dalli et al. 2015). DHA can also affect gene expression (Berger et al. 2006), the acylation patterns of membrane proteins (Webb et al. 2000), and the function of enzymes and ion channels (Matta et al. 2007). However a large and growing body of evidence indicates that DHA induces cell death only after it has become incorporated into membrane phospholipids and that the initial triggering event in cell death is a membrane-based phenomenon (Stillwell and Wassail 2003; Stillwell et al. 2005; Calder 2012). There is substantial physiological, biochemical, biophysical, and morphological evidence that DHA-containing phospholipids change the structure of cell membranes (Mitchell et al. 2003; Niu and Mitchell 2005; Chapkin et al. 2008; Shaikh 2010; Rockett et al. 2012; Teague et al. 2013; Pinot et al. 2014). Indeed, whether provided as a dietary component to an individual organism (Lien 2009) or as a component of the incubation medium of cultured cells (Zerouga et al. 1996; Williams et al. 1998; Williams et al. 1999), DHA is taken up by cells and incorporated into the Washington Academy of Sciences 15 phospholipids of membranes. The plasma membrane in particular appears to be a primary location of action for the tumor cell killing properties of DHA (Jenski et al. 1993; Pascale et al. 1993; Williams et al. 1998; Williams et al. 1999). Of particular interest in this regard is the influence of DHA-containing phospholipids on the membrane microdomain structures known as lipid rafts. Lipid rafts serve as platforms for the regulation of cell processes and represent a selective cellular compartment that can co-localize and modulate the activities of enzymes, receptors and other proteins (Simons and Ikonen 1997; Lingwood and Simons 2010). There is evidence that DHA-containing phospholipids induce cell death by altering the structure or organization of lipid rafts, and that this influence on membrane structure is the first and most important step in DHA-induced cell death (Stillwell et al. 2005; Schley et al. 2007; Chapkin et al. 2008). Other evidence strongly suggests that DHA causes cell death in tumor cells by the induction of apoptosis (Blanckaert et al. 2010; Kang et al. 2010). There are two distinct activation pathways for apoptosis. The extrinsic pathway involves plasma membrane-associated death receptors and a cysteine-dependent aspartate-directed protease, caspase-8. The intrinsic pathway involves the release of cytochrome c from mitochondria and the activation of caspase-9. These two initiating events then cause the activation of downstream effector caspases including caspase-3 which in turn cleaves a series of intercellular substrates to continue the apoptotic cascade. Lipid rafts are importantly involved in the extrinsic apoptotic pathway as the death receptors, a subset of the tumor necrosis factor receptor superfamily, are among those receptors regulated by lipid rafts (Gajate et al. 2009; Lang et al. 2012). Thus, there is evidence that DHA causes the death of many types of tumor cells, that the cause of cell death in many of these instances is the induction of apoptosis, that DHA alters the structure of lipid rafts, and that lipid rafts regulate the receptors involved in initiating the extrinsic pathway of apoptosis. This study attempts to connect these links by testing the hypothesis that DHA causes cell death in leukemia cells by specifically triggering the extrinsic pathway of apoptosis. We show that DHA is selectively incorporated into the plasma membrane of murine leukemia (T27A) cells. We use both morphological and biochemical Summer 2015 16 means to demonstrate that DHA induces apoptosis in these cells. By monitoring and manipulating the activities of caspases -8, -9 and -3, we further show that all three caspases are activated by DHA and that the inhibition of any one of them rescues T27A cells from DHA-induced apoptosis. Together these and previous data support the hypothesis that DHA causes cell death by inducing alterations in the structure of lipid rafts that lead to the ligand-independent activation of death receptors and apoptosis. Methods Materials T27A murine leukemia cells were obtained from American Type Culture Collection (Manassas, Va). Fatty acids and fatty acid methyl ester (FAME) reference standards were purchased from Nu-Chek-Prep (Elysian, MN). RPMI-1640 culture medium supplemented with 2 mM glutamine, 25 mM HEPES, 50 pg/mL streptomycin and 100 units/mL penicillin, was from Cambrex Bio Science (Walkersville, MD). Bovine calf serum was from Hyclone (Logan, UT). Irreversible, cell-permeable inhibitors of caspases -3 (Z-D[0-Me]E[0-Me]VD[0-Me]-FMK), -8 (Z- IE[0-Me]TD[0-Me]-FMK), and -9 (Z-LE[0-Me]HD[0-Me]-FMK) were from Calbiochem (EMD Biosciences, Inc., La Jolla, CA). The colorimetric assay kits for measuring the activities of caspases -3, -8 and -9 were from BioVision (Mountain View, CA.). Staurosporin, SiCE (“Celite”), and dimethyl sulfoxide (DMSO) were from Sigma Chemical Co. (St. Louis, MO). All other chemicals were from Sigma or Thermo Fisher Scientific (Waltham, MA). Cell culture Except where noted, T27A cells were cultured in RPMI-1640 medium supplemented as described above and with 10% (vol/vol) bovine calf serum in 25 cm2 culture flasks maintained at 37°C under an atmosphere of 5% CO2 in humidified air. As noted previously (Zerouga et al. 1996; Williams et al. 1998; Williams et al. 1999), under these conditions cultures doubled every 12 to 15 hours. Cell viability was monitored by trypan blue exclusion (0.04% in phosphate buffered saline [PBS, 0.154 M NaCl, 0.016 M NaH2P04, pH 7.2]). Washington Academy of Sciences 17 Supplementation of culture media with fatty acids DHA and oleic acid (OA, 18:1 n-9) were added to RPMI culture medium using the methods of Spector and Hoak (1969) exactly as described by Williams et al. (1998). The fatty acid was dissolved in hexane and transferred to an Erlenmeyer flask containing SiCU Ten g of SiCb were used per mmol of fatty acid. The hexane was removed completely by a gentle stream of N2 before the dry mixture was transferred to a solution of fatty acid free bovine serum albumin (1% fatty acid free BSA in RPMI supplemented as above, but excluding serum). After stirring for 30 min in the dark, the RPMI/fatty acid mixture was centrifuged for 30 min at 600 grav to remove the Si02 and the medium was sterilized by filtration (0.22 pm). Bovine calf serum was added to 10% (vol/vol) of the total just before use. Calf serum contributes a small amount of fatty acids to the final culture medium, but less than 1 % of that is DHA (Williams et al. 1998). Unless noted otherwise, cells were incubated in fatty acid-enriched medium for 3 days (68-76 hours). Under these conditions T27A cells take up considerable DHA, and at DHA concentrations below 0.61 mM they remain >90% viable (Williams et al. 1998; Williams et al. 1999; and see below). Assay of caspase activity and caspase inhibition The activities of caspases -3, -8, and -9 were measured spectrophotometrically in 90-well plates. For each assay, T27A cells were cultured in RPMI medium containing no additions, 1.3 pM staurosporin, or 0.61 mM DHA. After 16 h of culture, cells from each flask were harvested by low-speed centrifugation. Cell viability (always greater than 90% in control cells) was assessed by trypan blue exclusion and cell density was determined by duplicate counts on a hemacytometer. For each treatment, 3 x 106 cells were treated with 50 pF of lysis buffer according to the manufacturer’s instructions. After centrifugation, 30 pF of cell lysate were mixed with 20 pF of caspase assay medium in a well of the plate, mixed, and allowed to incubate at 37°C for 1 hour before the absorbance was read at 405 nm. Background values were subtracted from all absorbances and all treatment values were expressed as percentage of the control. For caspase inhibition experiments, cells were exposed to 10 pM inhibitor in DMSO (0.1% final concentration) for 30 min before exposure to control or fatty acid-enriched medium. Summer 2015 18 The linearity of the caspase assays was confirmed using p- nitroanaline as a standard. Regression analyses of the resulting standard curves yielded lines with n >0.990. Staurosporin was used as a positive control and only those assays that showed staurosporin-induced caspase activity were analyzed further. Isolation of plasma membranes After 48 h of culture in either normal (control) medium or medium enriched with 0.3 mM DHA, cell cultures were disrupted by sonication and the resulting homogenate was fractionated by the centrifugation protocol of Kaduce et al. (1977) using the buffers of Molnar et al. (1969) as described by Williams et al. (1998; 1999). Briefly, T27A cells were collected by centrifugation (500 grav for 15 min), resuspended in 0.25 M sucrose buffer (0.25 M sucrose, 40 nrM NaCl, 100 mM KC1, 5 mM MgSO-t. 7 HrO, 20 mM Trizma base, pH 7.2 with HC1), and disrupted (on ice) by sonication for 2 x 35 sec using a tip-type sonicator (Fisher Scientific Model 500, 35 seconds, pulse on 1 sec, pulse off 1.5 sec). The cell homogenate was centrifuged at 27 kgrav for 10 min to remove undisrupted cells and cellular debris and the supernatant over the resulting pellet was spun for 1 hour at 105 kgrav to produce a mixed membrane pellet. The mixed membrane pellet was layered onto a pad of 1.1 M sucrose (remaining composition as above) and spun at 107 kgrav for 16 hours. The white interfacial material was collected and washed twice in excess PBS. The resulting membrane represents a better than 8-fold purification of plasma membrane over the mixed membrane fraction (Kaduce et al. 1977) and has been used in previous studies to determine the effects of DHA on membrane structure and composition in T27A cells (Williams et al. 1998; Williams et al. 1999). Lipid extraction and gas chromatography of membrane fatty acids Total lipids where extracted from whole cell preparations and from isolated plasma membranes using CHCI3/CH3OH (Bligh and Dyer 1959) and concentrated under a stream of dry N2 gas. Phospholipids separated from neutral lipids ( e.g ., triacylglycerols) by silicic acid chromatography (Wren 1960; Williams and Somero 1996) were transesterified into FAMEs using methanolic sodium methoxide (Eder et al. 1992). FAMEs were resolved using a 0.25 mm x 30 nr HP-23 cis/trans Washington Academy of Sciences 19 FAME column in a Hewlett-Packard 6890 gas chromatograph. The instrument was programmed to produce a temperature ramp from 1 80°C to 240° at 2°C/min starting 2 minutes after sample injection. Peaks corresponding to individual FAMEs were identified by comparison ot retention times to those of authentic standards. Peak areas were calculated using Hewlett-Packard’s ChemStation software. Statistics Statistical analyses were carried out using version 2.15.3 of R (R Development Core Team, 2008; http://www.r-project.org/). Probabilities < 0.05 were considered significant (and labeled *). Percent data were arcsine transformed (sin'Wproportion) before statistical analyses as recommended (Sokal and Rohlf 1981). The normality of distribution of each data set was assessed using the Shapiro-Wilk test. The homogeneity of variances among data sets was tested using Fligner- Killeen test as it has been shown to be least sensitive to departures from normality (Conover et al. 1981). The slopes of regression lines were compared to each other and to slope = 0 using the linear model function of R. Group means were compared using one-way analysis of variance (ANOVA) followed by Tukey's HSD mean separation test, or where appropriate, the Kruskal-Wallis test followed by Wilcoxon rank sum tests. Results When cells of the murine leukemia line T27A were cultured in media supplemented with DHA, they took up the fatty acid and incorporated it into cellular phospholipids (Table 1). In phospholipids isolated from whole cells, DHA levels were 25 times that found in control cells. The increased proportion of DHA was associated with a large reduction in the proportions of stearic acid (18:0) and the n-6 isomer of 18:3. Proportions of palmitic acid (16:0) and oleic acid (18:1) increased. By contrast, DHA incorporation into phospholipids of the plasma membrane represented an 1 8-fold increase over that of control cells and resulted in a final proportion of DHA almost twice that observed in phospholipids isolated from whole cells. In the plasma membrane, DHA largely displaced oleic acid (1 8:1), as well as arachidonic acid (20:4) and other long chain polyunsaturated fatty acids. The DHA-induced alteration of membrane lipid composition of both whole cells and plasma membrane Summer 2015 20 is reflected in the near inversion of the n-6/n-3 ratios (Table 1) after treatment with DHA. The dramatic accumulation of DHA in phospholipids of the plasma membrane of T27A cells can also be clearly seen when comparing the ratios of palmitate to stearate (16:0/18:0) and of DHA to stearate (22:6/1 8:0) in phospholipids extracted from plasma membrane and whole cells cultured in control versus DHA-enriched media (Figure 1). The results shown in Table 1 and Figure 1 closely mirror previously reported observations on the effects of DHA on the lipid composition of plasma membranes isolated from these cells (Zerouga et al. 1996; Williams et al. 1998; Williams et al. 1999) and indicate that the experiments presented here both compliment and expand those earlier works. Table 1. The distribution of phospholipid fatty acids, as percent of total fatty acids, extracted from whole cells and from isolated plasma membranes after 48 h of culture in control or DHA-enriched (0.3 rnM) medium. Minor fatty acids, i.e. those comprising less than 1% of the total, are excluded from the analysis. The data represent the means of two independent experiments. Fatty Acid Whole Cells Control + DHA Plasma Membranes Control + DHA 14:0 1.1 1.4 1.2 1.1 14:1 1.1 0.4 0.2 0.1 16:0 12.2 17.5 15.0 16.9 16:1 0.8 0.8 1.3 0.8 18:0 73.5 59.1 49.4 50.6 18:1 3.6 6.4 20.3 7.7 18:2 n-6 1.2 1.9 4.9 2.9 18:3 n-6 4.6 1.2 0.0 0.0 20:3 n-6 0.0 0.0 1.6 0.4 20:4 n-3 0.4 0.0 3.6 1.0 22:5 n-3 0.0 0.0 1.3 0.0 22:6 n-3 0.4 10.3 1.1 18.5 total n-6 5.8 3.1 6.5 3.3 total n-3 0.8 10.3 6.0 19.5 n-6/n-3 7.25 0.30 1.08 0.17 Washington Academy of Sciences 21 Figure 1. The ratios of mean values of palmitate to stearate (16:0/18:0) and of DHA to stearate (22:6/18:0) in phospholipids extracted from whole cells or plasma membrane (PM) after the cells had been cultured for 48 h in control medium or in medium containing 0.3 mM DHA. Figure 2. The density of T27A cells three days after exposure to the indicated concentrations of fatty acid and expressed as a percentage of control. Squares, OA; circles, DHA. Linear modeling revealed that the slope of the regression of the OA response is not significantly different from zero. The slope of the regression of the DHA response is highly significantly different from both slope = 0 and the OA response (p < 0.001 in both cases). Each point represents the mean ± 1 standard error of the mean from n = 7-14 (DHA) or n = 3-6 (OA) independent determinations of different cultures. Summer 2015 22 Culture in DHA-enriched medium caused a significant reduction in the rate of leukemic cell proliferation. Figure 2 shows a DHA-dose dependent reduction in cell density compared to control cultures and to cultures similarly exposed to OA. The proportion of viable cells in the DHA-enriched cultures also fell significantly, while the viability of cells in cultures exposed to OA remained indistinguishable from that of the controls (Figure 3). Together these data show that DHA caused significant cell death over a three day exposure to concentrations of DHA in the culture medium from 0.3 to 0.9 mM. Phase contrast microscopy revealed that unlike control cells or cells cultured in OA-enriched medium, cells cultured in DHA-enriched medium were irregularly shaped and exhibited conspicuously higher internal complexity including extensive cytoplasmic vacuolization. In addition, the external surfaces of control cells and of cells cultured in OA- enriched medium were even and regular, whereas the surfaces of cells cultured in DHA-enriched medium were uneven and displayed numerous exvaginations of the plasma membrane (commonly referred to as “blebs”; e.g. Charras 2008). Figure 4 shows that the percentage of T27A cells exhibiting blebs increased steadily with DHA dose until at the highest doses tested these structures appeared on nearly 75% of all cells present in the culture. Culture of T27A cells for 16 h in a medium containing 0.61 mM DHA resulted in a significant elevation of the activities of caspases-3, -8, and -9 (Figure 5). When cell cultures were individually treated with 10 pM of an inhibitor specific for each of these caspases for 30 min prior to culture in DHA-enriched medium they did not undergo cell death and cell densities were similar to those of control cultures (Figure 6). Washington Academy of Sciences 23 Figure 3. The viability of T27A cells as assessed by trypan blue exclusion three days after exposure to the indicated concentrations of fatty acid. The slope of the regression of the OA response is not significantly different from zero and that of the regression of the DHA response is highly significantly different from both slope = 0 and the OA response (p < 0.001 in both cases). Squares, OA; circles, DHA. Each point represents the mean ± 1 standard error of the mean from n = 3 independent determinations of different cultures. 100 _c _o CO CD 75 4-» IE r- X UJ t/5 50 V o 25 r~ <5 u 5 C- 0 0.0 0.2 0.4 0.6 0.8 [Fatty Acid] (mM) Figure 4. The percentage of T27A cells exhibiting plasma membrane exvaginations (“blebs”, inset) after 1 6 h as a function of the concentration of fatty acid in the culture medium. Square, OA; circles, DHA. Each point represents the mean ± 1 standard error of the mean from n = 3 independent determinations of different cultures. Summer 2015 24 Figure 5. The effect of DHA on cellular caspases. The activities of caspases (casp-) 3, 8, and 9 in T27A cells after 16 h of culture in normal medium (control) and in medium containing 0.61 mM DHA. The activity of caspase-9 is significantly (*, p < 0.05) different from the control value. The activities of caspases -3 and -8 are not significantly different from caspase-9 and are marginally significantly (0.05 < p < 0.1) different from the controls. The data are presented as percent of activity found in control cells and represent the means ± 1 standard error of the mean from n = 3 (caspase-3) or n = 4 independent assays using separate cell cultures. inhibitor then Dl IA Figure 6. Density of T27A cell cultures expressed as a percentage of that in control flasks after 48 hours in the presence of medium containing no additions, 0.1% (vol/vol) DMSO (carrier control), and medium enriched with 0.61 mM DHA. These are compared to the densities of cell cultures exposed for 30 min to 10 (iM of an inhibitor specific to each one of the indicated caspases before the 48 h exposure to medium containing 0.61 mM DHA. Each bar represent the mean ± 1 standard error of the mean from n = 12 (control, DMSO, and DHA) or n = 4 (inhibitors) independent cultures and assays. The bar labeled with the asterisks is significantly (p < 0.05) different from the control value. Washington Academy of Sciences 25 Discussion T27A is a line of murine B lymphoblast cells with well-described susceptibility to DHA-induced cell death (Zerouga et al. 1996; Kafrawy et al. 1 998). Under normal growth conditions they possess very little DHA (Table 1). During culture in media supplemented with DHA, T27A cells incorporated considerable amounts of the fatty acid into phospholipids of their plasma membrane (Table 1, Figure 1). Delivering exogenous metabolites and drugs to cells as albumin conjugates is thought to simulate physiological delivery conditions and buffers the availability of the delivered substance. Other advantages of using albumin as a biological carrier molecule are described elsewhere (Kratz 2008; Elsadek and Kratz 2012). In phospholipids extracted from whole cells, DHA increased from less than one-half of 1% of the total in control cells to over 10% of total phospholipid fatty acids in cells cultured with supplemental DHA. In purified plasma membrane preparations the percentage rose from close to 1% to over 18%. These results agree well with previous data from these cells (Williams et al. 1998) and with reports showing that DHA has a powerful effect on both the composition and structure of their plasma membranes (Zerouga et al. 1996; Zerouga et al. 1997; Williams et al. 1998; Williams et al. 1999). These observations suggest that the metabolism of fatty acids in these leukemic cells favors the non-random incorporation of DHA into cell membranes with a preferential incorporation of DHA into phospholipids of the plasma membrane. Preferential incorporation of DHA into the plasma membranes of T27A cells has been observed previously (Jenski et al. 1993; Pascale et al. 1993; Williams et al. 1998; Williams et al. 1999). Culture of T27A cells in DHA-enriched media caused a dose- dependent decrease in cell density and cell viability, an increase in the percent of cells exhibiting blebs, and the activation of cellular caspases. OA did not induce these effects (Figures 2 and 3). We chose OA as the control fatty acid for this study because it is the most abundant fatty acid in many cell types, it is not toxic to T27A cells (Kafrawy et al. 1 998) and because in other cells types and in model membranes it neither induces apoptosis nor influences membrane raft function or structure (Kishida et al. 2006; Shaikh et al. 2009; Shaikh et al. 2009a). Figures 2 and 3 show Summer 2015 26 that when cells were cultured in media containing DHA at concentrations above approximately 0.3 mM, their rate of proliferation was slowed and significant numbers of cells died. The observation of induction of cell death near 0.3 mM is consistent with what we and others have found in this cell line (Zerouga et al. 1996; Williams et al. 1998; Williams et al. 1999), and may have implications for human health. In a study of healthy men and women the serum concentration of DHA-phospholipids was found to be near 0.15 mM. That concentration rose to over 0.35 mM after six weeks of dietary supplementation with DHA capsules (Conquer and Holub 1998). In a separate study of 234 healthy men, the mean serum concentration of DHA-phospholipids was 0.18 mM and was elevated to over 0.31 mM by similar DHA capsule supplementation (Grimsgaard et al. 1997). These studies show that the DHA levels used to reduce the growth and proliferation of mouse leukemia cells in vitro can be achieved in humans by dietary manipulation. At all concentrations of DHA examined, cells exhibited distinct exvaginations or blebs on their plasma membranes (Figure 4). Though the significance of these structures is not well understood, they are widely recognized as a hallmark of apoptosis (Charras 2008). The definitive indicator of apoptosis is the presence of active caspases (Galluzzi et al. 2011) and in these cells culture in medium containing 0.61 mM DHA resulted in the activation of caspases -3, -8, and -9 (Figure 5). These observations establish that DHA induces apoptosis in T27A cells. In general, apoptosis can be triggered by two separate, but linked, pathways: the intrinsic and extrinsic pathways (Portt et al. 2011; Galluzzi et al. 2011). The intrinsic pathway originates with mitochondria and involves the release from the intermembrane space of pro-apoptotic molecules, particularly cytochrome c. The released cytochrome c initiates a series of events that result in the conversion of inactive procaspase-9 into active caspase-9. Caspase-9 then activates caspase-3 which is responsible for setting off the series of down-stream events characteristic of apoptosis. The extrinsic pathway involves death receptors located in the plasma membrane of the cell. Binding of an appropriate ligand to a death receptor initiates an apoptotic cascade that begins with the conversion of inactive procaspase-8 into active caspase-8. Depending on the type of cell, caspase-8 then activates caspase-3 directly or indirectly Washington Academy of Sciences 27 by converting the protein Bid into tBid which activates caspase-9 (Portt et al. 2011; Galluzzi et al. 2011). Caspases -3, -8, and -9 are all active in T27A cells after exposure to DHA (Figure 5) and the inhibition of any one of them prevents the cells from undergoing apoptosis (Figure 6). Since caspase-3 is an effector caspase acting downstream of the initiator caspases -8 and -9, it appears that a linear cascade of activation events occurs whereby one initiator caspase activates the other (i.e. either caspase-8 activates caspase-9 or vice versa) and the latter then activates caspase-3. In some cell types both caspases -8 and -9 are able to activate caspases-3 directly (Slee et al. 1 999; Peter and Krammer 2003), but apparently in T27A cells under the conditions used here one of these caspases is unable to do so. It is possible that one of the initiator caspases (-8 or -9) activates capsase-3, then caspase-3 activates the remaining initiator caspase (Ozoren and El-Deiry 2003), but this is also not the case here because the activation of caspase- 3 initiates the irreversible stages of apoptosis (the execution pathway) and thus the inhibition of the initiator caspase that was activated by caspases- 3 would not result in the rescue from cell death shown in Figure 5. The data presented here suggest that one of the initiator caspases activates the other yet is itself unable to activate caspase-3. These results are consistent with T27A cells belonging to the type II group of apoptotic cells (Scaffidi et al. 1998; Ozoren and El-Deiry 2002). In cells able to undergo type I apoptosis, death receptor/ligand binding results in the direct activation of effector caspases like caspase-3. Most cells undergo type II apoptosis, in which death receptor/ligand binding is indirectly linked to the activation of effector caspases through the mitochondrion-dependent pathways via Bid and tBid (Scaffidi et al. 1998; Ozoren and El-Deiry 2002; Blanarova et al. 2011). These observations are consistent with a pathway in T27A cells in which DHA induces apoptosis by first triggering caspase-8 which in turn activates caspase-9 to initiate the effector caspases. Lipid rafts are dynamic and ephemeral laterally segregated assemblies of the plasma membrane that are rich in sphingolipids, cholesterol, and acylated and glycosylphosphatidylinositol (GPI)- anchored proteins (Simons and Ikonen 1997; Lingwood and Simons 2010). Lipid rafts serve as important platforms for the regulation of cell processes by confining and concentrating receptors and enzymes from the Summer 2015 28 surrounding membrane. They represent another selective cellular compartment that can co-localize and modulate the activities of these proteins (Simons and Ikonen 1997; Lingwood and Simons 2010). Death receptors, a subset of the tumor necrosis factor receptor superfamily, are among those receptors that have been shown to be regulated by lipid rafts. They include tumor necrosis factor receptor- 1 (TNF-R1, p55), death receptor (DR) 3 (WSL-l/APO-3), DR4 (tumor necrosis factor-related apoptosis-inducing ligand receptor- 1 [TRAIL-R1]), DR5 (TRAIL- R2/APO-2), DR6 and CD95 (Fas/APO-1) (Ashkenazi and Dixit 1998; Lavrik 2011). These receptors initiate extrinsic apoptosis after ligand binding or ligand-independent clustering of receptors (Fumarola et al. 2001; Scheel-Toellner et al. 2004). Only when located within lipid rafts do death receptors facilitate the activation of caspase-8 and down-stream events leading to apoptosis. Death receptors do not activate caspase-8 when located in non-raft regions of the membrane (Xu et al. 2009; Gajate et al. 2009; Blanarova et al. 2011). Lipid raft dysfunction has previously been implicated in the DHA- induced cell death of T27A cells (Williams et al. 1998; Williams et al. 1999). Other work has shown that DF1A alters the structure (Wassail and Stillwell 2008), size (Chapkin et al. 2008; Rockett et al. 2012) and protein composition (Rogers et al. 2010) of lipid rafts. Recently, Shaikh's group has shown that DHA has profound effects on mammalian immune function and that these effects arise from the influence of DHA on the lipid rafts of B cells (Rockett et al. 2012; Gurzell et al. 2013). Other evidence convincingly shows that DHA alters the raft-localization of epidermal growth factor receptor (Schley et al. 2007; Rogers et al. 2010), caveolin-1 (Li et al. 2007), toll-like receptors (Wong et al. 2009), the major histocompatibility complex (MHC) class I proteins (Ruth et al. 2009; Shaikh et al. 2009), the signaling molecules SFK, Lck, Fyn, and c- Yes (Stulnig et al. 1998; Stulnig et al. 2001; Chen et al. 2007), the interleukin-2 receptor (Li et al. 2005), phospholipase D1 (Diaz et al. 2002), endothelial nitric oxide synthase (Li et al. 2007; Matesanz et al. 2010), and protein kinase C (Fan et al. 2004). Combined with the data presented here, these observations suggest that DHA has an influence on death receptor-mediated apoptosis via an action on lipid rafts. This conclusion is reinforced by studies showing that a number of structurally diverse anti-tumor agents selectively induce apoptosis in cancer cells by Washington Academy of Sciences 29 triggering apoptosis thorough Fas-clustering in lipid rafts (Xu et al. 2009; Mollinedo et al. 2010; Blanarova et al. 2011). Conclusions When T27A leukemic cells are cultured in media enriched with DHA, the cells take up the fatty acid and incorporate it into their membranes, particularly the plasma membrane. Culture in DHA-enriched medium also causes cell death by inducing apoptosis. This induction of apoptosis is caused by the initiation of the extrinsic apoptotic pathway and with a linear activation of caspases in the sequence caspase-8, caspase-9, then caspase-3. Coupled with previous observations by us and others, the data suggest that the first step in DHA-induced cell death in T27A leukemia cells involves the activation of plasma membrane-associated death receptors by an influence of DHA on lipid rafts. Acknowledgements I am very grateful for the excellent assistance of Meagan E. O. Smith and Marina Acocella. I also wish to thank Matt Anderson, Lisa Fitzgerald-Miller, Kendra Model, Katrina Moncure, Melissa Moore, Shaun Smith and Jeremy White. I thank Drs. J. Stribling and P. Erikson for reading and commenting on drafts of the manuscript. The support of the Department of Biological Sciences and of the Henson School of Science and Technology of Salisbury University is acknowledged and greatly appreciated. Summer 2015 30 References Ashkenazi A, Dixit VM. 1998. Death receptors: signaling and modulation. Science 281:1305-1308. Berger A, Roberts MA, Hoff B. 2006. How dietary arachidonic- and docosahexaenoic- acid rich oils differentially affect the murine hepatic transcriptome. Lipids Health Dis 5:10. Blanarova OV, Jelinkova I, Szoor A, Skender B, Soucek K, Horvath V, Vaculova A, Andera L, Sova P, Szollosi J, Hofmanova J, Vereb G, Kozubik A. 201 1 . Cisplatin and a potent platinum (IV) complex-mediated enhancement of TRAIL-induced cancer cells killing is associated with modulation of upstream events in the extrinsic apoptotic pathway. Carcinogenesis 32:42-51. Blanckaert V, Ulmann L, Mimouni V, Antol J, Brancquart L, Chenais B. 2010. Docosahexaenoic acid intake decreases proliferation, increases apoptosis and decreases the invasive potential of the human breast carcinoma cell line MDA-MB-231. Int J Oncol 36:737-742. Bligh EG, Dyer WJ. 1 959. A rapid method of total lipid extraction and purification. Can J Biochem Physiol 37:91 1-917. Calder PC. 2012. Mechanisms of action of (n-3) fatty acids. J Nutr 142:592S-599S. Chapkin RS, Kim W, Lupton JR, McMurray DN. 2009. Dietary docosahexaenoic and eicosapentaenoic acid: emerging mediators of inflammation. Prostag Leukotr Ess 81:187-191. Chapkin RS, Wang N, Fan YY, Lupton JR, Prior IA. 2008. Docosahexaenoic acid alters the size and distribution of cell surface microdomains. Biochim Biophys Acta 1778:466-471. Charras GT. 2008. A short history of blebbing. J Microsc 231:466- 478. Chen W, Jump DB, Esselman WJ, Busik JV. 2007. Inhibition of cytokine signaling in human retinal endothelial cells through modification of caveolae/lipid rafts by docosahexaenoic acid. Invest Ophth Vis Sci 48:18-26. Colas RA, Shinohara M, Dalli J, Chiang N, Serhan CN. 2014. Identification and signature profiles for pro-resolving and inflammatory Washington Academy of Sciences 31 lipid mediators in human tissue. Am J Physiol - Cell Physiol 307:C39- C54. Conover WJ, Johnson ME, Johnson MM. 1981 . A comparative study of tests for homogeneity of variances, with applications to the outer continental shelf bidding data. Technometrics 23:351-361. Conquer JA, Holub BJ. 1998. Effect of supplementation with different doses of DHA on the levels of circulating DHA as non-esterified fatty acid in subjects of Asian Indian background. J Lipid Res 39:286-292. Dalli J, Ramon S, Norris PC, Colas RA, Serhan CN. 2015. Novel proresolving and tissue-regenerative resolvin and protectin sulfido- conjugated pathways. FASEB J. Diaz O, Berquand A, Dubois M, Di AS, Sette C, Bourgoin S, Lagarde M, Nemoz G, Prigent AF. 2002. The mechanism of docosahexaenoic acid-induced phospholipase D activation in human lymphocytes involves exclusion of the enzyme from lipid rafts. J Biol Chem 277:39368-39378. Eder K, Reichlmayr-Lais AM, Kirchgessner M. 1992. Studies on the methanolysis of small amounts of purified phospholipids for gas chromatographic analysis of fatty acid methyl esters. J Chromatog 607:55-67. Elsadek B, Kratz F. 2012. Impact of albumin on drug delivery - New applications on the horizon. J Control Release 157:4-28. Fan YY, Ly LH, Barhoumi R, McMurray DN, Chapkin RS. 2004. Dietary docosahexaenoic acid suppresses T cell protein kinase C theta lipid raft recruitment and IL-2 production. J Immunol 1 73:6151-6160. Fumarola C, Zerbini A, Guidotti GG. 2001. Glutamine deprivation- mediated cell shrinkage induces ligand-independent CD95 receptor signaling and apoptosis. Cell Death Differ 8:1004-1013. Gajate C, Gonzalez-Camacho F, Mollinedo F. 2009. Involvement of raft aggregates enriched in Fas/CD95 death-inducing signaling complex in the antileukemic action of edelfosine in Jurkat cells. PLoS One 4:e5044. Accessed on July 23 2015 and available at http://joumals.plos. org/plosone/article?id=l 0. 1 37 1 /journal, pone.0005044. Galluzzi L, Vitale I, Abrams JM, Alnemri ES, Baehrecke EH, Blagosklonny MV, Dawson TM, Dawson VL, El-Deiry WS, Fulda S, Gottlieb E, Green DR, Hengartner MO, Kepp O, Knight RA, Kumar S, Lipton SA, Lu X, Madeo F, Malorni W, Mehlen P, Nunez G, Peter ME, Summer 2015 32 Piacentini M, Rubinsztein DC, Shi Y, Simon HU, Vandenabeele P, White E, Yuan J, Zhivotovsky B, Melino G, Kroemer G. 2011. Molecular definitions of cell death subroutines: recommendations of the Nomenclature Committee on Cell Death 2012. Cell Death Differ 19:107- 120. Grimsgaard S, Bonaa KH, Hansen JB, Nordoy A. 1997. Highly purified eicosapentaenoic acid and docosahexaenoic acid in humans have similar triacylglycerol-lowering effects but divergent effects on serum fatty acids. Am J Clin Nutr 66:649-659. Gurzell EA, Teague H, Harris M, Clinthorne J, Shaikh SR, Fenton JI. 2013. DHA-enriched fish oil targets B cell lipid microdomains and enhances ex vivo and in vivo B cell function. J Leukocyte Biol 93:463- 470. Jenski LJ, Sturdevant LK, Ehringer WD, Stillwell W. 1993. Omega-3 fatty acid modification of membrane structure and function. I. Dietary manipulation of tumor cell susceptibility to cell- and complement- mediated lysis. Nutr Cancer 19:135-146. Jenski LJ, Zerouga M, Stillwell W. 1995. Omega-3 fatty acid containing liposomes in cancer therapy. Proc Soc Exp Biol Med 210:227- 233. Kaduce TL, Awad AB, Fontenelle LJ, Spector AA. 1977. Effect of fatty acid saturation on a-aminoisobutyric acid transport in Ehrlich ascites cells. J Biol Chem 252:6624-6630. Kafrawy O, Zerouga M, Stillwell W, Jenski LJ. 1998. Docosahexaenoic acid in phosphatidylcholine mediates cytotoxicity more effectively than other omega-3 and omega-6 fatty acids. Cancer Lett 132:23-29. Kang KS, Wang P, Yamabe N, Fukui M, Jay T, Zhu BT. 2010. Docosahexaenoic acid induces apoptosis in MCF-7 cells in vitro and in vivo via reactive oxygen species formation and caspase 8 activation. PLoS One 5 :e 10296. Accessed on July 23 2015 available at http://journals.plos.org/plosone/article7icUl 0. 1 371 /journal. pone. 00 1 0296. Kishida E, Tajiri M, Masuzawa Y. 2006. Docosahexaenoic acid enrichment can reduce L929 cell necrosis induced by tumor necrosis factor. BBA - Mol Cell Biol L 1761 :454-462. Washington Academy of Sciences 33 Kratz F. 2008. Albumin as a drug carrier: Design of prodrugs, drug conjugates and nanoparticles. J Control Release 1 32: 1 71 -1 83. Lang I, Fick A, Schafer V, Giner T, Siegmund D, Wajant H. 2012. Signaling active CD95 receptor molecules trigger co-translocation ot inactive CD95 molecules into lipid rafts. J Biol Chem 287:24026-24042. Lavrik I. 2011. Regulation of death receptor-induced apoptosis induced via CD95/Fas and other death receptors. Mol Biol 45:150-155. Li Q, Wang M, Tan L, Wang C, Ma J, Li N, Li Y, Xu G, Li J. 2005. Docosahexaenoic acid changes lipid composition and interleukin-2 receptor signaling in membrane rafts. J Lipid Res 46:1904-1913. Li Q, Zhang Q, Wang M, Liu F, Zhao S, Ma J, Luo N, Li N, Li Y, Xu G, Li J. 2007. Docosahexaenoic acid affects endothelial nitric oxide synthase in caveolae. Arch Biochem Biophys 466:250-259. Lien EL. 2009. Toxicology and safety of DHA. Prostag Leukotr Ess 81:125-132. Lingwood D, Simons K. 2010. Lipid rafts as a membrane-organizing principle. Science 327:46-50. Matesanz N, Park G, McAllister H, Leahey W, Devine A, McVeigh GE, Gardiner TA, McDonald DM. 2010. Docosahexaenoic acid improves the nitroso-redox balance and reduces VEGF-mediated angiogenic signaling in microvascular endothelial cells. Invest Ophth Vis Sci 51:6815-6825. Matta JA, Miyares RL, Ahern GP. 2007. TRPV1 is a novel target for omega-3 polyunsaturated fatty acids. J Physiol 578:397-41 1. Mischoulon D, Freeman MP. 2013. Omega-3 fatty acids in psychiatry. Psychiat Clin N Am 36:15-23. Mitchell DC, Niu SL, Litman BJ. 2003. DFlA-rich phospholipids optimize G-Protein-coupled signaling. J Pediatr 143:S80-S86. Mollinedo F, de la Iglesia-Vicente J, Gajate C, de Mendoza AEH, Villa-Pulgarin JA, de Frias M, Roue G, Gil J, Colomer D, Campanero MA, Blanco-Prieto MJ. 2010. In vitro and in vivo selective antitumor activity of edelfosine against mantle cell lymphoma and chronic lymphocytic leukemia involving lipid rafts. Clin Cancer Res 16:2046- 2054. Summer 2015 34 Molnar J, Markovic G, Chao H, Molnar Z. 1969. Glycoproteins of Ehrlich ascites carcinoma cells: Separation of plasma and endoplasmic reticular membrane fragments. Arch Biochem Biophys 134:524-532. Niu SL, Mitchell DC. 2005. Effect of packing density on rhodopsin stability and function in polyunsaturated membranes. Biophys J 89: 1833- 1840. ' OzorenN, El-Deiry WS. 2002. Defining characteristics of Types I and II apoptotic cells in response to TRAIL. Neoplasia 4:551-1557. Ozoren N, El-Deiry WS. 2003. Cell surface death receptor signaling in normal and cancer cells. Semin Cancer Biol 13:135-147. Pascale AW, Ehringer WD, Stillwell W, Sturdevant LK, Jenski LJ. 1993. Omega-3 fatty acid modification of membrane structure and function. II. Alteration by docosahexaenoic acid of tumor cell sensitivity to immune cytolysis. Nutr Cancer 19:147-157. Peter ME, Krammer PH. 2003. The CD95 (APO-l/Fas) DISC and beyond. Cell Death Differ 10:26-35. Pinot M, Vanni S, Pagnotta S, Lacas-Gervais S, Payet LA, Ferreira T, Gautier R, Goud B, Antonny B, Barelli H. 2014. Polyunsaturated phospholipids facilitate membrane deformation and fission by endocytic proteins. Science 345:693-697. Portt L, Norman G, Clapp C, Greenwood M, Greenwood MT. 2011. Anti-apoptosis and cell survival: A review. BBA - Mol Cell Res 1813:238- 259. Rockett BD, Teague H, Harris M, Melton M, Williams J, Wassail SR, Shaikh SR. 2012. Fish oil increases raft size and membrane order of B cells accompanied by differential effects on function. J Lipid Res 53:674- 685. Rogers KR, Kikawa KD, Mouradian M, Hernandez K, McKinnon KM, Ahwah SM, Pardini RS. 2010. Docosahexaenoic acid alters epidermal growth factor receptor-related signaling by disrupting its lipid raft association. Carcinogenesis 31:1523-1530. Ruth MR, Proctor SD, Field CJ. 2009. Feeding long-chain n-3 polyunsaturated fatty acids to obese leptin receptor-deficient JCR:LA- cp rats modifies immune function and lipid-raft fatty acid composition. Br J Nutr 101:1341-1350. Washington Academy of Sciences 35 Salem N, Kim H-Y, Yergey JA. 1986. Docosahexaenoic acid: membrane function and metabolism. In: Simopoulos AP, Kifer RR, Martin RE, editors. Health effects of polyunsaturated fatty acids in seafoods. Orlando , Florida: Academic Press, p 263-317. Scaffidi C, Fulda S, Srinivasan A, Friesen C, Li F, Tomaselli KJ, Debatin KM, Krammer PH, Peter ME. 1998. Two CD95 (APO-l/Fas) signaling pathways. EMBO J 17:1675-1687. Scheel-Toellner D, Wang K, Assi LK, Webb PR, Craddock RM, Salmon M, Lord JM. 2004. Clustering of death receptors in lipid rafts initiates neutrophil spontaneous apoptosis. Biochem Soc Trans 32:679- 681. Schley PD, Brindley DN, Field CJ. 2007. (n-3) PUFA alter raft lipid composition and decrease epidermal growth factor receptor levels in lipid rafts of human breast cancer cells. J Nutr 137:548-553. Serhan CN, Dalli J, Colas RA, Winkler JW, Chiang N. 2014. Protectins and maresins: New pro-resolving families of mediators in acute inflammation and resolution bioactive metabolome. BBA - Mol Cell Biol L 1851:397-413. Serini S, Piccioni E, Merendino N, Calviello G. 2009. Dietary polyunsaturated fatty acids as inducers of apoptosis: implications for cancer. Apoptosis 14:135-152. Shaikh SR. 2010. Diet-induced docosahexaenoic acid non-raft domains and lymphocyte function. Prostag Leukotr Ess 82:159-164. Shaikh SR, LoCascio DS, Soni SP, Wassail SR. Stillwell W. 2009. Oleic- and docosahexaenoic acid-containing phosphatidylethanolamines differentially phase separate from sphingomyelin. BBA - Biomembranes 1788:2421-2426. Shaikh SR, Rockett BD, Salameh M, Carraway K. 2009a. Docosahexaenoic acid modifies the clustering and size of lipid rafts and the lateral organization and surface expression of MHC class I of EL4 cells. J Nutr 139:1632-1639. Siddiqui RA, Harvey K, Stillwell W. 2008. Anticancer properties of oxidation products of docosahexaenoic acid. Chem Phys Lipids 153:47- 56. Summer 2015 36 Siddiqui RA, Shaikh SR, Sech LA, Yount HR, Stillwell W, Zaloga GP. 2004. Omega 3-fatty acids: health benefits and cellular mechanisms of action. Mini Rev Med Chem 4:859-871. Simons K, Ikonen E. 1 997. Functional rafts in cell membranes. Nature 387:569-572. Slee EA, Harte MT, Kluck RM, Wolf BB, Casiano CA, Newmeyer DD, Wang HG, Reed JC, Nicholson DW, Alnemri ES, Green DR, Martin SJ. 1999. Ordering the cytochrome c-initiated caspase cascade: hierarchical activation of caspases-2, -3, -6, -7, -8, and -10 in a caspase- 9-dependent manner. J Cell Biol 144:281-292. Sokal RR, Rohlf FJ. 1981. Biometry. New York: W.H. Freeman and Company. 428 p. Spector AA, Hoak JC. 1969. An improved method for the addition of long-chain free fatty acid to protein solutions. Anal Biochem 32:297-302. Stillwell W. 2008. Docosahexaenoic acid: a most unusual fatty acid. Chem Phys Lipids 153:1-2. Stillwell W, Shaikh SR, Zerouga M, Siddiqui RA, Wassail SR. 2005. Docosahexaenoic acid affects cell signaling by altering lipid rafts. Reprod Nutr Dev 45:559-579. Stillwell W, Wassail SR. 2003. Docosahexaenoic acid: membrane properties of a unique fatty acid. Chem Phys Lipids 126:1-27. Stulnig TM, Berger M, Sigmund T, Raederstorff D, Stockinger H, Waldhausl W. 1998. Polyunsaturated fatty acids inhibit T cell signal transduction by modification of detergent-insoluble membrane domains. J Cell Biol 143:637-644. Stulnig TM, Huber J, Leitinger N, Imre EM, Angelisova P, Nowotny P, Waldhausl W. 2001. Polyunsaturated eicosapentaenoic acid displaces proteins from membrane rafts by altering raft lipid composition. J Biol Chem 276:37335-37340. Teague H, Ross R, Harris M, Mitchell DC, Shaikh SR. 2013. DHA- fluorescent probe is sensitive to membrane order and reveals molecular adaptation of DHA in ordered lipid microdomains. J Nutr Biochem 24:188-195. Wassail SR, Stillwell W. 2008. Docosahexaenoic acid domains: the ultimate non-raft membrane domain. Chem Phys Lipids 153:57-63. Washington Academy of Sciences 37 Webb Y, Hermida-Matsumoto L, Resh MD. 2000. Inhibition of protein palmitoylation, raft localization, and T cell signaling by 2- bromopalmitate and polyunsaturated fatty acids. J Biol Chem 275:261- 270. Williams EE, Jenski LJ, Stillwell W. 1998. Docosahexaenoic acid (DHA) alters the structure and composition of membranous vesicles exfoliated from the surface of a murine leukemia cell line. Biochim Biophys Acta 1371:351-362. Williams EE, May BD, Stillwell W, Jenski LJ. 1999. Docosahexaenoic acid (DHA) alters the phospholipid molecular species composition of membranous vesicles exfoliated from the surface of a murine leukemia cell line. Biochim Biophys Acta 1418:185-196. Williams EE, Somero GN. 1996. Seasonal-, tidal-cycle- and microhabitat-related variation in membrane order of phospholipid vesicles from gills of the intertidal mussel Mytilus californianus . J Exp Biol 199:1587-1596. Wong SW, Kwon MJ, Choi AMK, Kim HP, Nakahira K, Hwang DH. 2009. Fatty acids modulate Toll-like receptor 4 activation through regulation of receptor dimerization and recruitment into lipid rafts in a reactive oxygen species-dependent manner. J Biol Chem 284:27384- 27392. Wren JJ. 1960. Chromatography of lipids on silicic acid . J Chromatog 4:173-195. Xu ZX, Ding T, Haridas V, Connolly F, Gutterman JU. 2009. Avicin D, a plant triterpenoid, induces cell apoptosis by recruitment of Fas and downstream signaling molecules into lipid rafts. PLoS One 4:e8532. Accessed on July 23 2015 available at http://joumals.plos.org/plosone/article7idH0.1371/journal.pone.0008532. Yamagami T, Porada CD, Pardini RS, Zanjani ED, Almeida-Porada G. 2009. Docosahexaenoic acid induces dose dependent cell death in an early undifferentiated subtype of acute myeloid leukemia cell line. Cancer Biol Ther 8:331-337. Zerouga M, Jenski LJ, Booster S, Stillwell W. 1997. Can docosahexaenoic acid inhibit metastasis by decreasing deformability of the tumor cell plasma membrane? Cancer Lett 1 19:163-168. Summer 2015 38 Zerouga M, Stillwell W, Stone J, Powner A, Dumaual AC, Jenski LJ. 1996. Phospholipid class as a determinant in docosahexaenoic acid's effect on tumor cell viability. Anticancer Res 16:2863-2868. Bio Eugene Williams is an educator and scientist who works in the area of cell physiology. He is interested in how fishes acclimatize and adapt to changes in temperature and in understanding the curious relationship between the oils of cold-water fishes and cancers. Williams earned a Ph.D. at Arizona State University and is currently a Professor of Biological Sciences at Salisbury University. Washington Academy of Sciences 39 A Nineteenth Century Historical Analysis of Game Warden Efforts: Focus on Rabbits and Hares Kelsey Gilcrease South Dakota School of Mines and Technology Abstract Leporids (rabbits and hares) were widely sought-after game animals to many people in the nineteenth century. But how often were offenses to the game laws caught? That answer depends on the number of wardens, the amount of prosecuted leporid offenses (as compared to other offenses), and how complex it was to catch an offense. The aim of this paper is to determine the types of offenses that game wardens enforced, the number of prosecuted leporid offenses, the specific types of leporid offenses in two states — New Jersey and Massachusetts — and the New Jersey counties where leporid enforcements occurred. This investigation uncovered three key findings: (1) The more uniformity of time spent between wardens on offenses could be a success factor in catching leporid offenses; (2) There was a correlation between the number of wardens who cited leporid offenses and the number of counties involved with the leporid offenses; and (3) Outside of those years, there was a disproportionate number of leporid offenses when correlating for the numbers of wardens and numbers of counties. Furthermore, the results of this investigation offer implications toward our understanding of past leporid conservation, most notably findings related to the uniformity of the number of wardens prosecuting leporid offenses and the years when prosecuted leporid offenses were prominent in the nineteenth century. Introduction The historical emphasis of managing wildlife was largely synonymous with managing game species and predators (Bolen and Robinson, 2003, pp. 20, 183). For example, the first game laws in North America occurred in 1639 to close the white-tailed deer hunting season for six months (Bolen and Robinson, 2003). Protection of game is critical in efforts to conserve wildlife because of the need to understand biology at the organismal level and to conserve habitats (Willis et al., 2008). However, the very beginnings of enforcement of wildlife laws in the United States tell a rugged story. Beginning with colonial times, game law enforcers intended that private citizens act as “advisors'” regarding game laws to ensure that fellow hunters were following the laws (Lund, 1980), Summer 2015 40 which would mostly help to inform the privileged class in how to protect or manage game resources on their own lands. Yet if private citizens were enforcing the game laws, how effective were they at protecting wild game? In historical analyses of wildlife enforcement during the nineteenth century, Lund (1980), Tober (1981), and Stockdale (1993) indicated that enforcement of laws pertaining to wildlife was weak or non-existent, since it was the people who were conducting the enforcements and this perception was criticized by the local townspeople. Early wildlife laws too, ignored the bag limit, further making enforcement of the laws difficult (Lund, 1980). As enforcement of laws and compliancy is important to the conservation of wildlife, the historical nineteenth-century rationale or prioritization of the protection of leporids (rabbits and hares) in New Jersey and Massachusetts (the two states that are the focus of this study) appears unclear1. What offenses did the wardens focus their time on? And which counties were enforcing the leporid laws? These questions lead to further questions as to whether certain counties in New Jersey and Massachusetts reported more leporid offenses than others, and what factors were most strongly associated with whether and how game law violations were reported. Furthermore, research regarding wildlife conservation officers is limited (Archbold, 2012) so there is a need for more of a historical underpinning on the number of conservation offenses, the number of wardens, and the county distribution of the offenses. These questions are important because leporids were often hunted for food ( Omaha Daily Bee , 1887, Si. Paul Daily Globe , 1887) at a time when some North American leporid populations started declining. Let’s look at the two states which, again, are the focus of this study - New Jersey and Massachusetts. The leporids of New Jersey include the Eastern cottontail (Sylvilagus floridanus ), introduced species of the European hare ( Lepus europaeus), the black-tailed jackrabbit (Lepus calif or nicus), the white- tailed jackrabbit (Lepus townsendii ) (State of New Jersey, 2004) and, at one time, the snowshoe hare (Lepus americanus ) (Rhoads, 1903). However, current records indicate no presence of the snowshoe hare in New Jersey (Murray and Smith, 2008). Washington Academy of Sciences 41 The leporids of Massachusetts also include the Eastern cottontail, snowshoe hare, and black-tailed jackrabbit (Massachusetts Executive Office of Energy and Environmental Affairs, 2014). In Massachusetts, the New England cottontail ( Sylvilcigus transUionalis ) currently has a conservation status of candidate species (USFWS, 2015). The purpose of this paper is to determine the types of leporid offenses; the number of prosecuted leporid offenses; the types of offenses game wardens enforced; and the counties that prosecuted the enforcements — rather than to decipher why game wardens focused on specific leporid species offenses. Methods Based on the Annual Reports of the Board of Fish and Game Commissioners in New Jersey (1894-1899) and the Reports of the Commissioners on Inland Fisheries and Game in Massachusetts (1889- 1 899), I categorized the nineteenth-century wildlife offenses for those two states into these eight groups: fish, illegal fishing, lobster, Sunday offenses, pollution, illegal game, trespassing, and any offenses related to game generally, such as squirrels, deer, ducks, birds, and leporids. Table 1 presents a description of each of these eight categories. Each year from 1894-1899 for New Jersey and 1889-1899 for Massachusetts, I recorded the: • total number of wardens who were holding a warden status in New Jersey and Massachusetts, • total number of offenses2, • number of fish, illegal fishing, Sunday hunting, pollution, illegal game, trespassing, and other game offenses (including the number of wardens who cited leporid offenses), • number of counties with leporid offenses, • names of the wardens who cited leporid offenses, and • counties in which the offenses occurred. The Simpson’s E or Even-ness index (widely employed for biodiversity studies), was used to determine the “even-ness” of the prosecuted leporid offenses and how “evenly” the leporid offenses occurred in each county (National Center for Ecological Analysis and Synthesis Summer 2015 42 (NCEAS), 2014). Simpson’s E can range from 0 to 1, with 1 being the highest uniformity (NCEAS, 2014). Because there was usually only one warden spending time on leporid offenses in Massachusetts, the even-ness index had to include the total number of wardens available, regardless of whether all the wardens worked on game such as leporids. Table 1. Offenses pursued in New Jersey and Massachusetts and characterization of the offenses Offense Description of the Offense Fish Any fishing offense related to possession of a certain fish species caught, or if the fish was under a legal limit size Illegal fishing Offenses related to the use of nets or other equipment used to illegally catch fish Lobster Any offense related to lobster catching - e.g., size was too small, mutilation was involved, offense was over the limit Sunday hunting Offenses occurring on Sunday (and no hunting or fishing was allowed on Sunday) Pollution Any pollution that occurred Illegal game Any attempt to take game or the act of possessing illegal game Trespassing Offenses where people were caught trespassing on property not belonging to them Squirrels, deer, ducks, song birds, or leporids Offenses where any of these animals were caught out of season or over the legal limit; any attempt to kill these animals; the use of certain illegal methods to catch these animals (i.e., snares) Regarding the county data: In New Jersey only the years 1 894-1895, 1898, and 1899 contained county data. Massachusetts county data were not applicable because county data were sparse, and due to the nature of the historical documents, it was difficult to decipher where the offenses originated. For the New Jersey data, correlation analysis was used to determine the relationship between the number of wardens who cited leporid offenses and the number of counties involved with leporid offenses in the state of New Jersey. Washington Academy of Sciences 43 Results Findings on New Jersey In New Jersey, fishing and illegal fishing activities were the more commonly-reported offenses in 1894-1895 (49% of the overall time). In 1896, Sunday hunting and offenses with song birds were the more commonly reported offenses (63% of the overall time). In 1897, illegal fishing activities and offenses with song birds were more commonly reported offenses in the state (60% of the time). By 1898 and 1899, Sunday hunting and offenses with song birds were more commonly reported (57% of the time for both years). The above overall prosecuted offenses provide the metrics by which annual percentages of prosecuted leporid offenses are calculated for the state of New Jersey during those same years (Figure 1). Percentage of leporid offenses in New Jersey Figure 1. New Jersey: Percentage of leporid offenses, among other types of offenses. In New Jersey, the number of wardens each year did not uniformly reflect the number of leporid offenses. The highest number of wardens involved with leporid offenses occurred in 1896 with 14 wardens. At the same time, 1896 had the lowest percentage of wardens reporting leporid offenses (Figure 1). Simpson E values related to the even-ness of wardens prosecuting leporid offenses in each year are reported in Table 2. Summer 2015 44 Table 2. New Jersey: Even-ness of wardens prosecuting leporid offenses. Year 1894- 1895 1896 1897 1898 1899 Even-ness .7475 .78 .394 .809 .685 In New Jersey, the number of counties in which leporid offenses were prosecuted each year did not uniformly reflect the total leporid offenses in the state. The most leporid offenses (relative to other offenses) occurred in 1 898 and 1 899 (Figure 1); however, the least number of counties that cited leporid offenses occurred in 1899 (Figure 2). The year 1898 had the most even number of wardens spending time on leporid offenses (Table 2) and 1898 is also significant for the even-ness of the counties that cited leporid offenses (see Table 3, which shows Simpson E values for the even- ness of counties with prosecuted leporid offenses in 1894-1895, 1898, and 1 899). The year 1 899 was the least uniform in terms of counties with leporid offenses (Table 3), with Bergen County carrying over half of the total leporid offenses (Figure 2). The number of wardens who prosecuted leporid offenses correlates strongly to the number of counties in which offenses were prosecuted in New Jersey. Correlation co-efficient analysis showed r = 0.989 association between the number of wardens who cited leporid offenses and the number of counties involved with the leporid offenses. In 1 896 and 1 898, the number of wardens who worked on leporid offenses was more than 50% of all wardens (Figure 3); however, the most counties prosecuting leporid offenses occurred in 1898 with twelve counties involved (Figure 4). In the years 1894-1895, eight counties were involved in leporid offenses (Figure 5). Washington Academy of Sciences 45 New Jersey counties with leporid offenses in 1899 cu County Figure 2. New Jersey counties prosecuting leporid offenses, 1899. Table 3. New Jersey: Even-ness of counties prosecuting leporid offenses. Year 1894- 1895 1896 1897 1898 1899 Even-ness .7475 NA NA .842 .488 Wardens prosecuting leporid offenses in New Jersey 0.7 i 1894 1895 1896 1897 1898 1899 Figure 3. New Jersey: Number of wardens who prosecuted leporid offenses, expressed as a percentage of the total number of wardens. Summer 201 5 46 The most prominent year for prosecuting leporid offenses in New Jersey was 1898, based upon the previously discussed percentage of wardens and percentage of prosecuted leporid offenses in New Jersey (Figures 3 and 1), and also based upon the number of wardens prosecuting leporid offenses being the most uniform (Table 2). The most un-even number of wardens prosecuting leporid offenses occurred in 1897, given that, for example, Warden Dunham carried 44% of the total leporid offenses. Also, in 1897, there was a decrease (from the previous year) in the number of wardens contributing to leporid offense prosecutions (Figure 3). New Jersey counties with leporid offenses in 1898 County Figure 4. New Jersey counties prosecuting leporid offenses, 1898. New Jersey had various types of prosecuted leporid offenses, with the majority of the offenses involving killed rabbits and the lowest number of offenses for selling rabbits (Figure 6). The methods used by hunters in the killed-rabbit offenses was not mentioned. Washington Academy of Sciences 47 New Jersey counties with leporid offenses in 1894-1895 I I I I I i r County .1 Figure 5. New Jersey counties prosecuting leporid offenses, 1894-1895. Type of leporid offenses in New Jersey 14 12 10 8 6 4 2 0 “ 1 F \ FI 1 J l j ■ j l xO^ <£" * / * / * « . Retrieved on August 3, 2014. Fowler, J., L. Cohen, and P. Jarvis. 1998. Practical Statistics for Field Biology’. John Wiley and Sons. West Sussex, England. Lueck, D. 1995. Property Rights and the Economic Logic of Wildlife Institutions. Natural Resources Journal 35: 625-670. Lund, T. A. 1976. Early American Wildlife Law. New York University Law Review 5 1 : 703-730. Lund, T. A. 1980. American Wildlife Law’. University of California Press. Berkeley, CA. Murray, D. and A. T. Smith. 2008. Lepus americanus. The IUCN Red List of Threatened Species. Version 2014.2. . Retrieved on 06 August 2014. National Center for Ecological Analysis and Synthesis (NCEAS). 2014. Available at: . Retrieved August 8, 2014. Omaha Daily Bee. 1887. March 7, 1887 page 4 Available at: . Rhoads, S. N. 1903. The Mammals of Pennsylvania and New Jersey: A Biographic, Historic and Descriptive Account of the Furred Animals of Land and Sea, both living and extinct, known to have existed in these states. Wickersham Printing Company. Lancaster, PA. State of New Jersey. 1 894. Annual Report of the Board of Fish and Game Commissioners of the State of New Jersey. 1 894. Trenton, NJ. State of New Jersey. 1895. Annual Report of the Board of Fish and Game Commissioners of the State of New Jersey. 1 895. Trenton, NJ. State of New Jersey. 1 896. Annual Report of the Board of Fish and Game Commissioners of the State of New Jersey. 1 896. Trenton, NJ. State of New Jersey. 1 897. Annual Report of the Board of Fish and Game Commissioners of the State of New Jersey. 1897. Trenton, NJ. State of New Jersey. 1 898. Annual Report of the Board of Fish and Game Commissioners of the State of New Jersey. 1898. Trenton, NJ. State of New Jersey. 1 899. Annual Report of the Board of Fish and Game Commissioners of the State of New Jersey. 1 899. Trenton, NJ. State of New Jersey. 2004. Available at . Retrieved on August 3, 2014. Washington Academy of Sciences 55 St. Paul Daily Globe. 1887. February 18, 1887. Available at . Retrieved on August 9, 2014. Stockdale, M. 1993. English and American Wildlife Law: Lessons from the Past. Proceedings of the Annual Conference of the Southeastern Association of Fish and Wildlife Agencies 47: 732-739. Tober, J. A. 1981. Who Ch\’ns the Wildlife? The Political Economy of Conservation in Nineteenth-Century America. Greenwood Press. Westport, CT. United States Fish and Wildlife Service. 2015. New England Cottontail (Sylvilagus transitionalis). Available at . Retrieved on August 10, 2015 Willis D., C. Scalet, and L. D. Flake. 2008. Introduction to Wildlife and Fisheries. W. H. Freeman and Company. New York, NY. Bio Kelsey Gilcrease is a biology lab and ecology instructor at the South Dakota School of Mines and Technology, Department of Chemistry and Applied Biological Sciences. Her main research is focused on leporid conservation in North America, biogeography, and conservation biology histories in the nineteenth and early twentieth centuries. She may be contacted at Kelsev.Gilcrease@sdsmt.edu. Summer 2015 56 Washington Academy of Sciences 57 Uranus and Neptune Revisited Sethanne Howard USNO, retired Abstract Uranus and Neptune are two planets not known in ancient times. Once discovered, however, astronomers were eager to obtain their vital statistics. Of course once Voyager flew by them everyone had the information, but before there was Voyager, there were many attempts to measure things like the rotational period, the mass, the brightness, etc. This is the story of the two people who obtained the rotational period of both planets before Voyager got there. They confirmed that the spin of Uranus is retrograde and that of Neptune direct. Uranus rotates on its side. Their estimates for the periods of rotation are, for Uranus, 24 ±3 hr., and for Neptune, 15 ±3 hr. Introduction For most of human history humanity knew of only five planets (plus our own): Mercury, Venus, Mars, Jupiter, and Saturn. These are the planets that are visible to the naked eye at various times during the year. We learned relatively recently that there are two other planets in our Solar System: Uranus and Neptune. Note the proper pronunciation of Uranus (accent on the first syllable). First a little history on Uranus and Neptune. The question being what did we know before the Voyager flyby of Uranus. Uranus had been observed on many occasions before its recognition as a planet, but it was generally mistaken for a star. Then Sir William Herschel observed the planet qua planet on March 13, 1781. This was the first planet added to the Solar System since the dawn of history. See Figure 1 for a drawing of the telescope he used. He decided to name the new planet Georgium Sidus (George’s Star), in honor of his new patron, King George III of England. As one might expect, this was not popular outside England. Bode (a German astronomer) opted for Uranus , the Latinized version of the Greek god of the sky, Ouranos (the only planet with a name of Greek origin). Bode argued that just as Saturn was the father of Jupiter, the new planet should be named after the father of Saturn. Ultimately, Bode’s suggestion became the most widely used, and became universal in 1850 when Her Majesty’s Nautical Almanac Office, the final holdout, switched from using Georgium Sidus to Uranus. Summer 2015 58 Figure 1 — Discovery telescope for Uranus Uranus is the seventh planet from the Sun. It has the third-largest radius and fourth-largest mass in the Solar System. It revolves around the Sun once every 84 Earth years. Uranus has a ring system and numerous moons. The Uranian system has a unique configuration among the planets because its axis of rotation is tilted sideways, nearly into the plane of its revolution about the Sun. Its north and south poles therefore lie where most other planets have their equators. Uranus’s orbital elements (the shape of its orbit) were first calculated in 1783 by Pierre-Simon Laplace.' Over time, discrepancies began to appear between the predicted and observed orbits, and in 1841, John C. Adams" first proposed that the differences might be due to the gravitational tug of an unseen planet."1 In 1 845, Urbain Le Verrierlv began his own independent research into Uranus’s orbit. On September 23, 1 846, Johann G. Gallev was the first to see the new planet close to the position predicted by Le Verrier.vl There was considerable controversy over the name for the new planet. At first, Neptune was simply called “the planet exterior to Uranus” or “Le Verrier’s planet.” However, eventually the name Neptune , Roman god of the sea, was accepted. Neptune is the eighth and farthest planet from the Sun in the Solar System. It is the fourth-largest planet by diameter and the third-largest by Washington Academy of Sciences 59 mass. It revolves around the Sun once every 164.8 Earth years. Like Uranus, Neptune has a ring system and several moons. Uranus - the Details Uranus was known to be a bit strange. It was already suspected that its axis of rotation was off kilter. Most of the Solar System planets have rotation axes that are close to perpendicular to the plane of the orbit. The Earth, for example, has an axis tilted only 23.5° off perpendicular. Uranus, on the other hand, was thought to have a rotation axis almost in its orbital plane. No one was quite sure of this, though. Its major moons are Ariel, Umbriel, Titania, Oberon, and Miranda - names taken from Shakespeare. There are about 27 moons known. The internal structure of Uranus is shown in Figure 2. The wildly off-center magnetic field is shown in Figure 3. Figure 2 — Structure of Uranus The planet is thought to have a very small central almost rocky core, surrounded by a plasma ocean, surrounded in turn by an atmosphere with lots of hydrogen, helium, and methane (CEE). It is the methane that Summer 2015 60 makes Uranus appear cyan in color. Note in Figure 3 that the Magnetic North Pole points “downward” - below the orbital plane. The magnetic field is not centered on or near the center of the planet. This is quite unusual. This unusual geometry results in a highly asymmetric magnetosphere. By comparison, the magnetic field of Earth is roughly the same at either pole, and its “magnetic equator” is roughly parallel with its geographical equator. Figure 3 — magnetic field for Uranus (left) and Neptune (right) Neptune - the Details Figure 4 shows the structure of Neptune. Although smaller than Uranus as seen from the Earth, when seen with a large telescope it is visible as a disk. Voyager found the axial tilt of Neptune to be 28.32° - similar to the Earth's tilt. Triton is its major moon - very large as moons go. Unlike other large planetary moons in the Solar System, Triton has a retrograde orbit, indicating that it was captured rather than formed in place. There are about 14 known moons for Neptune. The magnetic field of Neptune (Figure 3) is also a bit off center although not as much as Uranus. Never visible to the naked eye, Neptune requires a 4 meter class telescope to capture its spectra, and a 50 inch telescope to work in the near infrared part of the spectrum. To me it is a beautiful planet because its color is a deep, rich blue. The atmosphere is mainly hydrogen and helium with trace amounts of CFU that contribute to that beautiful color. Washington Academy of Sciences 61 Planetary Rotation All planets rotate. The Earth rotates about its axis once a day. This is how we define a "day’. Planets revolve around the Sun and rotate about their individual axes. It is fairly straightforward to obtain the rotation periods (i.e., the length of the planetary day) of the five naked eye planets - one simply watches them. We can’t watch the more distant Uranus and Neptune. It takes a large telescope to determine their rotational periods. Figure 4 — The internal structure of Neptune: 1 . Upper atmosphere, top clouds 2. Atmosphere consisting of hydrogen, helium, and CH4 gas 3. Mantle consisting of water, ammonia, and CH4 ices 4. Core consisting of rock (silicates and nickel-iron) By using various techniques people tried to determine the rotation periods for Uranus and Neptune. A visual technique means watching the planet as it spins. This is rather like watching the Great Red Spot on Jupiter as Jupiter rotates: once around is a ‘day’ on Jupiter. Theory means that the rotation period is derived from planetary theory (using the mass and shape of the planet to derive its period), not by using a telescope. Photometry means that a telescope is used with a filter in selected wavelength bands ( e.g ., a color like infrared) to measure changes in the light from the planet. A regular and repeatable change in the light can Summer 2015 62 represent the length of the planetary day. The spectra technique means that the planet's spectral lines are used to determine the period. This last is the most difficult to do because it means measuring the minute tilt of the spectral lines, and from that tilt, the rotational period. Some of the early attempts are listed in Table I and Table II. Table I - early values for Uranus Date Period Technique Person 1872 12h Visual Buffam 1900 7h < P <12h34m Theory Houzeau 1902 retrograde Spectra Deslandres 1912 10h45m Spectra Lowell, Slipher 1916 10"49m Photometry Campbell 1930 10h50m Spectra Moore, Menzel It appears that people were closing in on a rotational period between 10h and 1 lh, and this value showed up in textbooks of the time. Actually the notion that Uranus has a rapid rotation goes back to Herschel who thought he saw a polar flattening of the planet. v" Somewhat later Laplace provided further qualitative support for HerscheTs deduction by noting that the observed co-planar nature of the satellite orbits implied that Uranus needed a substantial equatorial bulge to counteract the disruptive perturbations of the Sun. However, the moon Miranda (smallest and innermost) is on an orbit substantially inclined to the common plane of the remaining satellites. Curious. The first substantial datum on Uranus’s rotation was provided by Deslandres in 1902, who detected the tilt, induced by rotation, of reflected Fraunhofer lines in the planet's spectrum, thus proving the retrograde sense of the planet’s spin. Table II - early values for Neptune Date Period Technique Person 1884 7.92h Photometry Hall 1896 1 3 h < P < 18h Theory Tisserand 1928 15.8” Spectra Moore, Menzel 1955 12.4311 Photometry Gunther There was considerable scatter in the proposed period for Neptune. It is farther from Earth than Uranus, so it is more difficult to observe. Even the sense of its spin was uncertain (prograde or retrograde). It was often supposed that the planet probably rotated in a retrograde sense. Finally, Washington Academy of Sciences 63 since Moore and Menzel’s work on Neptune was unique, a re-examination of the sense of spin was worthwhile. The sense of the spin is a crucial factor in understanding the evolution of Triton’s retrograde orbit. In the mid 1970’s Mike Belton and Sethanne Howard (Hayes), who both worked at Kitt Peak National Observatory (in Tucson, Arizona), decided to re-measure the rotational periods of Uranus and Neptune. The Voyager mission was due to encounter Uranus in 1986 so they had to get their data before Voyager got there so their work would help prepare the Voyager mission for the Uranus encounter. They wanted to determine the rotation period, sense of spin, and orientation of the spin axis. They decided to use the spectra method for Uranus, and spectra and photometry methods for Neptune. This is their story of how this was done in the days before the Internet, thumb drives, and laptops. The details of the math are omitted for simplicity. For the spectral work they chose reflected Fraunhofer lines. These spectral lines come from sunlight reflected by the planet’s atmosphere. Of course, planets do not shine on their own. They are seen by reflected sunlight. The visible and near-infrared spectra of Uranus and Neptune have strong Fraunhofer absorption lines making them good candidates for the tilted spectral line approach. No one had ever observed any regular variation in the light of Uranus (Campbell’s work was questioned) so they did not use the photometry method for Uranus. Actually Howard had done a small project in the mid-1970s where several images of Uranus were co-added together to increase the signal-to-noise. The result was a fairly featureless planet with an increase in contrast in one hemisphere. Not believing the results, the project was dropped. That was perhaps unfortunate because Voyager later showed the same feature. For the photometric work, they already knew that Neptune showed variations in infrared light so they chose that spectral waveband for the observations. Gathering the Observations Belton received observing time on the Kitt Peak 4 meter Mayall telescope for this project. It was unusual to get 4 meter time for planetary work. He used Kodak Illa-J plates to record the data/"1 Before there were digital data, there were glass plates with a photographic emulsion Summer 2015 64 embedded on them. The plate was baked in N2 for several hours to increase its sensitivity. After cutting the plate to the proper size (in the dark), one exposed the plate to the object of interest. The developed plate looked like a negative, lighter where the spectral lines appeared (Figure 7). Astronomical spectra typically look like a series of lines, some wide some narrow. In this case, each strip of a line represents a chemical element or molecule in the atmosphere of the planet. Figure 5 shows a standard absorption spectrum spanning blue to red. The lines are not tilted. Figure 6 shows a planetary spectrum with a tilted linelx. 400 500 600 700 I I I I wavtlcnqth in runometcri (10 * m) Figure 5 — standard absorption spectrum Figure 6 — tilted lines. The top is coming towards the observer (blue shift), the bottom is going away from the observer (red shift). The laboratory line (no tilt) is shown in white in front of the tilted line. Belton oriented the slit of the spectrograph so that it spanned the planet from one side to the other. He took a timed exposure. Then he would rotate the spectrograph slit by a few degrees and take another spectrum. He was finished when he had rotated the slit all the way around the planet. He obtained a set of nice spectral data from Uranus and Neptune. Some spectra are shown in Figure 7 which shows three position angles (angle of the slit on the planet) of Uranus and Neptune and the lunar spectrum used to set the plate scale, x dispersion, Xl and intrinsic line tilt. This is a developed plate (a negative of the original) so the lines are not dark, they are light in hue. Why do the spectral lines tilt? Uranus is rotating about its axis. At any given time one side comes toward us, the other away from us. This Washington Academy of Sciences 65 results in a classical Doppler shift of the light we see. Place the spectrograph slit entirely across the planet. Then one end of the slit has light receding from the observer (red shift). The other end of the slit has light coming towards the observer (blue shift). A red shift will shift the position of the spectral line to the right just a bit. A blue shift will shift the position of the spectral line to the left just a bit. The amount of shift varies with the location of the slit on the planet. Near the planet center there will be no shift at all. The farther from the center, the greater the shift, hence a tilt to the whole line as it covers the planet. MOON Figure 7 — spectra of Uranus, Neptune, and the Moon Note that the widths of the spectra are different for the two planets. That is because as seen from Earth Neptune is smaller than Uranus. Figure 8 (upper portion) shows a schematic of a planet with the spectrograph slit across it at approximately a 45° angle (this is called the position angle). It took the powerful 4 meter telescope to do this because the image of the planet had to be large enough to encompass the whole Summer 2015 66 slit. Figure 8 (lower portion) shows how the placement of the slit connects to the spectral line. Belton developed the glass plates and handed them over to Howard for reducing (i.e., obtaining the data). Figure 8 — Image of Uranus with overlaid slit (top image). The parallel black lines represent the slit. Tilted spectral lines (bottom image) are shown with two dotted lines showing where the light from the planet appears on the spectral line. Washington Academy of Sciences 67 The photometry method for Neptune was handled differently from the spectral line method. The Kitt Peak 50" telescope had an infrared filter, so the observer would see infrared light and not much else.x" The light from Neptune passed through the telescope and filter to land on a photomultiplier tube - turning photons into electrons. From there the signal would appear on a Brown chart recorder (an antique observing tool) which fed a continuous strip of paper through - similar to what happens with a lie detector test. The strip of paper recorded the infrared signal from Neptune for as long as one could observe through the night. If the signal never changed as the night wore on, then there was little to see as Neptune rotated. But if the signal dropped occasionally and in a regular manner then the time between drops would give an estimate of the rotation period. They hoped for the best and indeed found this particular signature for Neptune. X1" However, the data were rough and not well defined. Nevertheless, they agreed fairly well with the spectroscopic results. Figure 9 shows a sample of the Neptune data. Time (in terms of fractional periods) is plotted along the horizontal axis, brightness along the vertical axis. Note the dip in the middle of the graph. They were seeing something (unknown) that caused the light from the planet to decrease in a regular way. Figure 9 — infrared photometry of Neptune Summer 2015 68 Data Reduction Howard began reducing the spectral data. She chose to measure O 0 orders 46 (near 5000 A ) and 47 (near 4900 A ). These lines were near the center of the plate, and were of good exposure with no overlapping orders. She identified individual lines using the lunar spectrum as a reference so that she used only Fraunhofer lines. Step one was to “digitize” the data with an automated scanning machine called a microdensitometer. The exposed glass plate was placed on the platen and automatically moved step by step to measure the density of each spot on the plate. She used a 20 x 20pm aperture stepping every 10pm both along and across the spectra. Each order was digitized in overlapping strips 20pm wide and 10,240pm long. In this way each spot on the glass plate was turned into a number stored on a 7-track magnetic tape. The microdensitometer was controlled by a PDP 8 computing machine (way back there in early computers). It took weeks of work just to get the numbers stored on a 7 track tape.xlv Today the data would be taken with a CCD (Charge Coupled Device) chip and recorded digitally right at the start. There were no floppy disks, DVDs, or thumb drives. This was long before the days of the laptop; so she used a Vax minicomputer for the actual data processing. Vaxen were very nice, robust machines.xv Kitt Peak had one of these that everyone shared. A Vax can read a 7-track tape. She wrote several computer programs to read the data from the tape. Each microdensitometer scan was cross-correlated with a running set of weights approximating the function that represented the slit. The position of the line was defined as the zero-crossing of the numerical derivative of the cross-correlation. The tilt angle was found by a least squares fit of a straight line through the zero-crossings. In other words, she digitally reproduced the tilted line. The results for two lines are shown in Figure 10. The left side shows a good fit. The right side shows a poor fit. As a back-up she also made large hard copy prints of the plates (aka Figure 7). She hand measured the tilts with a compass, protractor, and ruler as a check on the automated procedure. Interestingly enough the errors in the hand check were about the same as the errors with the automated procedure. Washington Academy of Sciences 69 Figure 10 — two spectral line fits In the case of Uranus the root mean squared (rms)xvl deviation of the line tilt angles was about ±1.7° and the standard error of the mean approximately ±0.4°. The rms deviation for Neptune was much larger. It was about ±3° with a standard error of the mean about ±0.9°. Knowing the measured tilt angle is not enough. The line tilt must be corrected for the astronomical seeing. This is the largest source of possible error. “Seeing” is a measure of the quality of the sky. Is the image a pinpoint or sharp (good seeing) or is it smeared out (poor seeing)? Celestial objects blur and twinkle because of turbulent mixing in the Earth’s atmosphere. Astronomers always hope for a clear night with good seeing. It can happen that a night can be quite clear yet unusable because of poor quality seeing. The angular size of Uranus as seen from Earth is a known value. One can then estimate how large the observed planet is with respect to that known value. This is a measure of the seeing - in essence, how “fuzzy” is Uranus. The same method is used for Neptune. Of course “fuzzy” can be a bit subjective, hence the possibility of error. In this case the value of the ‘seeing’ is the full width at half intensity of the Gaussian smoothing function required to explain the distribution of density of the plate across the dispersion (i.e., how tall is the spectrum). In other words, use a Gaussian distribution to map the width of the planet (or height of the spectrum). From there one estimates the effective seeing corrections by matching the cross-dispersion distribution of intensity on each plate with Summer 2015 70 the intensity in a model spectrum that results from a convolution of a Gaussian function and a model planetary limb darkeningxv" function at the spectroscopic slit. That is a lot of fancy words that mean match the observed profile of Uranus with the known profile of Uranus. The difference between the two is a measure of the seeing. The estimate of seeing therefore depends on the radius assumed for each planet. Howard and Belton assumed Uranus to have a radius of 25,900 km (actual value 25,362 km) and Neptune to have a radius of 24,500 km (actual value 24,622 km). For Uranus this correction meant the line tilt needed to increase by 32%. For Neptune it was a 202% increase (Neptune was small and fuzzy)! How good was this estimate of the seeing? The uncertainty in the corrected tilt arising from estimating the seeing was about ±8% for Uranus and about ±12% for Neptune. At long last they had the line tilt, 6. Onward to the rotation. The angular velocityxvm, co \ is a measure of the rotation period, co is directly related to the line tilt and is independent of the radius of the planet. There are a number of other things to consider when getting a rotation rate, co \ from the line tilt. Ultimately the relationship between the angular velocity, co. of a planet near opposition and the tilt of the spectral line, tan 0. in reflected light is (see the original paper for derivation of this equation):XIX 206265c £>(/ t)tan(9 (l + cos^sd X where c is the velocity of light, D(k ) the plate dispersion, s the plate scale, cf> the planetary phase angle, d the distance to the planet, i/a \ the position angle of the slit, ^poiethe position angle of the pole, and f0 the latitude of the observer (us) with respect to the planet’s equator. All the parameters on the right hand side are known. Thus, one can solve for the parameters on the left side: position angle of the pole, y/po ie, the rotation period, co, and the spin direction (clockwise or counterclockwise). The position angle is the angle measured in the plane of the sky going counterclockwise from north. It is a standard tool in astronomy, One can see that the actual data reduction is rather complex. After all the work and calculations were done they had their answers. (ycos/’@sin(^pole - y/s) Washington Academy of Sciences 71 They found that the spin of Uranus is retrograde; the spin ot Neptune is prograde. This confirmed early spectroscopic results. For Uranus the direction of its pole points a little south of the orbital plane, thus making Uranus truly a sideways planet. The position angle of the pole, projected onto the plane of the sky, is 283° ±4 ( i.e ., 13° south of the equator). As it revolves, Uranus rotates like a drunken astronomer rolling around the floor. Near the time of Uranian solstices, one pole continuously faces the Sun, and the other one faces away. Only a narrow strip around the equator experiences a rapid day-night cycle with the Sun low over the horizon as in Earth’s Polar Regions. Each pole gets about 42 years of continuous sunlight, followed by 42 years of darkness. Near the time of the equinoxes, the Sun faces the equator of Uranus giving a period of day-night cycles similar to those seen on most of the other planets. For Neptune the pole points north of the orbital plane, agreeing with earlier results. The value for Neptune is 32° ±11. Remember the Earth’s pole is tilted about 23.5°. The rotation periods Howard and Belton found differed significantly from earlier work. For Uranus they obtained: 24 ± 3 hr. and for Neptune: 22 ± 4 hr. As an additional check they were fortunate to obtain from Lowell Observatory the original plates taken in 1912 by Lowell and Slipher for Uranus and Mars. They used the same data reduction method and found significant differences between their Uranus/Mars data and Lowell and Slipher’ s values. They have no explanation for these differences. They also used spectra of Jupiter reduced the same way and found agreement with the known rotation period. Now done, they published the results. The work made a sizable splash in the news, even showing up in the TVew York Times and Popular Mechanics. Naturally one rarely abandons a scientific project. A few years (1980) later they discovered that their estimates of the seeing were in error. The error had little effect on Uranus, but had a much greater effect on Neptune due to its farther distance and smaller angular size.xx Basically they learned that the limb darkening of Neptune may be about the same as Uranus. They had assumed that the two planets had different limb darkening values. Once they changed this parameter the corrected value Summer 2015 72 for Neptune's rotation period became 15.4 ±3 hr. If the limb darkening is less for Neptune, then the rotation period is lengthened. Conclusion What did Voyager find when it got there? For Uranus the rotation period is 17h 14m. They were a bit off there. For Neptune the rotation period is 16h 6m. So, oddly, they were closer for the more distant planet. The axial tilt for Uranus is 97.77° (about 8° south of the orbital plane, they got 13°). The axial tilt for Neptune is 28.32° - not too far from their determination. All in all, a real visit is worth the price of admission. The two Voyager missions are still operating as they move through the heliosheath - the place where the interstellar gas meets the solar wind. They have long since left the Solar System planets behind. Their current locations are continually updated on the Voyager website http://voyager.jpl.nasa.gov/. Check it out. In 2015 Voyager had its 38lh birthday, and it is the longest operating of any NASA satellite. Uranus as seen by Voyager Washington Academy of Sciences 73 Summer 2015 74 I Laplace (1749 - 1827) was a French mathematician and astronomer. II English astronomer/mathematician. III This effect had been suggested by the English astronomer Mary Somerville (1780 - 1872). IV French mathematician. v German astronomer. V1 Galileo had probably observed Neptune, but he thought it was a star. v" Basically, the flatter the planet the faster it rotates. V1U Astronomers no longer use glass plates. Today everything is digital. But in the mid 1970’s glass plates were still common. Ix The spectral lines from galaxies also tilt. One can use the line tilt to determine the mass of the galaxy. x The plate scale can be described as the number of degrees, or arcminutes or arcseconds, corresponding to a number of inches, or centimeters, or millimeters {etc.) at the focal plane (where an image of an object is “seen”). XI The dependence of refraction on the wavelength of light is called dispersion. A lens or prism disperses light. XII Of course, no filter is perfect. They had to correct for leaks in the filter. X1U When Voyager encountered Neptune it saw a “large spot” a storm rather like the Great Red Spot on Jupiter. They must have been observing that spot as it rotated. x,v People do not use 7-track tapes for data storage any more. xv Vaxen are almost gone now too. XV1 . A measure of the error. xvu Limb darkening refers to the diminishing of intensity in the image of a star or planet as one moves from the center of the image to the edge or “limb” of the image. xvni Angular speed at which the planet is rotating. xlx Howard-Hayes, S. and Belton, M., “The Rotational Periods of Uranus and Neptune”, Icarus, 32, 383-401 (1977). xx Belton, M. J. S, Wallace, L., Howard-Hayes, S., and Price, M. J., “Neptune’s Rotation Period: a Correction and a Speculation on the Difference between Photometric and Spectroscopic Results”, Icarus , 42, 71-78 (1980). Bio Sethanne Howard is an astronomer who has held positions with U.S National Observatories, NASA, the National Science Foundation, and the U.S. Navy. She was Chief of the U.S. Nautical Almanac Office, 2000- 2003. Her research specialty is galactic dynamics. She has also been active in science education, especially concentrating on the history of women in science. Washington Academy of Sciences 75 Washington Academy of Sciences 1200 New York Avenue, NW Room 1 1 3 Washington, DC 20005 Membership Application Please fill in the blanks and send your application to the Washington Academy of Sciences at the address above. We will contact you as soon as your application has been reviewed by the Membership Committee. Thank you for your interest in the Washington Academy of Sciences. (Dr. Mrs. Mr. Ms.) Business Address Home Address Email Phone Cell Preferred mailing address Please indicate: Type of membership Business Home Regular Student Schools of Higher Education Attended Degree Dates Present Occupation or Professional Position Please list memberships in scientific societies - and include office held: Summer 2015 76 Journal of the Washington Academy of Sciences Instructions to Authors The Washington Academy of Sciences publishes its interdisciplinary peer-reviewed Journal of the Washington Academy of Sciences four times a year — Spring, Summer, Fall, and Winter. 1 . Deadlines for quarterly submissions are: Spring - February 1 Fall - August 1 Summer - May 1 Winter - November 1 2. Draft manuscripts using a word processing program (such as MSWord), not PDF. 3. Papers should be 6,000 words or fewer. With 7 or more graphics, reduce the number of words by 500 for each graphic. 4. Include an abstract of 1 50-200 words. 5. Graphics must be in black & white or greytone. They must be referenced in the text. 6. Use endnotes, not footnotes. The bibliography may be in a style considered standard for the discipline or professional field represented by the paper. 7. Submit papers as email attachments to the editor. 8. Include the author’s name, affiliation, and contact information — including postal address. Membership in an Academy-affiliated society may also be noted. 9. Manuscripts are peer reviewed and become the property of the Washington Academy of Sciences. 1 0. There are no page charges. Please see the Academy’s web site, www.washacadsci.org, for the library subscription rate, listing of articles dating to 1899, and information on accessing them. Washington Academy of Sciences Washington Academy of Sciences Affiliated Institutions National Institute for Standards & Technology (NIST) Meadowlark Botanical Gardens The John W. Kluge Center of the Library of Congress Potomac Overlook Regional Park Koshland Science Museum American Registry of Pathology Living Oceans Foundation National Rural Electric Cooperative Association (NRECA) Summer 2015 78 Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies Acoustical Society of America, Washington Chapter American Assoc, of Physics Teachers, Chesapeake Section American Astronomical Society American Ceramics Society American Fisheries Society American Institute of Aeronautics and Astronautics American Meteorological Society, Washington, DC Chapter American Nuclear Society, Washington DC Section American Phytopathological Society, Potomac Division American Society for Cybernetics American Society for Metals, Washington Chapter American Society of Civil Engineers, National Capital Section American Society of Mechanical Engineers, Washington Section American Society of Microbiology, Washington Branch American Society of Plant Biologists, Mid-Atlantic Anthropological Society of Washington ASM International Association for Women in Science, DC Metropolitan Chapter Association for Computing Machinery, DC Area Chapter Association for Science, Technology, and Innovation Association of Information Technology Professionals Biological Society of Washington Botanical Society of Washington Capital Area Food Protection Association Chemical Society of Washington District of Columbia Institute of Chemists District of Columbia Psychological Association Eastern Sociological Society Electrochemical Society, National Capital Section Entomological Society of Washington Geological Society of Washington Historical Society of Washington DC Human Factors and Ergonomics Society, Potomac Chapter Institute of Electrical and Electronics Engineers, Northern Virginia Section (continued on next page) Paul Arveson Frank R. Haig, S. J. Sethanne Howard Vacant Lee Benaka David W. Brandt Vacant Charles Martin Vacant Stuart Umpleby Vacant Vacant Daniel J. Vavrick Vacant Mark Holland Vacant Toni Marechaux Jodi Wesemann Alan Ford F. Douglas Witherspoon Chuck Lowe Stephen Gardiner Chris Puttock Keith Lempel Elise Ann Brown Vacant Tony Jimenez Ronald W. Mandersheid Vacant Vacant Jeffrey B. Plescia Jurate Landwehr Vacant Gerald P. Krueger Murty Polavarapu Washington Academy of Sciences Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies (continued from previous page) Institute of Electrical and Electronics Engineers, Washington Section Institute of Food Technologies, Washington DC Section Institute of Industrial Engineers, National Capital Chapter International Association for Dental Research, American Section International Society for the Systems Sciences International Society of Automation, Baltimore Washington Section Instrument Society of America Marine Technology Society Maryland Native Plant Society Mathematical Association of America, Maryland-District of Columbia- Virginia Section Medical Society of the District of Columbia National Capital Area Skeptics National Capital Astronomers National Geographic Society Optical Society of America, National Capital Section Pest Science Society of America Philosophical Society of Washington Society for Experimental Biology and Medicine Society of American Foresters, National Capital Society Society of American Military Engineers, Washington DC Post Society of Manufacturing Engineers, Washington DC Chapter Society of Mining, Metallurgy, and Exploration, Inc., Washington DC Section Soil and Water Conservation Society, National Capital Chapter Technology Transfer Society, Washington Area Chapter Virginia Native Plant Society, Potowmack Chapter Washington DC Chapter of the Institute for Operations Research and the Management Sciences (WINFORMS) Washington Evolutionary Systems Society Washington History of Science Club Washington Paint Technology Group Washington Society of Engineers Washington Society for the History of Medicine Washington Statistical Society World Future Society, National Capital Region Chapter Richard Hill Vacant Neal F. Schmeidler J. Terrell Hoffeld Vacant Vacant Hank Hegner Jake Sobin Vacant D. S. Joseph Vacant Vacant Jay H. Miller Vacant James Cole Vacant Eugenie Mielczarek Vacant Daina Apple Vacant Vacant E. Lee Bray Terrell Erickson Richard Leshuk Vacant Russell Wooten Vacant Albert G. Gluckman Vacant Alvin Reiner Alain Touwaide Michael P. Cohen Jim Honig Washington Academy of Sciences Room 113 1200 New York Ave. NW Washington, DC 20005 NONPROFIT ORG US POSTAGE PAID MERRIFIELDVA 22081 PERMIT# 888 Return Postage Guaranteed Was 03*2- Volume 101 Number 3 Fall 2015 Journal of the WASHINGTON ACADEMY OF SCIENCES Board of Discipline Editors iii Introduction K. Borne iv Big Data Analytics and Workforce Issues C. McNeely. 1 Big Data Adoption in the Health Care Domain E. Kuiler. 1 1 Big Data: Who's Accountable? J. Halm 23 Exploring Bias and Error in Big Data Research K. Seely-Gant, L. Frehill 29 Educating Data Scientists H. Topi, M. L. Markus 39 Everything Old is New Again L. Frehill. 49 Social Media Analysis for Higher Education .A. Berea et ai 63 Privacy in a Networked World H. Xu, H. Jia 73 Membership Application 85 Instruction to Authors 86 Affiliated Institutions 87 Affiliated Societies and Delegates 88 ISSN 0043-0439 Issued Quarterly at Washington DC MCZ LIBRARY HARVARD UNIVERSITY Washington Academy of Sciences Founded in 1898 Board of Managers Elected Officers President Mina Izadjoo President Elect Mike Coble Treasurer Ronald Hietala Secretary John Kaufhold Vice President, Administration Vice President, Membership Sue Cross Vice President, Junior Academy Vice President, Affiliated Societies Gene Williams Members at Large Paul Arveson Michael Cohen Frank Haig, S.J. Neal Schmeidler Mary Snieckus Past President Terrell Erickson Affiliated Society Delegates Shown on back cover Editor of the Journal Sethanne Howard Journal of the Washington Academy of ■ScvewcesOSSN 0043-0439) Published by the Washington Academy of Sciences email: journal@washacadsci.org website: www.washacadsci.org The Journal of the Washington Academy of Sciences The Journal is the official organ of the Academy. It publishes articles on science policy, the history of science, critical reviews, original science research, proceedings of scholarly meetings of its Affiliated Societies, and other items of interest to its members. It is published quarterly. The last issue of the year contains a directory of the current membership of the Academy. Subscription Rates Members, fellows, and life members in good standing receive the Journal free of charge. Subscriptions are available on a calendar year basis, payable in advance. Payment must be made in US currency at the following rates. US and Canada $30.00 Other Countries $35.00 Single Copies (when available) $15.00 Claims for Missing Issues Claims must be received within 65 days of mailing. Claims will not be allowed if non- delivery was the result of failure to notify the Academy of a change of address. Notification of Change of Address Address changes should be sent promptly to the Academy Office. Notification should contain both old and new addresses and zip codes. POSTMASTER: Send address changes to WAS, Rm 1 1 3, 1200 New York Ave. NW Washington, DC 20005 Academy Office Washington Academy of Sciences Room 113 1200 New York Ave. NW Washington, DC 20005 Phone:(202) 326-8975 1 200 New York Ave Suite 11 3 Washington DC 20005 www. washacadsci.org Journal of the WASHINGTON ACADEMY OF SCIENCES Volume 101 Number 3 Fall 2015 Contents Editorial Remarks S. Howard ii Board of Discipline Editors iii Introduction K. Borne iv Big Data Analytics and Workforce Issues C. McNeely 1 Big Data Adoption in the Health Care Domain E. Kuiler 1 1 Big Data: Who’s Accountable? J. Hahn 23 Exploring Bias and Error in Big Data Research K. Seely-Gant, L. Frehill 29 Educating Data Scientists H. Topi, M. L. Markus 39 Everything Old is New Again L. Frehill 49 Social Media Analysis for Higher Education A. Berea et al 63 Privacy in a Networked World H. Xu, H. Jia 73 Membership Application 85 Instructions to Authors 86 Affiliated Institutions 87 Affiliated Societies and Delegates 88 ISSN 0043-0439 Issued Quarterly at Washington DC Fall 2015 11 Editorial Remarks Big Data is a buzzword we hear often these days. Datasets can be so large or so complex that traditional data processing applications are inadequate. We shall peek into a conversation that speaks to big data validity, credibility, applicability, and its broader implications. The Fall issue of the Journal is dedicated to a Symposium entitled “Big Data Analytics and Workforce Issues: Initiatives, Research, and Challenges” which was part of the 7lh annual Dupont Summit held this past December in Washington, DC, and sponsored by the Policy Studies Organization. The objectives of the Symposium included articulating critical research and policy questions on big data and identifying problems that must be faced to answer them. Moreover, the presenters discussed the explosion of big data in and across different contexts (academia, industry, and government - the “triple helix”) and at different levels of analysis. The organizer of the Symposium was Connie McNeely from George Mason University. The moderator was Jong-on Hahm also from George Mason University. Panelists were Philip Bourne, National Institutes of Health; Heng Xu, National Science Foundation; Erik Kuiler, Systems Made Simple, Inc.; Lisa Frehill, Energetics Technology Center; Michelle Schwalbe, National Research Council; and Laurie Schintler, George Mason University. We are fortunate that Connie McNeely gathered papers from the panelists for publication in the Journal. Her paper leads the discussion and introduces the rest of the contributors. Enjoy an unusual and unique look at Big Data. Kirk Borne starts us off describing the concept of Big Data and providing an introduction to the papers. Sethanne Howard Editor Washington Academy of Sciences Journal of the Washington Academy of Sciences Editor Sethanne Howard sethanneh@msn.com Board of Discipline Editors The Journal of the Washington Academy of Sciences has a 12-member Board of Discipline Editors representing many scientific and technical fields. The members of the Board of Discipline Editors are affiliated with a variety of scientific institutions in the Washington area and beyond — government agencies such as the National Institute of Standards and Technology (NIST); universities such as Georgetown; and professional associations such as the Institute of Electrical and Electronics Engineers (IEEE). Anthropology Emanuela Appetiti Astronomy Sethanne Howard Biology/Biophysics Eugenie Mielczarek Botany Mark Holland Chemistry Deana Jaber eappetiti@hotmail.com sethanneh@msn.com mielczar@physics.gmu.edu maholland@salisbury.edu djaber@marymount.edu Environmental Natural Sciences Terrell Erickson Health Robin Stombler History of Medicine Alain Touwaide Operations Research Michael Katehakis Physics Katharine Gebbie Science Education Jim Egenrieder Systems Science Elizabeth Corona terrell.ericksonl@wdc.nsda.gov rstombler@aubuiTistrat.com atouwaide@hotmail.com mnk@rci.rutgers.edu katharine.gebbie@nist.gov i im@deepwater.org elizabethcorona@gmail.com Fall 2015 IV Introduction Big Data Nation - Foundations, Applications, and Implications Kirk Borne Principal Data Scientist, Booz Allen Hamilton Data is the new oil, the new natural resource, the new black, the new bacon, and the new gold rush. Such statements have been made in one form or another, and most have been labeled as “big data” hype. Nevertheless, despite the hype, the growth in data is unmistakably a real (and really big) phenomenon. Fortunately, the growth is not just in the volume of our data collections, but also in the value, opportunities, and insights that organizations can now achieve through the exploration and exploitation of their massive (and growing) data assets. The papers published here cover several dimensions of this data-driven revolution in the business of everything: business, education, research, government, finance, healthcare, natural resources, our personal and social lives, and more. In the paper by Topi and Markus (“Educating Data Scientists in the Broader Implications of their Work”), the authors categorize data science into three bodies of knowledge: Applications, Infrastructure, and Implications. If we re-label the second of these as “Foundations”, then we have not only a useful mnemonic (i.e., the three "-ations"), but we also have a sensible categorization of the papers that are presented here. The foundations upon which we build big data and data science applications for discovery, insights, and innovation include basic research and engineering. That includes academic research as well as data engineering for infrastructure research and development. The paper “Social Media Analysis for Higher Education” by Berea, Rand, Wittmer, and Wall is an excellent example of academic research on mining a particular type of data: social media text data. The authors explore students’ views of their higher education experience through a common social media analytics technique, sentiment analysis. The paper “Big Data: Who’s Accountable?” by Hahm takes us through three case studies where bias and mis- categorization of data have led to inaccurate (and sometimes controversial) results - the importance of starting with a proper foundation in data sampling, data integration, and interpretation of data analytics conclusions is emphasized throughout. A related paper “Exploring Bias and Error in Big Washington Academy of Sciences V Data Research” by Seely-Gant and Frehill examines research ethics (foundations and implications) of sample bias and erroneous interpretations, which are increasingly common in the big data era (especially when working with social media, open source platforms, and online user data). The applications of big data and data science are everywhere, and there are papers here that examine those cases. The paper “Big Data Adoption in the Health Care Domain: Challenges and Perspectives” by Kuiler looks at applications in healthcare (improving patient care and population health), while also addressing foundational issues (workforce development) and implications (data anonymization and data privacy confront data sharing). In the paper “Everything Old is New Again: The Big Data Workforce”, Frehill looks at the abundant, novel, and game-changing applications of big data across all sectors ( e.g ., business, health, and finance). The author then explores how this changes workforce development, education programs, consumer/customer experience, and the landscape of data-driven decision-making in organizations. Finally, the implications of what we are doing with our data collections deserve special attention, both here as well as in data science and analytics education programs. One of those areas of focus in academic programs should be data ethics. There are very few examples of ethics courses that specifically address the big data era - one of those is the “Data Ethics in an Information Society” course in the George Mason University Computational and Data Sciences degree program. Several papers here address such societal and workforce implications. The aforementioned paper by Topi and Markus specifically examines the legal, ethical, and societal implications of analytics and data science, specifically in the context of training data scientists and analysts to be aware of and diligent in minimizing the possible harmful consequences of their analytics applications. The paper “Big Data Analytics and Workforce Issues: Prospects and Challenges in the Information Society” by McNeely examines both applications and implications of big data analytics, with a central focus on the latter, specifically: challenges across technical, social, political, and economic dimensions. In the insightful paper “Privacy in a Networked World: New Challenges and Opportunities for Privacy Research” by Xu and Jia, the authors investigate new concepts, consequences, and concerns related to privacy in our increasingly digital Fall 2015 VI lives. They examine human-data interaction, information linkability, information ephemerality, and information identifiability. All of this naturally leads to concerns about information liability, something which all data analytics professionals must weigh on the balance sheet of information and data assets. As data emerges from the era of big data hype, contributing value beyond our large data repositories, and blooms into a major organizational asset and a major organizational product, the papers presented here will provide valuable guidance, insights, and perspectives concerning the foundations, applications, and implications of data analytics in a data- drenched world. Bio Kirk Borne, PhD is an astrophysicist, Big Data science consultant, public speaker, and the Principal Data Scientist in the Strategic Innovation Group at Booz Allen Hamilton. He previously spent 12 years as tenured Professor at George Mason University in the Computational Science, Informatics, and Data Science programs. Before that, he worked 18 years on various NASA contracts, as research scientist, as a manager on a large science data systems contract, and as the Hubble Telescope Data Archive Project Scientist. He also actively promotes data literacy by disseminating information related to data science and analytics on social media, where he has been named consistently since 2013 among the top worldwide influences in big data and data science. Washington Academy of Sciences Big Data Analytics and Workforce Issues: Prospects and Challenges in the Information Society Connie L. McNeely George Mason University Abstract Big data is one of the most critical features marking and defining our world today. It constitutes an analytical space encompassing processes and technologies that can be applied across a wide range of domains in the current and growing information society. The articles presented in this issue address related challenges and prospects as crucial considerations in technical, social, political, and economic power and relations in national and international contexts. With particular attention to conceptual delineations, analytical applications, and educational and workforce dynamics, they attend to both instrumental and intrinsic aspects of big data relative to society in general. Together, the articles constitute a conversation that speaks to big data validity, credibility, applicability, and broader societal implications — both positive and negative — today and in the future. Introduction Big data, in all of its manifestations and applications, is the beating heart of today’s burgeoning information society. On the one hand, big data has been acclaimed in line with promises for societal benefits. However, on the other hand, big data also has sparked controversies and debates on the challenges and vulnerabilities that it has created relative to social, political, and economic power and relations. Whether addressed in terms of technical, social, or organizational perspectives, relevant topics are in the forefront of initiatives in academia, government, and industry. Moreover, the advent of big data has raised questions about those who use it and those who work with it, especially in light of socio-cultural and structural dynamics and disparities in terms of educational and workforce dynamics. Accordingly, the objectives of the articles in this collection include articulating critical research and policy questions and identifying challenges to answering them and to engaging big data effectively. Employing perspectives that address both instrumental and intrinsic aspects of big data, they offer an effective and comprehensive view on the prospects and challenges of big data analytics and workforce issues in and across different contexts and at different levels of analysis. Together, the articles constitute a conversation Fall 2015 2 that speaks to big data validity, credibility, applicability, and broader societal implications now and in the future. New opportunities and prospects, but also new challenges, controversies, and vulnerabilities, have marked the explosion of big data as a phenomenon in and of itself. In 2000, only a quarter of all stored information was digital; by 2013, more than 98 percent of the world’s stored information was digital (Mayer-Schonberger and Cukier 2013). Indeed, “the world contains an unimaginably vast amount of digital information which is getting ever vaster ever more rapidly. This makes it possible to do many things that previously could not be done: spot business trends, prevent diseases, combat crime, and so on. Managed well, the data can be used to unlock new sources of economic value, provide fresh insights into science, and hold governments to account” (Economist 2010) — but they also create a host of new problems, with misuse and misinformation, security concerns, privacy violations, etc. at the top of many related policy agendas. The ever- increasing body of data is a core operational feature in virtually every sector of society, and how we understand and use big data is increasingly the defining feature of our times. Conceptual Dimensions Big data is a multidimensional concept referring to the exponential growth and availability of both structured and unstructured data (SAS 2013), embracing technology, decision making, and policy. Big data has largely been interpreted in terms of the “3 Vs”: volume, velocity, and variety. That is, "big data is high volume, high velocity, and/or high variety information assets that require new forms of processing to enable enhanced decision making, insight discovery, and process optimization" (Beyer and Laney 2012; Laney 2001). Volume indicates the increasing amount of data, velocity indicates the speed of data, especially the rate at which it is created or becomes available, and variety indicates the range of data types and sources (Laney 2001). The compilation of large complex datasets has made for massive volumes of data characterized by variety that reflect the different types of structured and unstructured data that are collected; velocity refers to how quickly these data can be made available for analysis (UA 2015). Together, these dimensions comprise a basic model for describing big data. Washington Academy of Sciences 3 However, other “Vs” also have been included, especially variability and veracity , such that reference to the “5 Vs” has become common. Along with the variety and complexity that mark big data, variability is reflected in inconsistencies in data flows (SAS 2013). The veracity of the data represents an especially critical issue. Veracity is an indication of data integrity and the extent to which it can be trusted for analytical and decision- making purposes (UA 2015). Methods for data verification and validation, as specifically applied to big data, are of particular importance in this regard. In addition, another “V” — value — is sometimes discussed as a separate dimension of big data, highlighting the value-added capacity of big data (IDC 2012). In any case, while there is a lack of consistent definition, the term “big data” has reached some general agreement among various stakeholders as constituting at least some indication of volume, signaling the size of datasets as the critical factor (Ward and Barker 2013). After all, the allusion is to “big” data in relative terms. Big data is derived from various sources, in particular streaming data as the Internet of Things, social media data, and publicly available open data. The conversion of large collections of documents from print to digital format is giving rise to massive archives of unstructured data, and social media, crowdsourcing platforms, and various applications are producing reams of information from the real-time transactions of people around the world. The complex structure, behavior, and permutations of datasets are a fundamental consideration in describing data as big (Ward and Barker 2013). However, having said that, “big data is less about data that is big than it is about a capacity to search, aggregate, and cross-reference large data sets” (boyd and Crawford 2012, p. 663). Underlying the concept of big data are the technologies — the tools and techniques — that are used to process massive or complex datasets. Hence, we can refer to big data as a term describing the storage and analysis of large and/or complex datasets using a series of applicable techniques (Ward and Barker 2013). From an expanded theoretical and practical perspective, big data also has been described as a cultural, technological, and scholarly phenomenon, resting on the interplay of technology, analysis, and mythology (boyd and Crawford, p. 663): Fall 2015 4 1) Technology : maximizing computation power and algorithmic accuracy to gather, analyze, link, and compare large data sets. 2) Analysis : drawing on large data sets to identify patterns to make economic, social, technical, and legal claims. 3) Mythology : the widespread belief that large data sets offer a higher form of intelligence and knowledge that can generate insights that were previously impossible, with the aura of truth, objectivity, and accuracy. In more encapsulated terms, big data reflects “a point of view, or philosophy, about how decisions will be — and perhaps should be — made in the future” (Lohr 2013). Data-to-Knowledge-to-Action Analytics “The challenge of big data is to convert it into useable information by identifying patterns and deviations from those patterns” (UA 2015). In epistemological terms, information is comprised by a collection of data, and knowledge is established through different strands of information (. Economist 2010), leading to questions that speak to the process of converting data to knowledge to action. For example, what are analytical and policy implications of the data in light of the how and why they are collected, categorized, and aggregated? Do such data tasks reflect on how they are or should be used? Furthermore, as elsewhere queried by McNeely and Hahm (2014), will the analysis of big data provide insights and information that will allow the development of answers to big questions, or will it simply provide larger scale versions of answers already attained with smaller data? Frankly, actual understanding is not stressed in most big data approaches; correlations are the rule, representing a move away from actually understanding phenomena to simply indicating associations (Mayer-Schonberger and Cukier 2013). The challenge of turning data into knowledge reflects matters of data interpretation and re-purposing relative to secondary data markets (Washington 2014). In practice big data might yield information, but not necessarily understanding. Keep in mind that data gain meaning only in context. What critical or fundamental factors must be considered for true understanding? The socio-technical limitations of big data rest on considerations of context and meaning and, as such, big data must be engaged with an appreciation of both its power and its limitations. More to Washington Academy of Sciences 5 the point, while big data is increasing, the ability to translate it into knowledge and, more, to extract wisdom from it is relatively rare (McNeely and Hahm 2014, p. 307; Economist 2010). Large datasets have long been around and in use in various fields. However, the big data revolution invokes a different frame for engaging them. The integration of data from various sources and the use of that data for purposes beyond those for which it was originally collected or created are principal tasks associated with big data use (Berman 2013). Moreover, at this point, it appears certain that data will continue to get “bigger and bigger.” The Internet of Things is expected to comprise tens of billions of objects by the end of this decade and is actively and instantaneously sensing data on virtually every aspect of our lives and environment. Noting this trend, Kuiler (p. 11) looks to the volumes of clinical, financial, and consumer information available to healthcare organizations. Mapping a complex multi-disciplinary approach to big data analytics, he focuses on questions related to health and bioinformatics. In application, he categorizes and reviews a wide range of structured and unstructured data and offers an imiovative approach to performance measurement in the healthcare domain. Overall, he provides evidence on the use of big data analytics for reducing operational costs and optimizing performance, for improving regulatory compliance, and for increasing returns on investments, while also delineating future trends in big data analytics. However, he also explores challenges and barriers to big data analytics and use, discussing limitations and difficulties incurred in, for example, industry refusals to share data, institutional barriers, and information governance. Framing big data as a trope for a number of different technological and institutional factors, he points to the problems of an abundance of data relative to a scarcity of information, noting that more data is not always better. Approaching such issues from a different direction, practical questions of data veracity also are fundamental for converting data to knowledge to action. Problems of sampling bias are particularly relevant in this regard. Sampling bias is inherent in many big datasets. How might that affect policy development and implementation? Seely-Gant and Frehill (p. 29) examine related complications along these lines, discussing how sampling issues, especially selection bias, associated with big data sources can have far reaching implications for analysis and interpretation. Furthermore, in the Fall 2015 6 same vein, Hahm (p. 23) offers a commentary on accountability and data veracity, pointing out how sampling and sorting bias and errant categorizations can lead to inaccurate conclusions, which can be particularly dangerous for informing policy decisions. While also encompassing these problems, one of the most prominent and controversial issues that arises in discussions of big data is privacy. Xu and Jia (p. 73) probe this topic, examining changing conceptions of privacy in today’s big data environment in terms of information identifiability, ephemerality, and linkability. They apply this conceptual approach to investigate threats to information privacy in light of the collection and analysis of large-scale data from social networking sites. Focusing on human-data interaction, they turn to problems and risks associated with, for example, data de-identification and re-identification, data integration, and legal obligations and developments with regard to privacy issues. Their primary emphasis is on mapping privacy regulations into actionable information technology requirements that are re-usable across systems. Education and Workforce Dynamics Big data engagement and related topics are relevant within and across sectors and require examination from technical, social, and organizational perspectives. The skills, training, and education necessary for big data related jobs in industry, government, and academia have become a focus of discussions on educational attainment relative to workforce trajectories. Especially given assertions of a skills gap for manipulating, analyzing, and understanding big data, the relationship between education and the development of the big data workforce is a critical point of departure for delineating the field in general. Further, the role of big data in affecting social, political, and economic relations and power come into play as reflected in questions of educational and workforce opportunity and access, and also raising questions of the “digital divide.” Do gatekeepers come into play with big data, as in other fields, precluding certain individuals or groups from accessing data or participating in relevant fields? In general, basic questions on building proficiency in big data and workforce development are at the forefront of debates in different sectors (NRC 2014). That is, what should be taught, by whom, to whom, and how? Washington Academy of Sciences 7 Technically speaking, training and education for big data jobs typically require a basic knowledge of statistics, quantitative methods, or programming, upon which applicable skillsets can be built. Such background can be acquired in a number of fields that have long incorporated related preparation. For example, Frehill (p. 49) notes that “social scientists have worked with exceptionally large data sets for quite some time, historically accessing remote space, writing code, analyzing data, and then telling stories about human social behavior from these complex sources.” However, she differentiates between traditional large “designed” datasets and the new “organic” big data that are calling for more and more trained knowledge workers with the required “deep skills and talent.” Examining the role between higher education and the development of the big data workforce, she addresses basic questions about participation and also considers key lessons regarding gender differences in the big data workforce. Overarching changes in occupational roles and practices in the face of technological shifts have led to revised workforce expectations and needs. Some estimates suggest a shortage by 2018 of some 190,000 data scientists in the United States, in addition to 1.5 million analysts and managers with knowledge and skills to use analyses of big data to make effective decisions (GovLab 2013; MGI 2011). Frankly, when an industry or field is growing rapidly, “it is not unusual for a shortage of workers to occur until educational institutions and training organizations build the capacity to teach more individuals, and more people are attracted to the needed occupations” (CEA 2014, p. 41). Thus, Topi and Markus (p. 39) investigate the growing number of analytics and data science programs, arguing for the need to include an emphasis on the implications and consequences of practices and applications in related fields. They note the need for big data workers “who are sensitive to data downsides as well as upsides” to achieve the benefits of big data while avoiding harmful consequences. From yet a different perspective, the investigation presented by Berea, Rand, Wittmer, and Wall (p. 63) rests on social media analysis, using big data itself in their research on big data analytics within education and related policies and reflecting the changing data landscape. Fall 2015 8 Conclusion The effects of big data are being felt everywhere. As an analytical space, big data encompasses processes and technologies that can be applied across a wide range of domains “from business to science, from government to the arts” ( Economist 2010), with positive and negative implications depending on perspective and application. As such, Big Data triggers both utopian and dystopian rhetoric. On one hand, Big Data is seen as a powerful tool to address various societal ills, offering the potential of new insights into areas as diverse as cancer research, terrorism, and climate change. On the other, Big Data is seen as a troubling manifestation of Big Brother, enabling invasions of privacy, decreased civil freedoms, and increased state and corporate control. As with all socio-technical phenomena, the currents of hope and fear often obscure the more nuanced and subtle shifts that are underway, (boyd and Crawford 2012, pp. 663-664) Big data raises new issues and concerns related to, for example, privacy, liability, security, and access, and has been invoked relative to new ways of thinking about the world and relations across contexts. It has led to new possibilities and prospects for research and policy, with fundamental issues turning on cultural, organizational, and technological capacities at the heart of debates and practices within and across academia, industry, and government. Attending to issues of research and knowledge production, of education and workforce dynamics, of socio-cultural, political, and economic relations, the articles presented in this issue interrogate and examine critical related issues from various perspectives, addressing challenges and prospects for big data in theory and application in the growing information society. Washington Academy of Sciences 9 References Berman, J.J. 2013. Principles of Big Data: Preparing, Sharing, and Analyzing Complex Information. Burlington, MA: Morgan Kaufman. Beyer, M.A., and D. Laney. 2012. “The Importance of ‘Big Data’: A Definition.” Stamford, CT: Gartner. boyd, D., and K. Crawford. 2012. “Critical Questions for Big Data: Provocations for a Cultural, Technological, and Scholarly Phenomenon.” Information, Communication, and Society 15(5): 662-679. Chmura Economics and Analytics (CEA). 2014. “Big Data and Analytics in Northern Virginia and the Potomac Region.” Northern Virginia Technology Council. [https://gwtoday.gwu.edu/sites/gwtoday.gwu.edu/files/downloads/BigData%20repo rt%202014%20for%20Web.pdf (accessed 3 September 2015)] Economist. 2010 (27 February). “Data Data Everywhere: A Special Report on Managing Information.” [https://www.emc.com/collateral/analyst-reports/ar-the-economist- data-data-everywhere.pdf (accessed 3 September 2015)] Eliot, T.S. 1934. The Rock: A Pageant Play. New York: Harcourt, Brace, and Company. GovLab. 2013. “The GovLab Index: The Data Universe.” [http://thegovlab.org/govlab- index-the-digital-universe (accessed 3 September 2015)] IDC. 2012. “Worldwide Big Data Technology and Services, 2012-2015 Forecast.” IDC Market Analysis #233485. Laney, D. 2001. “3D Data Management: Controlling Data Volume, Velocity, and Variety.” [http://blogs.gartner.com/doug-laney/files/2012/01/ad949-3D-Data- Management-Controlling-Data-Volume-Velocity-and-Variety.pdf (accessed 3 September 2015)] Lohr, S. 2013 (19 June). “Sizing Up Big Data, Broadening Beyond the Internet.” New York Times, [http://bits.blogs.nytimes.com/2013/06/19/sizing-up-big-data- broadening-beyond-the-intemet/?_r=0 (accessed 3 September 2015)] Mayer-Schonberger, V., and K. Cukier. 2013. Big Data: A Revolution That Will Transform How We Live, Work, and Think. New York: Houghton Mifflin Harcourt. McKinsey Global Institute (MGI). 2011. Big Data: The Next Frontier for Innovation, Competition, and Productivity.” www.mckinsey.com/mgi. (Accessed 3 September, 2015) McNeely, C.L., and J. Hahm. 2014. “The Big (Data) Bang: Policy, Prospects, and Challenges.” Review of Policy Research 3 1 (4): 304-3 1 0. National Research Council (NRC), U.S. 2014. Training Students to Extract Value from Big Data. Washington, DC: National Academies Press. Fall 2015 10 SAS. 2013. “What is Big Data?” [http://www.sas.com/en_be/insights/big-data/what-is- big-data.html (accessed 3 September 2015)] University Alliance (UA). 2015. “What is Big Data?” [http://www.villanovau.eom/resources/bi/what-is-big-data/#.VjBFMSuK820 (accessed 3 September 2015)] Ward, J.S., and A. Barker. 2013. “Undefined by Data: A Survey of Big Data Definitions.” [http://arxiv.org/pdf/1309.5821vl.pdf (accessed 3 September 2015)] Washington, A. 2014. “Government Information Policy in the Era of Big Data.” Review of Policy Research 3 1(4): 319-325. BIO Connie L. McNeely is a Professor in the School of Policy, Government, and International Affairs at George Mason University, where she is also the Co-Director of the Center for Science and Technology Policy. Her teaching and research address various aspects of science, technology, and innovation, healthcare, organizational behavior, public policy, governance, social theory, and culture. Dr. McNeely directs major projects on big data analytics, on scientific networks, and on migration and diversity in the science and technology workforce, and leads an International Research Group on Global Innovation in Science and Technology. Washington Academy of Sciences Big Data Adoption in the Health Care Domain: Challenges and Perspectives Erik W. Kuiler George Mason University Abstract Due to recent technological advancements, health care organizations now have access to large volumes of clinical, financial, and consumer information from which to identify patterns and trends. As with other industries, health care is grappling with the best ways to decipher and leverage these big data sets, with the ultimate goals to enhance patient care and improve population health. The sheer magnitude of the number of available data is both a boon and a hurdle. When interpreting data, more information is not always better, unless an organization assesses these data to discern what are noise and what are not. This paper explores a number of challenges and barriers to big data analytics and use. Introduction The healthcare domain has witnessed a rapid growth in the delivery of data-driven medicine resulting from the introduction of, for example, electronic health records, digital imaging, digitized procedures, increasing sophistication in lab test formulation, the real-time availability of sensor data, and, what stands out in the popular press, the introduction of genomics-related projects (Ohno-Machado 2012; Shah and Tenenbaum 2012). Information technology (IT) advances have led to a discourse on the applicability of big data (a term coined by the Gartner Group, an IT industry market research organization) to health data analytics (for example, Sahoo et al. 2013). This study summarizes a presentation made at the 2014 Dupont Summit, held in Washington DC, and explores topics considered in that discussion. The paper concludes with a preliminary assessment of future trends. Adopting the Gartner Group’s definition, trade journals tend to emphasize three big data properties, collectively referenced as the three V’s: volume - to denote an exponentially large data set, ranging in size from one or more terabytes ( 1 0 1 2) to multiple petabytes ( 1 0 1 5) or exabytes ( 1 0 1 8); velocity - to indicate data that arrive as continuous streams, rather than as transaction or database files; and variety- to designate data sets that contain both structured and unstructured data that may be subject to different Fall 2015 12 semantics and in different formats, gathered from diverse sources.1 Discussing big data in its historical perspective, Jacobs (2009) offers a definition of the term that is perhaps more useful because it places big data in its proper IT context: “Big data should be defined at any point in time as "data whose size forces us to look beyond tried-and-true methods [of storage and manipulation] that are prevalent at that time.,,, From Jacobs’ point of view, in the 1960’s data files that could not be managed effectively with a single tape mount could be considered as the big data of that era. Currently, the capabilities to ingest, analyse, and manage multi-petabyte data sets have underscored the limitations of our data analytics capabilities supported by Relational DataBase Management Systems. These data management limitations have led to the introduction of specific IT applications that address the volume and velocity requirements of big data, much as in the 1960-80’s the availability of multi-tape data sets led to the introduction of mechanical “tape monkeys” ( Jacobs’ term) to swap tapes in and out. Because of its use in the popular press, the term big data has become a trope for a number of different technologies and institutional conflicts between the rights of the states and the rights of the citizenry: cloud-based data and information analytics and big data management systems, data interoperability as well as NS A spying, insurance denials based on big data- based trend analyses, and security lapses that may lead to data breeches and the loss of personally identifiable information (PII) - any data that, collectively or severally, may potentially identify a specific individual human being. The scope of the healthcare domain is extensive, comprising the activities of diverse epistemic communities, each of which has its own institutional paradigms and cultural imperatives, resulting in a contested equilibrium (adapting Amartya Sen’s phrase 1982, 1999) between different interest groups: clinical health, focused on the delivery of patient-centered healthcare services; public health, including clinical case surveillance, syndromic surveillance, prevention, preparedness, and health promotion in a community; population health, focused on health outcomes of a group of individuals, including the distribution of such outcomes, in a population; environmental health, focused on physical, chemical, and biological factors external to a person, and all the related factors that may have an impact on Washington Academy of Sciences 13 individual behavior; and, since the 1990’s, genomics - genes, genomes, proteins, cells, ecological systems. Impetus to Big Data Adoption in the US In the United States, the impetus for big data-based health informatics came during 2008-2010. Under the Health Information Technology (HIT) for Economic and Clinical Health (HITECH) component of the American Recovery and Reinvestment Act of 2009, the Centers for Medicare and Medicaid reimburse health service providers for using electronic documents in formats certified to comply with HITECH’s Meaningful Use (MU) standards. The Patient Protection and Affordable Care Act of 2010 (ACA) promotes access to health care and greater use of electronically transmitted documentation. Health informatics are expected to provide a framework for the electronic exchange of health information that complies with all legal requirements and standards and, consequently, expands the delivery of comparative effective- and evidence-based medicine. HITECH MU and ACA support the adoption of Electronic Health Records (EHR) as the preferred method for data interoperability among patients, healthcare providers, and healthcare payers (HHS 2015). Benefits of Big Data Adoption Today, the availability of large data sets is the norm rather than the exception. With the adoption of data analytics, end-users have become increasingly more data literate, so that, while a simple spreadsheet would have sufficed earlier, now end-users expect to see more complex models, such as the results of time-series probability-based analyses to complement snap-shot descriptive statistics. The adoption of big data acquisition, management, and analytics provides a number of important benefits: large sample size - the larger the size, the greater the probability that the sample will accurately reflect the characteristics of the population; increased predictive power - studies based on big data samples are more likely to give statistically significant results; a strong foundation for puiposeful action - cluster and category analytics of very large data samples support the development of treatments and protocols that are more accurately tailored to the specific needs of patient populations (or cohorts). Fall 2015 14 In addition big data analytics support proactive wellness and disease management by discovering patterns; for example, snap-shots of current operations, likely future trends, metrics of program efficacy and efficiency, prospective needs of a population; decision support data to chart future directions, data to support knowledge-intensive problem definition and resolution (diagnostics, research, policy analysis, etc.). Also, big data analytics enable healthcare improvements by, for example, integrating clinical and claims data so that they are accessible, searchable, and reportable; aggregating data from patient encounters to support public and population health management; identifying and targeting individual patients and cohorts for outreach; assessing quality of care across provider networks; and correlating clinical and financial risk measures to optimize health care delivery. Additionally, big data provide answers to important questions, such as: How effective is a particular program, in terms of access and results; is a client population served as well as it could be? Looking at specific parameters, what policy changes should be enacted to make the program better? Should more resources be allocated and of what kind, when, and where? What is the likelihood of a patient suffering a stroke, given his or her lifestyle, and, based on these probabilities, what kind of ameliorative regimen should be proposed? What is the likelihood of a provider committing fraud, given certain characteristics, and, as a corollary, given the historical pattern of this provider’s behavior compared to that of other providers, is this one committing fraud? How can a fraud detection program be improved by operationalizing the analytics model? Barriers to Big Data Adoption While big data analytics can improve the delivery and quality of healthcare, there are institutional barriers to their adoption, including adversarial relationships between healthcare practitioners and HIT vendors, lack of government incentives, economic limitations, and ethical and moral constraints centered on data ownership, stewardship, and human rights. Information Governance and Management Challenges Prior to HITECH, health data sharing was usually limited to patient- physician communications, and data interoperability between healthcare providers was limited to facsimile distribution and similar dissemination methods. The Internet changed all this, and the frequently cloud-based Washington Academy of Sciences 15 aggregation of EHR data in very large collections (petabyte data sets, for example), sustained by big data management and analytics, offers opportunities to understand diagnoses, treatments, and protocols on a large scale, providing an important complement to clinical trials. Nevertheless, big data’s promise of increased data interoperability and information sharing has exacerbated issues of syntactic conformity and semantic clarity that have plagued data analytics since their inception in 1960’s automated data processing environments. These issues require more than technological solutions because the issues have their provenance in the cultural and institutional determinants of the epistemic domains to which they apply, rather than in the IT systems that support the analytics of such determinants. Current HIT capabilities support the integration of data from diverse sources that are frequently managed as data silos - for example, patient clinical data, adverse event data, product data (drugs, medical devices, blood, consumer, etc.), environmental and toxicological data, genomic data industry-provided data (insurance, product, etc.) - without considerations of data interoperability. Internally as well as externally, epistemic communities in the health domain support different lexica and ontologies, thereby restricting the possibility of efficient information sharing. For example, in the clinical community, the International Statistical Classification of Diseases and Related Health Problems Version 9 and 10 (ICD-9 and 10), which, although managed by the same agency, are not fully compatible. Furthermore, these two standards are not fully compatible with the Systematized Nomenclature of Medicine-Clinical Terms, another frequently used standard, and require resource-intensive “cross-walks.” In the genomics community, researchers employ at least two standards, depending on where they operate in the world community: the GenBank file format or the Swiss-Prott format. In the toxicology domain, the US National Institute of Environmental Health Services’ participation in the toxicogenomics ontology and global database initiatives is critically important in establishing a common lexicon and ontology. There are also different conveyance and transportation frameworks for transporting data: the HL7 Version 2.x (V2) messaging standard is, arguably, the most widely implemented standard for health data information exchange in the world. However, this standard is not compatible with the Fast Health Interoperable Resources Health Information Exchange and Fall 2015 16 Clinical Document Architecture (CDA) specifications maintained by the same organization. Moreover, neither the ICD-9/10 nor the HL7 standards are compatible with the American National Standard X12 Electronic Data Exchange transactions (the 274-278 and 834-835 series of transactions). There are also different communication architectures in use: point-to-point (peer-to-peer), and central repository (push/pull), etc. Divergent Views of Product Requirements Clinicians want HIT products that are tailored to support their specific processes and protocols (based on my conversations with practitioners and vendors at the 2014 AMI A national conference). Vendors want to capture the largest market share possible at the lowest cost; hence, the impetus to develop a generic, one-size-fits-all solution as the most efficient model. There are also industry barriers to health data interoperability. Many EHR vendors treat healthcare data as proprietary assets that can offer considerable market advantages. Also, many lifestyle- focused vendors (for example, in the tobacco, soft drink, and fast food industries) resist health research and, consequently, data sharing. Lack of Funding The Federal government has offered programmatic incentives to enhance healthcare delivery but these are frequently insufficient and not sustained. For example, a number of programs, such as the Beacon Community Cooperative Agreement Program, have come to an end. The purpose of this program was to demonstrate how health IT investments and the use of EHR’s could advance the vision of patient-centered care, while achieving better health and better care at lower cost. The Health and Human Services (HHS) Office of the National Coordinator for Health IT provided $250 million over three years to 17 selected communities, each with its unique population and regional context, throughout the United States that had already made inroads in the development of secure, private, and accurate systems of EHR adoption and health information exchange. When the funding dried up, a number of Beacon Communities incorporated elements of the program into their organizational structures and formed consortia at their own costs but it is not likely, in the long run, that these efforts can be sustained. Washington Academy of Sciences 17 In 2014, 87% of US hospitals had some form of EHR system (Cohen et al. 2014). In the US medical community many large institutions have adopted the use of EHRs, data sharing, and big data analytics; however, many small practices do not have the resources to adopt EHRs because they are expensive and there are few incentives to support their adoption. Local communities and regional governments usually do not have the resources to assist medical practices with adopting EHR and big data analytics. On the international level, many Southern Cone countries do not have the resources to provide basic healthcare to their citizens, let alone support an HIT infrastructure required to support EHR-based medicine, big data management, and analytics (to which UN’s efforts to reach its Millennium Goals can attest; see UN 2015). Likewise, International Non- Governmental Organizations that have limited financial, organizational, and temporal resources must frequently operate in adversarial environments created by host governments. Data Ownership and Data Stewardship There are also institutional barriers to health big data analytics that focus on data ownership and stewardship. Among the benefits of introducing EHR’s and Personal Health Records (PHR) is to institutionalize patient-focused healthcare so that patients become active partners in their healthcare paradigms (for example, to mitigate patients’ strategic ignorance: “my doctor knows what is best for me, and I expect her to notify me when things may go wrong.”).2 Patients own their data; the medical establishment and the government are data stewards. This uneasy alliance raises questions of when, and under what circumstances these personal data may be shared and how personal identity data can be protected against theft and unauthorized access. To ensure the privacy of individually identifiable health information in accordance with the Health Insurance Portability and Accountability Act of 1996 (HIPAA), health data records must be “anonymized” by removing all Personally Identifiable Information (PII) from such records prior to their use in data analytics. Data anonymization techniques are not fool-proof. A recent study noted that, in the absence of PII, it is still possible to join records with a reasonable degree of accuracy (for advertising purposes) from two discrete data sources based on date of birth, 5-digit residential zip code, and gender (cited, among others, by Fall 2015 18 Cavoukian and El Eman 2011; see also Kum et al. 2013 for an approach to preserving privacy in interactive record linkages). Legal , Moral, and Ethical Considerations In the health domain big data analytics exacerbates the antinomy between healthcare as a human right and healthcare as a commodity. The UN Human Rights Charter and the Convention on the Elimination of All Forms of Discrimination against Women formulate a concept of human rights that includes rights essential to human development, such as rights to adequate housing, healthcare, education, economic development (for example, employment at a fair wage), that apply to all humanity, regardless of gender, age, race, ethnicity, sexual orientation, etc. (UN 2002). Nussbaum (2000; see also Sen 1982) observes that bodily health is second only to life, supported by bodily integrity, as essential capabilities necessary to flourish as individual human beings. Sen (1999) posits five types of instrumental freedoms as essential to human freedom in the polity: political freedoms, including free speech and free elections, to help promote economic security; economic facilities, in the form of opportunities to engage in market activities and production; social opportunities, among them access to education and health care; transparency guarantees, those mechanisms and institutions necessary to guarantee full disclosure of information - the basis for trust; and protective security, those institutions necessary to prevent any human from sinking into destitution and abject poverty. Although there are moral strictures against using big data analytics to restrict insurance coverage to individuals, such practices occur. Similarly, in a “market model” of healthcare access and delivery, there are very few means, other than moral approbation, to restrain a pharmacological drug company from raising the price of a drug by, for example, 2000% or 5,500% (CBS News September 22, 2015; NBC News September 22, 2015). Big data analytics also faces other normative barriers. For example, patients are likely to accept the necessity of data sharing among providers to improve the quality of care, but the notion that their data will be shared with other non-provider third parties has proven to be controversial, especially when there are high-profile cases when data are shared without the owner’s consent. The case of Hilda Lacks comes to mind. She was a young African American woman who, in 1951, died of cervical cancer. Doctors took samples of her cells without her knowledge and shared them Washington Academy of Sciences 19 with other clinicians and researchers. Although labs were selling samples of what came to be known as the HeLa cells, Lacks’s family received no portion of the money generated by those sales and were not informed how these cells were used (Falik 2014). Misuse of big data analytics and their enabling technologies have fostered an increasingly greater wariness of citizens of their government. The majority of citizens understand that their data need to be shared to support the common weal but, as the activities of the National Security Agency’s spying on the U.S. population indicate, the citizenry is justified in its suspicions of its government. Big Data Analytics Workforce Challenges To be effective, big data analytics require a non-insular, non- compartmentalized ontological perspective and a multi-disciplinary, holistic approach to knowledge acquisition that incorporates skills from a variety of academic disciplines, including, for example, quantitative analysis (statisticians, computer scientists), finance (financial analysts, cost analysts, fraud analysts), healthcare (medical practitioners, biologists, chemists, product engineers), infrastructure and device engineering (communications and device engineers), social sciences (sociologists, medical healthcare economists), governance (policy analysts), information management (information governance experts and managers, librarians), deontology (ethicists), jurisprudence (legal professionals). The majority of universities and training institutes do not offer cross-field programs that emphasize the integration of these skills. As a result, one of the roles frequently overlooked in efforts to minimize the risks of the misuse of big data analytics is the role of the ethicist. If a little data analytics can lead to misuse, big data analytics can lead to even greater misuse because so many more data are available for abusive practices. Future Trends The study indicates a number of trends. Cloud-based big data analytics and usage will grow. Big data analytics and data interoperability have introduced increased concerns for effective privacy and security management, defense against data breaches, and data storage management. To address these concerns, government participation in developing, promulgating, and enforcing standards will increase (for example, NIST Standards for cloud-based security). Business intelligence (BI) vendors will Fall 2015 20 expand their offerings to accommodate very large data sets and big data analytics. Legal and human rights debates will become more contentious, especially about topics such as data ownership and stewardship as they relate to data sharing and individual privacy. In the health domain EHR adoption will continue sporadically and in geographic isolation, especially by the less-funded practices. Because of the extensive knowledge base required to perform big data analytics effectively and ethically, big data analytics will become increasingly the domain of intellectual and educated elites. References Cavoukian, A., and K. El Eman. 2011. Dispelling the myths surrounding de- identification: anonymization remains a strong tool for protecting privacy. Toronto, Canada: Infonnation and Privacy Commissioner of Ontario. Cohen, G., R. Amarasingham, A. Shah, B. Xie, and B. Lo. 2014. “The legal and ethical concerns that arise from using complex predictive analytics in health care.” Health Affairs, 33(7): 1139-1147. CBS News. 2015. CEO: 5,000-percent drug price hike "not excessive at all." [~http://www.cbsnews.com/news/turing-pharmaceuticals-ceo-martin-shkreli-defends- 5000-percent-price-hike-on-daraprim-drug/ (accessed 22 September 2015)]. Falik, D. 2014. “For Big Data, Big Questions Remain”. Health Affairs 33(7): 1111-1114. Jacobs, A. 2009. “The pathologies of Big Data”. Communications of the ACM 52(8): 36- 44. Kum, H., A. Krishnamurthy, A. Machanavajjhala, M. Reiter, and S. Ahalt. 2014. “Privacy preserving interactive record linkage (PPIRL).” Journal of the American Medical Informatics Association 21: 1-4. Ohno-Machado, L. 2012. “Big science, big data, and the big role for biomedical informatics.” Journal of the American Medical Informatics Association: 19: el. NBC News. 2015. Price Fiike for Tuberculosis Drug Cycloserine Rolled Back From 2,000% Jump, [http://www.nbcnews.com/health/health-news/price-hike- tuberculosis-drug-cycloserine-rolled-back-2-000-jump-n43 1716 (accessed 22 September 2015)]. Nussbaum, M. 2000. Women and human development: the capabilities approach. Cambridge: Cambridge University Press. Sahoo, S., C. Jayapandian, G. Garg, F. Kaffashi, S. Chung, and A. Bozorgi. 2014. “Heart beats in the cloud: distributed analysis of electrophysiological ‘big data’ using cloud Washington Academy of Sciences 21 computing for epilepsy clinical research.” Journal of the American Medical Informatics Association 2 1 : 263-27 1 . Shah, N. H., and J.D. Tenenbaum. 2012. “The coming of age of data-driven medicine: translational bioinformatics’ next frontier.” Journal of the American Medical Informatics Association 19: el-e2. Sen A. 1982. Commodities and capabilities. Oxford University Press. Sen, A. 1999. Development as freedom. New York: Knopf. United Nations (UN). 2002. Human rights: a compilation of international instruments. Vol. 1, Parts 1 and 2. New York: United Nations. United Nations. (UN) 2015. Millennium development goals reports. [http://www.un.org/millenniumgoals/reports.shtml (accessed 22 September 22 2015)]. United States Department of Health and Human Services (HHS). (2015). Beacon Community Program, [http://www.healthit.gov/policy-researchers- implementers/beacon-community-program (accessed 22 September 2015)]. United States Department of Health and Human Services (HHS). 2015. Health IT Legislation and Regulations, [http://www.healthit.gov/policy-researchers- implementers/health-it-legislation (accessed 22 September 2015)]. Endnotes 1 See, for example, periodic issues dedicated to the use of big data published in such trade journals Federal Computing Weekly or Healthcare Informatics. Frequency of webinars dedicated to big data offered by The Data Warehouse Institute (TWDI) may also prove instructive. 2 With the advent of big data and the increasing of HTML-based EHRs (with the HL7 Consolidated Clinical (CCDA), it is possible to embed genomic data in a patient’s EHR, offering the possibility of developing individualized medical protocols for patients. Fall 2015 22 Bio Erik W. Kuiler has spent most of his career as an Information Engineer, focusing on the development of lexicons, ontologies, and systems to support the management of data and information as enterprise assets. His research interests include software and information engineering, rhetoric and information theory, medieval studies and comparative literatures, and public policy. His recent publications are on big data analytics, development economics, and policy analysis. Washington Academy of Sciences 23 Big Data: Who’s Accountable? Jong-on Hahm George Washington University Abstract Data analysis. Big or Small, requires careful handling of data to ensure against sorting bias and errant categorization that can lead to inaccurate conclusions. Sorting error may be introduced when attempting to hannonize existing datasets with new datasets that offer many more parameters. Caution in data categorization, searching for specific factors, and drawing conclusions, is paramount for policymakers looking to use Big Data for societal benefits. A state decides to SET aside economic development zones and wants to encourage minority residents to establish businesses and hire workers. To determine target areas for public outreach, data are scraped from publicly available sources, cleaned, analyzed, and mapped to identify communities where resources should be expended. When program administrators pull up the first map, they are surprised to discover complete blanks in regions where significant populations of minorities are known to reside. A retailer planning an expansion into a new region pursues a marketing scheme intended to identify specific populations. The first advertising blitz includes a direct mail campaign where residents are offered free trials and samples of products. After the first mailers are sent, the retailer is bombarded with negative feedback and complaints about inappropriate and offensive product information. A consulting firm hired to improve efficiency and service in a surgical unit at a hospital talks to every staff member of the surgical unit. It devises an electronic tracking system to harmonize scheduling, smooth patient transfer protocols and keep the unit at high functional capacity. Within the first week, the schedule has bogged down completely and patients have had to be referred to nearby hospitals. In all of the above examples, something in the use of data has led to unintended, sometimes risky outcomes. As has become clear, the use of Big Data presents its own set of methodological and analytical challenges. The question then arises: when Big Data goes wrong, who’s accountable? Fall 2015 24 The advent of Big Data has seduced the unwary into the promise of the possible, overlooking the promise of accompanying problems. In enormous data sets, very slight differences in sorting and categorization can lead to large differences, particularly as data are collected into ever-larger sets. In a long-term study, when data increase by orders of magnitude over time, minute differences could grow into significance and result in greatly disparate impacts. When examining accountability, the goal should not be to conduct a forensic analysis of where, what, and how things went wrong, but rather to establish parameters from the outset that prevent such errors in the first place. In Big Data analyses, critical issues must be considered in the research design: sorting bias and harmonization of old and new datasets. Oftentimes, data collection is driven by the tools available. For example, if a research project will use a certain analysis program or approach, data will likely be collected in a format most amenable for use with that program. If the data become unwieldy or the desired granularity is different, sorting characteristics may be changed to make it work better with the analytical tool. Sometimes, data may be sorted according to a researcher’s own categorization algorithm without any conscious realization of such sorting. In 2013, the Wikipedia community noticed that its list of “American novelists” no longer contained any women. They had all been placed on a separate list of “American women novelists” (Filipacchi 2013). At the time, Wikipedia wanted to make the list for “American novelists” less unwieldy, and began creating subcategories (Neary 2013). This separation unleashed a firestorm of criticism from those who perceived the categorization as sexism, as there was no subcategory for “American men novelists” (McDonough 2013; Flood 2013). Whereas the constraints of Wikipedia’s platform may have led to this reclassification, the bias in characterizing standard “American novelists” to be male illustrates the potential for sorting bias in handling data even before analysis is attempted. Data can be altered through stratification, separation, or combination. Data cleaning can alter datasets such that nuances of the raw data, small shifts, offsets or trends that may indicate the influence of unanticipated factors is lost. Washington Academy of Sciences 25 The negative impact of bias in data analysis and its impact have been noted, sometimes in spectacular fashion. Photo auto-tagging programs developed by Yahoo and Google led to African-Americans being tagged with terms like “gorilla” and “ape” (Dwoskin 2015). Particularly in software that uses machine learning algorithms, a biasing factor such as selective data sets used for training can push the learning in an unintended direction. Unfortunately, avoiding bias is proving to be a much more daunting task than previously conceived. According to Valerio Pascucci, a leading researcher (XSEDE15, July 28, 2015), simply looking for something in a large data set will inject bias into the data analysis (Gibson 2015). Ideally, such unintended directional tacks will not occur in most research projects. Data will be derived, for the most part, from known data sets, analysis will use well-established, commonly used methods, data comparability and compatibility will not be an issue. The challenge arises when new data sets, offering richer, more informative views of a subject population become available, and researchers want to incoiporate the new information with the established set. At this point, the difficulty of data harmonization, in even the most basic ways, becomes evident. As computational capacity increases, an obvious difficulty arises in matching categories end to end. The US Census in 2020 will include an expanded range of choices for race and ethnicity to reflect the growing percentage of Americans who identify as multiracial (Pew Research Center 2015). In the example of the state wanting to spur minority entrepreneurship and workforce development, the state government may have wanted to harmonize existing data with newly available data on its minority populations. The state may have had to rely on existing district maps with much more limited population information. Constrained by budget restrictions and the tools available, harmonization of data sets may not have been as tailored as could be, leading to incorrect targeting of desired populations. The retailer may have introduced sorting bias that conflated demographics with product interest. The third example is based on an actual case and illustrates how sorting bias can creep into even Small Data analyses. The consultant hired Fall 2015 26 to help improve surgery scheduling developed an electronic scheduling system to replace a whiteboard on which surgery schedules were manually written with markers. The system degenerated because the consultants did not take into consideration the key role played by technicians who maintained supply stocks in the surgical suites. The technicians were not interviewed in the original data collection, nor were they provided access to the new electronic scheduling system. Hence, they did not stock the suites appropriately for each procedure. After a week of confusion and referral of patients to nearby hospitals (with concomitant loss of revenue), surgical scheduling reverted to the whiteboard system. The very nature of Big Data underscores the potential impact of infinitesimal differences that can become magnified in the petabytes of data being generated and mined for meaningful information. While analytical models can be changed with little imprint, impact on society and communities lingers on. References Dwoskin, E. 2015. “How social bias creeps into Web technology.” The Wall Street Journal, 21 August. http://www.wsi.com/articles/computers-are-showing-their- biases-and-tech-finns-are-concerned- 1440 102894 (accessed September 23, 2015). Filipacchi, A. 2013. “Wikipedia’s Sexism Toward Female Novelists.” The New York Times, 24 April, http://www.nytimes.com/2013/04/28/opinion/sunday/wikipedias- sexism-toward-female-novelists.html (accessed September 23, 2015) Flood, A. 2013. “Wikipedia bumps women from ‘American novelists’ category.” The Guardian, 23 April, http://www.theguardian.com/books/2013/apr/25/wikipedia- women-american-novelists (accessed September 24, 2015) Gibson, S. 2015. Exploring Farge Data for Scientific Discovery. HPCwire, 27 August. http://www.hpcwire.com/2015/08/27/exploring-large-data-for-scientific-discovery/ (accessed September 23, 2015) McDonough, K. 2013. American women novelists segregated by Wikipedia. Salon.com, 25 April. http://www.salon.com/2013/04/25/wikipedia moves women to american women novelists category leaves men in american novelists/ (accessed September 24, 2015). Washington Academy of Sciences 27 Neary, L. 2013. What’s in a Category? “Women Novelists” Sparks Wiki Controversy. National Public Radio , 29 April, http://www.npr.org/20 13/04/29/1 79850435/what-s- in-a-categorv-women-novelists-spark-wiki-controversv (accessed September 24, 2015). Pew Research Center. 2015. “Race and Multiracial Americans in the U.S. Census.” Multiracial in America, 11 June. http://www.pewsocialtrends.orR/20 15/06/1 1 /chapter- 1 -race-and-multiracial- americans-in-the-u-s-census/ (accessed September 28, 2015). BIO Jong-on Hahm is Special Advisor for International Research in the Office of the Vice President for Research at George Washington University, and is Distinguished Senior Fellow in the School of Policy, Government, and International Affairs at George Mason University. Her research focuses on global investments in science, technology, engineering, and mathematics (STEM) to spur innovation and economic growth; global STEM workforce migration; diversity in science; global intellectual property rights; and Big Data uses, analyses, and controversies. Fall 2015 28 Washington Academy of Sciences 29 Exploring Bias and Error in Big Data Research Katie Seely-Gant and Lisa M. Frehill Energetics Technology Center Abstract The availability and usability of massive data sets have added to the increasing popularity of big data research. However, common mechanisms of big data collection ( e.g ., social media, open source platforms, and other online user data) can be problematic. Sampling issues, especially selection bias, associated with these data sources can have far reaching implications for analysis and interpretation. This paper examines the types of sampling issues that arise in big data projects, how and why biases occur, and their implications. It concludes by providing strategies for dealing with sampling and selection bias in big data projects. From the dawn of humankind to the year 2003, it is estimated that 5 exabytes (101S bytes) of data were created by humans. Today, humans create about 2.5 exabytes of data every day (Intel IT Center 2012; Sagiroglu and Sinanc 2013). This explosion of information, due in large part to developments in data mining and collection, data warehousing and storage, and computational capacity and performance, has led to “the era of big data.” Individual-level data can now be collected and mined using online platforms, social media, and cell phone applications, giving big data researchers increasing levels of insight into previously unobserved behavioral patterns and other “found data” (Harford 2014; Tufekci 2014; Fan et al. 2014; Sagiroglu and Sinanc 2013; Yang and Wu 2013; De Mauro et al. 2014). With this influx of data researchers have been able to make significant strides in fields such as health care, finance and economics, and social science; however, big data research is not a panacea for data analytics. Though some data scientists subscribe to the “myth of large n” - i.e., when data are “big” enough biases are not significant -statistical errors and biases can still impact research findings regardless of the size of the dataset (Harford 2014; Anderson 2008; Lazer 2014). Due to the nature of data collection and mining, and the methods used therein, big data research may be particularly susceptible to sampling biases (Tufekci 2014; Fan et al. 2014; Harford 2014). Fall 2015 30 This paper will explore the dangers in the “myth of large n” by examining issues related to selection bias in big data research in particular, and will attempt to assess the extent to which big data projects may be affected by selection bias, the implications of this bias for research, and potential strategies for accounting for such limitations. We choose to focus on selection bias due to both the popularity of found data and mined social media data in big data analyses and the likelihood that these sources produce non-random samples. Big data are characterized by the “3 V’s”, volume (amount and size), velocity (real time or batch), and variety (structured or unstructured) (Sagiroglu and S inane 2013; De Mauro et al. 2014). The source and utility of big data can take many forms. Companies like Wal-Mart and Target collect and analyze real-time purchase data to predict consumer preferences, while economic researchers collect cell phone location data to determine the distance consumers are willing to travel to a shopping mall, a proxy for consumer demand and economic strength (Lazer 2014; Bollier 2010). Health and life science researchers have harnessed big data and increased computing power to revitalize genomic sequencing, compressing what had been a ten-year process to less than a week (Fan et al. 20 1 4; Harford 20 1 4). Covering all “3 V’s”, data mined from social media, cell phone applications, and open source online platforms provide big data researchers with unique insight into human behavior and interactions by providing large, real time data on their users and their content (Bollier 2010). With such large n’s, big data research offers interesting new ways to conduct analyses. In lieu of the traditional method of formulating hypotheses and theory before analyzing the data, big data researchers often take a high-level look at massive data sets, noting interesting or unexpected correlations, and then forming hypotheses and theories around those correlations (Lohr 2012; Harford 2014; Anderson 2008; Bollier 2010). This exploratory approach has led some to term the era of big data as the “end of theory” (Anderson 2008; Bollier 2010), suggesting that deductive reasoning grounded in previous research literature is no longer necessary with such large, timely, and varied data. By pursuing these exploratory approaches, and making claims based on discovered correlations, researchers risk falling prey to the traditional limitations and biases inherent in both statistical and social research. When Washington Academy of Sciences 31 researchers subscribe to the “myth of large n”, or the “n = all” suppositions, certain faults of big data may be ignored (Lazer 20 1 4; Bollier 20 1 0; Harford 2014). In particular, big data are especially susceptible to endogeneity, auto- correlation of errors, spurious correlations, and selection bias (Fan et al. 2014; Lazer 2014; Tufekci 2014; Harford 2014). The emphasis on social media data mining and other data collection from open source platforms and applications increases big data vulnerability to selection bias in particular (Fan et al. 2014). Selection bias describes the bias that is present when the selection of a sample or study group is such that proper randomization is not achieved and the sample is, therefore, not representative of the larger population (Berk 1983; Heckman 1979). Figure 1 shows a classic image that resulted from selection bias. The Chicago Tribune , which relied heavily on telephone surveys for their election predictions, prematurely claimed Dewey as the winner of the 1948 presidential election. The high cost of telephone lines led to biased results as affluent Americans were more apt to support Dewey than those who were less affluent. Selection bias is particularly problematic when relying on exploratory research and analyzing correlations, since self-selected samples often exhibit different correlational tendencies than random samples. Most important for big data analysis is the presence of confounding variables, or that persons who self-select into certain groups often have other variables in common that researchers are not accounting for such as demographic factors or similar environments, which cause the confounding variables (Tufekci 2014; Fan et al. 2014). More simply, selection bias describes the likelihood that certain persons or groups are more apt to be picked up by big data collection efforts than others, whether due to their use of social media and open source platforms, the availability of internet connectivity in certain areas, their ability to purchase smart phones and access applications, or any other number of omitted variables. For example, StreetBump, a smart phone based application rolled out in Boston, sought to record potholes while users were driving and report these potholes to the city for repair. Eventually when potholes were being disproportionately reported in affluent neighborhoods, a deeper analysis revealed an issue with selection bias. Residents in affluent neighborhoods were far more likely to own smart phones with network Fall 2015 32 access — and cars — than residents who lived in lower income neighborhoods (Harford 2014). Figure 1. Classic Case of Selection Bias: Chicago Tribune Declares “Dewey Defeats Truman” (photo credit: Associated Press). These demographic findings are consistent with 2014 Pew Research Center survey findings on smart phone users. Pew’s survey estimated that about 65 percent of American adults are smart phone users, but that population is, on average, under 40 years old, college-educated, and affluent (i.e., average incomes of more than $75,000 per year), and far more likely to be employed than non-smart phone users (Pew Research Center 2015). Given the demographics of smart phone users, it is problematic to use smart- phone data to make general claims about the U.S. population. Additionally, these problems may be compounded if researchers take exploratory approaches and simply analyze the data for interesting correlations, as additional context is usually needed to determine what variables are driving the correlation and what factors may cause the correlation to break down (Harford 2014). Selection bias has serious implications when left uncontrolled in a standard linear model as it creates a non-linear relationship between the dependent and key independent variable, such that a causal relationship may be misinterpreted and effects resulting from random noise in the model are mistaken for causal effects, affecting both internal and external validity. Washington Academy of Sciences 33 External validity is undermined when selection bias is present by underestimating the slope of the regression line, often leading to causal effects being underestimated. Internal validity may be similarly compromised if the effect of the “exogenous variable and the disturbance term are confounded”, leading to causal effects of an independent variable being confused with random noise in the data (Berk 1983, p. 388). By failing to formulate a theory and corresponding model, researchers are often unable to control for significant selection biases and jeopardize the validity of their research (Berk 1983; Heckman 1979). Other big data mining and collection efforts that have assumed “large n = no bias,” analyze social media user data. For example, Twitter-generated data has become quite popular among big data researchers because of the ease of data collection along with its connection to user data, such as tweet content, retweets, and engagement in trending topics. However, only 23 of U.S. adults already online use Twitter, and among that 23 percent, the population is largely minority (African-American and Hispanic) youth, with about 37% of Twitter users under the age of 30 (Duggan et al. 2015). Big data researchers often use Twitter to gauge public opinion on hot issues or learn more about consumer preferences; however, these data are problematic because Twitter users are not comparable to the U.S. population as a whole (Tufekci 2014). Using sources such as Twitter presents other unique sampling and selection issues. A common avenue of big data analysis using Twitter is “hashtag analysis”; that is, using Twitter’s linking system (tagging a post with a “#” connects that post with a live feed of all users tweeting about that particular topic) to gauge public opinion on timely, hot button issues. Research by Tufekci (2014) highlights the inherent problem of selecting cases based on the dependent variable. That is, users are only able to be included in the sample if they have tagged their tweet appropriately. This issue causes researchers to overlook cases where the user has not linked the post but is still engaged in the larger conversation, and is necessarily subject to self-selection bias, as the user has made a conscious choice to tag their post and include their content in the larger discussion (Tufekci 2014; Geddes 1990). Additionally, those hashtags that are used for analysis are those that were successful (/%., generated a large base of users engaging in Fall 2015 34 the conversation), and are therefore different than those hashtags that were unsuccessful in generating conversation. There are several ways researchers may account for selection bias in big data analysis. Traditionally, selection biases are controlled for through statistical modeling. Researchers should examine the overall demographics of their data source and attempt to control for confounding variables. Certain models and statistical methods are also better equipped to handle such biases, such as a regression discontinuity design (Taylor 2014), difference-in-difference models with a matching model as suggested by Heckman (1979), and a non-linear Tobit model, used by both Heckman (1979) and Berk (1983). Additionally not all selection bias is problematic. In some cases, notably market research, selection bias enables greater targeting of advertising. For example, retailers use “just-in-time” coupon delivery to target specific buyers. Entertainment services like Netflix and Amazon use similar methods to make suggestions to viewers. Likewise, in research projects where motivated participants are desired, the act of participating in a hashtag conversation, itself may be noteworthy. In the example of self-selection in hashtag analysis using Twitter, researchers can better account for the confounding variable issues present in a self-selected sample by going beyond exploratory, correlational analyses. Twitter datasets should not be considered random or representative; rather these data should be recognized as self-selected and missing data treated as “missing not at random”, or missing due to unobserved or unknown variables. In these analyses, it is also worthwhile for researchers to examine the cultural and social contexts of these “trending topics” and interpret their findings appropriately (Tufekci 2014; Meiman and Freund 2012). Additionally, big data research projects can be strengthened by pulling dependent variables from external, validated sources. For example, a study examining political attitudes using Twitter or Facebook content data as independent or control variables could be strengthened by using voting behavior or voting registration data from the U.S. Census as a dependent variable. This strategy, in particular, is useful in guarding against “selecting on the dependent variable”, as discussed by Tufekci (2014). Alternatively, researchers may use outside, reliable sources, like the U.S. Census, to Washington Academy of Sciences 35 benchmark their findings, acting as a sort of “gut check” for the findings. It may also be worthwhile for big data researchers to explore mixed-methods approaches to their studies. By complementing big data analyses with surveys, interviews, and other data collection methods, researchers can better understand the larger context of their data and provide more robust, representative findings. Big data research seems poised to revolutionize data analytics as we know it. By amassing such large amounts of data, researchers can observe correlations that may not manifest in smaller samples and can analyze large, near real time streams of data from numerous sources. While detecting previously missed correlations can spur new research questions and new understandings of processes in fields like health and social science, it is not a substitute for established theory, hypothesis development grounded in the extant social science literature, and statistical modeling. These data, as shown through this paper, may also be subject to selection biases, which can skew the findings and implications of the research. By incorporating more “small data” methods and techniques into research, such as mixed- method studies, alternate non-linear models, and benchmarking, big data analysts can strengthen their studies and findings and advance big data research. References Anderson, Chris. 2008. "The End of Theory: The Data Deluge Makes the Scientific Method Obsolete." Wired 16-07. Berk, Richard A. 1983. “An Introduction to Sample Selection Bias in Sociological Data”. American Sociological Review , 48(3), 386-398. Bollier, David, and C. M. Firestone. 2010. The Promise and Peril of Big Data. Washington, D.C.: Aspen Institute, Communications and Society Program. De Mauro, Andrea, Marco Greco, and Michele Grimaldi. 2015. “What is Big Data? A Consensual Definition and a Review of Key Research Topics”. In A IP Conference Proceedings (Vol. 1644, pp. 97-104). Duggan, Maeve, Nicole B. Ellison, Cliff Lampe, Amanda Lenhart, and Mary Madden. 2015. “Social Media Update 2014,” Pew Research Center. [Available at: http://www.pewinternet.oru/20 15/01 /09/social-media-update-20 1 4/ (Accessed 10 September 2015)] Fan, Jianqing, Fang Flan, and Han Fiu. 2014. “Challenges of Big Data Analysis”. National Science Review, 1(2), 293-314. Fall 2015 36 Geddes, Barbara. 1990. "How the Cases You Choose Affect the Answers You Get: Selection Bias in Comparative Politics." Political Analysis 2: 131-150. Harford, Tim. 2014. “Big data: A big mistake?” Financial Times, 11(5), 14-19. Heckman, James J. 1979. “Sample Selection Bias as a Specification Error”. Econometrica 47(1) 153-161. Heckman, James, Hidehiko Ichimura, Jeffrey Smith, and Petra Todd. 1998. “Characterizing Selection Bias Using Experimental Data”. Econometrica, 66(5), 1017-1098. LaValle, Steve, Eric Lesser, Rebecca Shockley, Michael S. Hopkins, and Nina Kruschwitz. 2013. “Big data, Analytics and the Path from Insights to Value”. MIT Sloan Management Review, 2 1 . Lazer, David, Ryan Kennedy, Gary King, and Alessandro Vespignani. 2014. “The Parable of Google Flu: Traps in Big Data Analysis”. Science 343(6176), 1203-1205. Lohr, Steve. 2012. “The Age of Big Data”. New York Times, Feb. 11 2012. Meiman, Jon, and Jeff E. Freund. 2012. "Large Data Sets in Primary Care Research." The Annals of Family Medicine 10(5) 473-474. Pew Research Center. 2015. “The Smartphone Difference” [Available at: http://www.pewintemet.org/20 1 5/04/0 l/us-smartphone-use-in-20 1 5/ (Accessed 10 September 20 1 5)] Price, Megan, and Patrick Ball. 2014. “Big Data, Selection Bias, and the Statistical Patterns of Mortality in Conflict”. SAIS Review of International Affairs, 34(1), 9-20. Raghupathi, Wullianallur, and Viju Raghupathi. 2014. “Big Data Analytics in Healthcare: Promise and Potential”. Health Information Science and Systems, 2(1) [http://www.hissjoumal.eom/content/2/l/3 (accessed 10 September 2015)] Sagiroglu, Serfef and Sinanc Duygu. 2013. “Big Data: A Review”. Proceedings for the Institute of Electrical and Electronics Engineers 2013 Annual Conference. 42-47 Taylor, Eric. 2014. “Spending More of the School Day in Math Class: Evidence from a Regression Discontinuity in Middle School”. Journal of Public Economics, 1 17, 162-181. Tufekci, Zeynep. 2014. “Big Questions for Social Media Big Data: Representativeness, Validity and Other Methodological Pitfalls”. In ICWSM ’ 14: Proceedings of the 8th International AAAI Conference on Weblogs and Social Media. 504-514. Yang, Qiang, and Xindong Wu. 2006. “10 Challenging Problems in Data Mining Research”. International Journal of Information Technology & Decision Making, 5(04), 597-604. Washington Academy of Sciences 37 Bios Katie Seely-Gant is an analyst with the Analytics Team at the Energetics Technology Center in Waldorf, Maryland (U.S.). Her work involves a wide variety of projects in areas including science and technology workforce and career pathways, diversity and inclusion in technical workforces, international and domestic teams, and mentoring. Lisa M. Frehill is Senior Analyst and Acting Director of the Analytics Team at the Energetics Technology Center (Waldorf, Maryland, U.S.). She is currently on detail as Organizational Evaluation and Assessment Researcher at the National Science Foundation. Dr. Frehill is an internationally recognized expert on human resources in science and engineering, designing and executing program evaluations, strategic workforce planning, and change management. Fall 2015 Washington Academy of Sciences 39 Educating Data Scientists in the Broader Implications of their Work Heikki Topi and M. Lynne Markus Bentley University, Waltham, MA Abstract The number of degree programs in analytics and data science is increasing rapidly. Because of the strong industry demand for highly qualified analytics professionals, the need for education will continue to grow. Current programs provide strong coverage of the infrastructure and applications of analytics and data science, but they are lacking in the coverage of their legal, ethical, and societal implications. We argue that every analytics and data science program should include a significant emphasis on the implications and potential consequences of data science applications. Including these elements in the programs will help analytics professionals understand better the complex and nuanced relationships between their work and various stakeholders of the context in which the work takes place. Data scientists and analysts who are sensitive to data downsides as well as upsides enable organizations to avoid harmful consequences of analytics applications but still achieve the benefits. Introduction Nothing is more crucial to the achieving the promises of big data than a workforce of capable individuals prepared to tackle the opportunities and challenges of analytics. Many commentators have mentioned projected shortfalls in the number of people qualified to fill the data scientist role (Craig et al. 2012, 2013; Manyika et al. 2011). A few analysts have also pointed to the need for technical support specialists who manage data preparation, storage, and related tasks (Woo 2013). Yet others have discussed deficiencies in the ability of managers and subject matter experts to sponsor, supervise, and take action on the results of analytic projects (Court 2015; Davenport 2013). The discussion has, however, paid less attention to what these people need to know. Through our participation in NSF-sponsored workshops on data science education1 and big data’s social, economic and workforce implications,2 we have identified important knowledge gaps related to the legal, ethical, and social implications of data science. In this contribution to the Symposium, we lay out the current state of data science Fall 2015 40 education and make the case for more attention to big data’s implications and consequences in data science education. Data Science Degree Programs The demand for professionals with proficiency in data science and analytic techniques has increased substantially during the last few years, and many media accounts have predicted a significant shortage of capable professionals in this area. A widely cited McKinsey report (Manyika et al. 2011) predicted a shortfall of nearly 200,000 knowledge professionals with in-depth preparation in analytics. Industry demand for graduates with this background has increased, and universities around the world have responded by launching both bachelor’s and master’s programs. North Carolina State University’s Institute for Advanced Analytics, an educational pioneer, maintains a database of related U.S. master’s programs. In September 2015, this list included 34 programs in Analytics, 19 in Data Science, and 54 in Business Analytics. Exact numbers of students are difficult to estimate, but program heads frequently boast of their success in attracting students from around the world. The largest programs admit hundreds of students annually, and all appear to bring in at least dozens. In total, data science and analytics programs already graduate thousands of students per year in the U.S. alone. The focus on data scientist preparation is not, however, only a U.S. phenomenon; new programs are also launching in Europe and Asia. We confidently expect the number of analytics graduates to continue to increase for some time. Data Science Curricula The disciplinary focus and orientation of data science and analytics programs vary significantly, and so do their curricula. University departments of statistics, mathematics, computer science, information systems, management science/operations research, and information science have set up programs, and so have schools and departments that focus on particular sectors, such as health care, finance, and the hard sciences. Some programs are interdisciplinary, crossing several departments or schools. Other programs ( e.g ., most, but not all, business analytics programs) are housed in single schools. Washington Academy of Sciences 41 Some data science and analytics programs focus primarily on the application of data science techniques to real world problems in science, engineering, health care, financial, educational policy, and the like. These programs tend to heavily emphasize data analytic techniques such as traditional statistical analysis, data mining, text mining, time series analysis, simulation, optimization, and machine learning, as well as domain knowledge. In addition, because graduates are expected to work closely with subject matter experts, these programs often include attention to general professional competences such as critical thinking, oral and written communication, collaboration and teamwork, and consulting skills for eliciting project requirements. The applications of data science increasingly involve continuous, real-time, algorithmic analysis of large quantities of data in a way that enables automated organizational decision-making. Therefore, some analytics programs (particularly in business schools) focus on the role of analytics in the digital transformation of organizations. Other data science programs focus more heavily on the activities involved in providing the underlying support for data science and analytics applications; these programs can be described as building skills in the technical infrastructure of data science. Examples of the topics emphasized in these programs are programming, algorithms and data structures, data visualization approaches, data warehousing, data management for structured and unstructured data, and so forth. A Missing Emphasis In addition to applications and infrastructure, a third key body of knowledge relevant to data science and analytics concerns their legal, ethical, and societal implications (see Table 1 that illustrates the three bodies of Data Science knowledge based on Markus and Topi, 2015). One can hardly mention the topic of big data without evoking privacy concerns, and discussion of security issues often follows closely behind. Also relevant are concerns about illegal discrimination, behavioral manipulation, harassment, and inappropriate “social sorting” (or labeling people via identity profiles) (Markus and Topi 2015). These legal, ethical, and societal implications of big data are rarely given attention proportionate to their importance in data science, analytics programs, or educational materials (Provost and Fawcett 2013). Instead, such topics are typically covered in educational programs of law, accounting information systems, social Fall 2015 42 sciences, and public policy, where data scientists-in-training may not be exposed to them. Table 1: Three Bodies of Data Science Knowledge (adapted from Markus and Topi, 2015) Applications • Use of data science tools and techniques to generate new insights in domains such as marketing, health care, law, finance, science • Use of data science tools and techniques to fully or partially automate previously manual decision-making processes such as the auctioning of advertising, mortgage or insurance underwriting, medical diagnosis, e-discovery, securities trading, identification of promising drug molecules • Development of new “apps” and data-oriented business processes and business models Infrastructure • Development of new tools and techniques for data handling (e.g., extracting, transferring and loading data, data storage, tagging and curating data, cleaning and verifying data) • Development of new software and hardware tools and techniques for data analysis and interpretation (e.g., text mining, data visualization, machine learning) Implications • Laws and regulations governing data protection, data security, and data management requirements (e.g., document retention and destruction) • Design of organizational structures and governance mechanisms that promote responsible (legal, ethical, and socially acceptable) data collection and use practices • Potential positive and negatives consequences of big data applications, tools for anticipating them, and strategies and techniques for minimizing negative side-effects. Washington Academy of Sciences 43 Most data science and analytics degree programs appear to take~ either implicitly or explicitly — a strong pro-innovation stance, implying that the consequences of big data are uniformly positive or, if negative, easily remediable. We believe that this stance creates important gaps in the preparation of the future data science and analytics professionals. In particular, we believe that they are underprepared for the significant ethical challenges they are likely to confront throughout their careers. Why an Implications Focus Is Essential Over the last century, thoughtful scientists, engineers, and technical professionals have taken powerful stands on the uncertain or potentially negative consequences of innovations such as fossil and atomic energy, genetic engineering, and nanotechnology. Appropriate responses to such issues are never easy to find and are always contested, but nothing is gained by sweeping the issues under the rug. Failure to raise concerns and debate the issues publicly generally only convinces the public that there is something to hide and creates opposition that can block beneficial innovations. This is as true of big data as it is of nuclear power. For instance, the InBloom big data educational innovation was terminated after parents voiced fears over possible secondary uses of their children’s data (Kharif 2014). When professionals are primed to understand and raise questions about the possible downsides of a proposed data innovation, better technology uses and outcomes are possible for all. Among the reasons for preparing data scientists to understand the broader implications of their work are the following: • The systems that feed and flow from data analytics are often highly complex, drawing data from numerous sources both external and internal to an organization and involving interconnections among independently developed systems. The sheer complexity of such systems can give rise to unexpected outcomes and glitches. Education is needed to anticipate, prevent, diagnose, and correct such outcomes. • Relying on intuition in the interpretation of analytics results can lead to serious practical errors. Data scientists need to be well attuned to the sources of error in data and algorithms and to human reactions to labeling, subtle guidance, and constraints on their behavior. Fall 2015 44 • Individual perspectives and value judgments regarding the implications of analytics-based systems and the potential that analytics infrastructures create vary quite significantly. Education can help analytics professionals leam ways to put personal biases aside and arrive at a more neutral analysis and a more successful resolution of complex sociotechnical situations. • Automated decision-making systems that operate without continuous human involvement can amplify the negative consequences of flawed data analyses. “Invisible technical workers” (Ribes etal. 2013) make choices — both at the time of original design and implementation of the systems and during system operation (which might be fully integrated) — that potentially have far- reaching consequences for both individuals and organizations. These consequences are often opaque to organizational clients and users. In some cases, even technical specialists may not understand why their algorithms produce the results they do. Greater awareness of this possibility is needed to produce and update algorithms that work well, to devise effective ways for humans to intervene, and to give affected people the opportunity for redress when errors are made. In short, realizing the potential benefits of big data without the possible harms requires data scientists and analysts who are sensitive to data downsides as well as data upsides. Knowledge Areas and Pedagogical Approaches We believe that data science and analytics programs need modules and course(s) on the implications and potential consequences of analytics. These courses must be specialized to the particular legal, ethical and societal issues raised by big data. For example, simply adding a course on business ethics (which may cover foreign corrupt practices, abusive labor practices, and environmental damage) to a business analytics program does not rigorously address issues like data protection law, personal information privacy, online harassment, or glitches in online trading. Naturally, ethical theories and general principles are shared across contexts, but the way these principles are applied requires in-depth understanding of the dependencies and connections discussed above. Washington Academy of Sciences 45 It is beyond the scope of this paper to present a detailed proposal for courses (or course modules) on the legal, ethical, and societal implications of analytics and data science. However, we propose that the following knowledge areas should be covered: • Categories of major implications and potential consequences of analytics-based systems (as laid out, for example, in the framework proposed by Markus and Topi 2015) • Methods for identifying, analyzing, and understanding complex organizational situations from multiple perspectives in an unbiased way, particularly from the perspective of implications and potential consequences of technology-enabled systems • Rich collections of relevant real-world examples that illustrate the positive impacts of thorough implications analysis • Material that allows students to understand themselves as ethical decision-makers. It is particularly important that the students have a good understanding of the sources of potential biases in data, algorithms, and decision-making processes. Pedagogically, the modules or courses required for developing these competences need a good balance between the elements that build a strong conceptual foundation and those that apply methods of active, participatory learning. It is essential to allow students to internalize the theories and make personal discoveries through case analysis, interviews and observations in organizational settings, role play, games, simulations, and other similar pedagogical approaches. The modules also need exercises that help the students discover their own value positions on analytics issues. Many of these materials do not currently exist, and developing them should be an important priority for the data science and analytics community. Conclusion In this paper, we have argued that every analytics and data science program should include a significant emphasis on the implications and potential consequences of big data. This can be accomplished through a set of components (or a single course) that provide students with the opportunity to engage, theoretically and experientially, with the legal, ethical and societal implications, and potential consequences of big data. Without a systematic approach to developing these competencies, even highly trained and technically competent experts may approach their work Fall 2015 46 with perspectives that are too narrowly focused on the potential benefits of innovation and too neglectful of potential harms. We firmly believe that integrating attention to implications and consequences into every analytics and data science program will lead to great value for individuals, organizations, and society. Acknowledgement This material is based upon work supported by the National Science Foundation under Grants No. 1348929 and 1545135. Endnotes 1 Workshop on Data Science Education (National Science Foundation award DUE: 1545135), Heikki Topi, Principal Investigator and Lillian (Boots) Cassel, Co-Principal Investigator. 2 Big Data, Big Decisions — A Research Agenda Setting Workshop on the Social, Economic, and Workforce Consequences of Big Data (National Science Foundation award IIS: 1348929), M. Lynne Markus, Principal Investigator. Washington Academy of Sciences 47 References Court, David. 2015. “Getting Big Impact from Big Data.” McKinsey Quarterly (January). [http://www.mckinsey.com/insights/business_technology/getting_big_impact_ from_big_data?cid=other-eml-alt-mkq-mck-oth-1501 (accessed September 25 2015)] Craig, Elizabeth, Charlene Hou, and Brian F. McCarthy. 2012. The Looming Global Analytics Talent Mismatch in Insurance. Accenture. [https://www.accenture.com/us-en/insight-looming-global-analylics-talent- mismatch-insurance.aspx (accessed September 25 2015)] Craig, Elizabeth, Charlene Hou, and Brian F. McCarthy. 2013. The Looming Global Analytics Talent Mismatch in Banking. Accenture. [https://www.accenture.com/us-en/insight- global-analytics-shortage-banking- summary. aspx (accessed September 25 2015)] Davenport, Thomas H. 2013. “Keep Up with Your Quants.” Harvard Business Review (July-August): 120-123. Kharif, Olga. 2014. “Privacy Fears Over Student Data Tracking Fead to InBloom's Shutdown.” Bloomberg Business. [http://www.bloomberg.eom/bw/articles/2014-05-01/inbloom-shuts-down- amid-privacy-fears-over-student-data-tracking (accessed September 6 2015)] Manyika, James, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles Roxburgh and Angela Hung Byers. 2011. Big Data: The Next Frontier for Innovation, Competition, and Productivity. McKinsey Global Institute. [http://www.mckinsey.com/ insights/business_technology/big_data_the_next_frontier_for_innovation (accessed September 25 2015)] Markus, M. Fynne and Heikki Topi. 2015. Big Data, Big Decisions for Science, Society, and Business- — NSF Project Outcomes Report. Waltham, MA: Bentley University. Provost, Foster and Tom Fawcett. 2013. Data Science for Business: What You Need to Know about Data Mining and Data-Analytic Thinking. Sebastopol, CA: O'Reilly. Ribes, David, Steven Jackson, Stuart Geiger, Matthew Burton, and Thomas Finholt. 2013. “Artifacts that Organize: Delegation in the Distributed Organization.” Information and Organization, 23: 1-14. Woo, Ben. 2013. “Combating the Big Data Skills Shortage.” Forbes, 18 Jan. [http://www.forbes.eom/sites/bwoo/2013/01/18/combating-the-big-data-skills- shortage/ (accessed September 25 2015)] Fall 2015 48 Bio Heikki Topi is Professor of Computer Information Systems at Bentley University in Waltham, MA. He has contributed to international computing curriculum development and evaluation efforts in various leadership roles since early 2000s (including CC2005 Overview Report; task force co-chair of IS 2010, the latest undergraduate IS curriculum revision; task force co-chair of MSIS 2016, ongoing revision of the IS master’s curriculum; and ongoing ACM Education Council initiative on Data Science education). He is co-author of a leading data management textbook Modern Database Management, now in its 12th edition. His co- edited Volume 2 Information Systems and Technology of CRC/Chapman & Hall’s Computing Handbook was published in May 2014. He has been a member of ACM’s Education Board since Spring 2006 and represented first AIS and then ACM on CSAB’s Board since 2005. M. Lynne Markus is the John W. Poduska, Sr. Professor of Information and Process Management at Bentley University, Visiting Professor at the London School of Economics, and Research Affiliate at MIT Sloan School’s Center for Information Systems Research. She was the Principal Investigator of an NSF workshop award to develop a research agenda on Big Data’s social, economic, and workforce consequences, and she participated in the White House 90-day review of Big Data. Markus was named a Fellow of the Association for Information Systems in 2004 and received the AIS LEO Award for Exceptional Lifetime Achievement in Information Systems in 2008. Washington Academy of Sciences 49 Everything Old is New Again: The Big Data Workforce Lisa M. Frehill Energetics Technology Center, Waldorf, MD Abstract “Big Data” is a relatively new term, often used imprecisely and often in contexts that imply a pressing need for workers with a newly-blended unique skillset. However, social scientists have worked with exceptionally large data sets for quite some time, historically accessing remote space, writing code, analyzing data, and then telling stories about human social behavior from these complex sources. Therefore, more than a half century of accumulated social science knowledge about extracting information from very large data sets to understand human social behavior provides a model for the emergent data science profession. In this article I present analyses of current and projected U.S. workforce data using various definitions of skillsets for data scientists, concluding with a discussion of the policy implications. Introduction “Big Data” is a relatively new term, often used imprecisely. Recent U.S. science and technology policy panels, including the President’s Council of Advisors on Science and Technology (PCAST 2015) and a U.S. National Academies of Science study group, deploy the term “big data” in a way that suggests that in the past, the issues of privacy, worker skills, and access were not encountered and are, therefore, new and in need of attention. However, social scientists have worked with exceptionally large data sets for quite some time, including implementing some of the much- touted benefits of big data such as merging multiple datasets from disparate sources into larger files, hierarchical file structures, and integrating quantitative and qualitative sources to derive insights into human social behavior. Hence, the methodological practices, ethical guidelines, data management, and statistical methods that have been honed in the social sciences' provide important workforce development lessons for the “new” big data. Additionally, depending on how one specifies the workforce skills requirements for big data, the size of the existing pool of talent varies, as does the prognosis for the labor market fortunes for data scientists as the Fall 2015 50 21st century’s “sexiest job” (Davenport & Patil 2012, RJMetrics 2015). In this article I start by differentiating the new big data as a form of what Groves (2011) describes as “organic” data, in contrast to the traditional large “designed” datasets that have been / continue to be collected, analyzed and reported on by social scientists. Next I will present analyses of current and projected U.S. workforce data using various definitions of skillsets for data scientists. Finally, I close by discussing the policy implications about the big data workforce. What is Big Data and How Does it Differ from Previous Types of Large Datasets? The term “Big Data” typically refers to what Groves of the U.S. Census Bureau has referred to as “organic” data (2011), in contrast to “designed data.” As these terms imply, designed data are the traditional raw material deployed by social scientists to answer research questions with carefully designed studies using tested and accepted methodologies to advance knowledge of the social world. Organic data are observational data generated by the day-to-day behaviors of people. Social scientists also gather organic data, but, again, such data collections are designed as opposed to the data mining approach associated with the new big data. Social scientists from many fields have worked with very large datasets. The computing power available in personal computers now far exceeds the capabilities of these machines just two decades ago. Those who worked with very large files such as those from the U.S. Census, the Panel Study of Income Dynamics, or the Department of Education’s High School and Beyond longitudinal study program (to name just a few), often accessed tape or cartridge-stored files using mainframe operating systems such as the IBM VAX. Researchers’ individual accounts at colleges and universities were typically insufficient in size to permit storage and analysis, necessitating that social scientists who were “quants” learn how to access and assign virtual space on which to park files and then perform statistical analyses using one of a number of statistical packages like BMDP, SPSS, or SAS (among others). Additionally, social scientists needed to learn the command syntax for these specialized packages, therefore, rudimentary programming skills were also a critical element of quantitative social scientists’ training. Washington Academy of Sciences 51 In the 1970s- 1990s, quantitative social scientists’ graduate programs typically included advanced courses in research design, including statistical analysis, but often did not emphasize visualizations, which were not as commonly deployed at that time in peer-reviewed research literature. Technical skills such as programming and command syntax were easily learned via workshops offered by campus computer centers, basic classes offered by computer science or management information science programs, or programming manuals. The academic program focused on the substance and design issues that provided a foundation for the analyses that social scientists would perform on the large datasets. Quantitative social scientists — in accordance with the deductive scientific method — started with ideas and then located the appropriate data, which could be used to answer research questions. With the substantial increases in desktop (and laptop) computing power, many social scientists now have the luxury of being able to store and analyze many of these same datasets on a local machine rather than mounting tapes and using remote computers. Additionally, many of the popular statistics packages developed windows-based products, which enable many quantitative social scientists today to more easily manipulate and analyze data without learning complicated command syntaxes. Turning to a consideration of the implications of big data with respect to observational data, three features now make organic/observational data “big”: velocity, variety, and volume, the three V’s (McAfee and Brynjolfsson 2012; De Mauro et al. 2015). Table 1 compares social science designed observational data and organic “big” data on the three Vs. In a nutshell, while the variety of designed and organic data are vast, the large organic designed data used by social scientists are orders of magnitude smaller than organic big data. Information derived from designed data requires more time to emerge than the pace with which the insights from organic data are demanded in business settings for data-driven decision making. Fall 2015 52 Table 1. Observational Data - Designed and Organic (i.e., Big) Factor Social Science - Designed Observational Studies Big Data - Organic Data Velocity Slow: once a study is designed, observational data are collected, sometimes using paper-and-pencil, sometimes using recording devices (audio and/or video as appropriate), and sometimes using minicomputers of various types. Once these data are gathered, analysis with qualitative analysis packages for the social sciences requires content to be manually tagged, sometimes by multiple coders. Fast: with the diffusion of smart phone technologies and video surveillance, individuals’ behaviors are able to be gathered instantly within a medium that may be used to analyze and influence subsequent behaviors. Variety In a designed study, scope is limited by the particular research question, time constraints, and technology associated with gathering relevant data. Though limited within a particular study, observational researchers have studied a dizzying array of topics. Large technology companies — including those that develop applications for smart phones — can access many types of data about many behaviors of individuals. Smartphone apps, access location information and are able to track individual preferences about entertainment, transportation choices, eating and exercise habits, and buying patterns (among others). Volume Generally megabyte-order of magnitude. “Big” is a misnomer for the petrabyte-sized files that are floating in cyberspace. Many of the techniques, tools, and protocols developed by social science research communities to manage and share large designed datasets — including attention to the ethical issues associated with collecting these data — hold important implications for the big data workforce. Since 1962 the Interuniversity Consortium for Political and Social Research (ICPSR) at the University of Michigan Institute for Social Research, has provided a repository for large social science datasets. ICPSR curates these data, provides support to social scientists via training sessions in quantitative methods, and has long-established and continuously evolving standards for the storage and use of data with attention to the issues of Washington Academy of Sciences 53 confidentiality and privacy. A 2014 PCAST report on privacy issues and big data recognized the potential value that could be added by the social sciences in its third recommendation: With coordination and encouragement from OSTP, the NITRD agencies should strengthen U.S. research in privacy- related technologies and in the relevant areas of social science that inform the successful application of those technologies. (PCAST 2014: xiii) The fourth recommendation indicates a need to incorporate education about privacy issues into the education and training of professionals who work with big data. Social science research methods classes — typically required courses at both the undergraduate and graduate levels — cover these issues. Additionally, online training platforms2 designed to educate and certify social, life, and medical scientists in issues associated with human subjects in research represent another mechanism by which those in disciplines that engage with big data-computer scientists, marketing researchers, and machine language programmers-could be educated in the complex issues associated with protection of individuals’ privacy. Workforce Considerations - What are the Skills Needed for “Data Scientists?” Estimates of the needs for the big data workforce vary widely because the skillset is a somewhat moving target. Starting in 2009, Hammerbacher suggested that the new big data required a new occupation, the data scientist, who, at Facebook, would be able to use a variety of programming skills — Hadoop, R, and Python — to access space for the new huge datasets and complete analyses. Similar emphasis on programming skills and alignment with computing and information technology (IT) disciplines was reflected in a 2010 PCAST report on Networking and Information Technology (NIT) Research and Development (NITRD): NIT is the dominant factor in America’s science and technology employment, and that the gap between the demand for NIT talent and the supply of that talent is and will remain large. Increasing the number of graduates in NIT Fall 2015 54 fields at all degree levels must be a national priority. Fundamental changes in K-12 education are needed to address this shortage. (PCAST 2010: 85) Other sources in 2010 and 2011, though, indicate additional skills beyond the technical computing skills cited by PCAST. A 2010 Economist article reported on the new profession, suggesting that data scientists “combine the skills of software programmer, statistician, and storyteller/artist to extract nuggets of gold hidden under mountains of data.” In 2011, McKinsey reported that the skills needed to exploit big data were so disparate, that three types of workers would be needed: Our research identifies three key types of talent required to capture value from big data: deep analytical talent — people with technical skills in statistics and machine learning, for example, who are capable of analyzing large volumes of data to derive business insights; data-savvy managers and analysts who have the skills to be effective consumers of big data insights — i.e., capable of posing the right questions for analysis, interpreting and challenging the results, and making appropriate decisions; and supporting technology personnel who develop, implement, and maintain the hardware and software tools such as databases and analytic programs needed to make use of big data. (Manyika et al. 2011: 103) As described above, these same skills are akin to those of quantitative social scientists who used programming skills to manipulate data and perform statistical analyses to extract information from very large datasets. Everything old is new again: data science is a new version of quantitative social science, but without the research foundation in human behavior and the ethical standards of the social sciences. It is important to recognize, however, that much of social science work with large datasets is basic research (i.e., fundamental knowledge creation), while the extraction of information from big data is applied research (i.e., enabling data-driven decision making in work settings). Most recently, however, some observers of the emergent data scientist profession are less optimistic about the future of this occupation. Washington Academy of Sciences 55 Darrow (2015) concludes, “Enjoy your fat salaries while you can data scientists, because the rising tide of new talent and — gasp — automation will take their toll.” The same Fortune article quotes Alex Cosmas of Booz Allen Hamilton (BAH), who indicates that BAH hires analysts and then trains them in the technical skills of data science: “We look for raw inquisitiveness, the intellectual curiosity which will repay you tenfold.” As in the past, the pool from which data scientists will be drawn is broader than the pool of those trained in computing or information technology. What is the Size of and Trends in the U.S. Data Science Workforce? There is no question about the proliferation of new occupations associated with computing and the burgeoning size of the information technology (IT) workforce. The ubiquity of computing technology has created the need for a host of workers in occupations that did not exist two decades ago. The rapidity with which demand for workers with IT skills as well as the variety of such skills have resulted in a number of mechanisms by which workers obtain these skills and, as well, how employers obtain the skilled workers they need. A recent report by RJMetrics (2015) used data from Linkedln to estimate the number of data scientists (worldwide) to be 1 1,400, many of whom held advanced degrees. Figure 1 shows U.S. Bureau of Labor Statistics (BLS) projections for growth in a number of science and engineering occupations for the 20 1 2- 2022 period along with the actual growth in these occupations in the previous ten-year period (2004-2014). Between 2004 and 2014, the number of jobs in the U.S. economy grew by 5.1 percent, with substantially more rapid growth in computing and mathematical sciences occupations, including those associated with software development, which grew by 37 percent and 27 percent, respectively. Architecture and engineering occupations barely grew, with a substantial contraction in hardware engineering. Projections of growth for the 2012-2022 decade match this pattern across occupations, with the most substantial growth projected for computing and mathematical sciences, especially software, both outpacing the 10.8 percent projected growth for the overall number of jobs. Fall 2015 56 U.S. Labor Force Growth - Actual and Projected □ Actual Growth, 2004-2014 ■ Projected Growth, 2012-2022 40% 30% 20% -C | 10% C? 0% -10% -20% Note: Software includes computer programming and software engineering, a subset of "Comp./Math. Sci.; Hardware includes Computer Hardware Engineers, a subset of "Arch. & Eng." Figure 1. Historical and Projected Demand for IT Workers Source: Analysis of data from the Bureau of Labor Statistics, 2014. “Table 1.2 Employment by detailed occupation, 2012 and projected 2022 (Numbers in thousands)” and Current Population Survey AAT-series Table 1 1 for 2004-2014. The plots in Figure 2 provides another way to understand the past decade of change in these technical occupations in comparative perspective. For both median weekly earnings of full time workers and the overall number of workers in each occupational category, a change ratio was computed as follows: Change ratio (Epcc, 2014 Epcc, 2004) Eocc, 2004 (Etotal, 2014 ~ Etotal, 2004) Etotal, 2004 The change ratio was computed for four categories of occupations (denoted as occ in the subscripts in the equation): computer and mathematical sciences, architects and engineers, software, and hardware. The x-axis shows the employment change ratio, while the y-axis plots the change ratio of median weekly earnings for these same four occupations. A change ratio of unity would indicate that the change in the specific Washington Academy of Sciences 57 occupation is on par with that in the rest of the economy, while ratios under one show slower change and ratios greater than one indicate faster growth. As shown in the Figure 2, only growth in software occupations outpaced the U.S. in both numbers and median weekly earnings, while computer and mathematical sciences outpaced the U.S. in earnings but not numbers. Both architects and engineers, including computer hardware engineers, grew slower than the U.S. economy over the same decade in terms of both earnings and number of workers. 8 7 ♦ C/MS 6 'o 5 ro J Cd 2 4 ♦ Software O (N t o 3 o <~\i _ oo 2 c ro -C U ! . 1 QO c c n ♦ Arch & Eng ra w ( -l -2 0 2 0 4 0 6 0 8 : 1 2 1 4 ♦ " J Employment Change 2004-2014 (Ratio) Hardware Figure 2. Earnings and Employment Change Ratios, 2004-2014 C/MS: Computer and mathematical sciences Software: Software developers, applications, and systems software Hardware: Computer hardware engineering Arch & Eng: Architects and engineers Source: Analysis of data from the Bureau of Labor Statistics, Current Population Survey AAT- series, Table 1 1 2004 and 2014 (employment) and Table 39 2004 and 2014 (Median weekly earnings for full time wage and salary workers). Fall 2015 58 Figures 1 and 2 show data only for the IT occupations associated with big data. It should be noted, though, that academic discipline silos complicate definition of the big data workforce. On the one hand, as described above, quantitative social scientists who work with very large data sets to generate new, basic science knowledge of human social behavior do not use the term “big data” to describe the work that they do. Computer scientists have embraced the term “big data.” Online forums emphasize the primacy of programming, which reinforces a professional boundary on the skills associated with accessing and analyzing these organic data. The popular pej oration of social science — clearly demonstrated by U.S. Congress members’ frequent attacks on social science projects funded by the National Science Foundation, for example — reinforces this barrier. The emphasis on technical programming skills and algorithm development has been suggested as a replacement for the theory development process with respect to social data, with one observer claiming that “the data deluge makes the scientific method obsolete.” (Anderson 2008). Conclusion Big data has abundant applications for business, health, and finance. The ability to rapidly analyze exceptionally large data sets from multiple sources to provide information to enable actions in real-time offers promise in a range of areas. For example, big data may enable more precise dosing of medications and has been used to develop sensor technology to determine when a football player needs to be side-lined because of his/her heightened risk of concussion. Consumers may experience more efficient service and process efficiencies may yield lower prices for consumers and higher profits for businesses. The emergence and evolution of the data science occupation bears on-going scrutiny. In just a few years, employers have seen the value associated with a cadre of workers who have both technical skills as well as the ability to tell a story with data. However, as noted by BAH’s Cosmas, locating inquisitive analysts and then training them up in the technical skills may be the likely direction that will be taken with this workforce. In this case, the potential recruitment pool is far wider than graduates of computer science programs and, indeed, computer science programs will need to provide students with experiences that encourage inquisitiveness about Washington Academy of Sciences 59 human social behavior and with robust training about privacy and confidentiality on par with that in social science methods. There are many potential benefits that may be derived from cross- pollination between computing, on the technical side, and social sciences, on the substantive side, to deploy big data as a tool for human advancement beyond capitalist accumulation. Both sets of fields, however, need to be wary of professional boundary heightening, which introduces inefficiencies. Time, energy, and effort are needed to develop data science as a truly transdisciplinary field that can yield both an advancement of basic science knowledge about human social behavior as well as applied science information for data-driven decision making in real world contexts. There is a place in such a transdisciplinary field for both designed and organic data, the latter of which may be more effectively translated into information when there is thoughtful consideration of research questions, the literature that informs those questions, and use of previously developed analytical methodologies. Better translation of social science research into actionable information may help diminish the challenges of its relevancy that have plagued public funding of social science. In the 1970s- 1990s, inquisitive social science practitioners demonstrated that the secrets of accessing and analyzing very large datasets were relatively easy to acquire; the current trends in big data analytics suggest this to be similar with respect to data science now. While the volume and velocity of basic research in the social sciences is smaller and slower than in big data, the same variety of data sources and implications for data quality -i.e., validity and reliability-are similar. So everything old is new again; more than a half century of accumulated social science knowledge about extracting information from very large data sets to understand human social behavior provides a model for the emergent data science profession. Fall 2015 60 Endnote 1 Social sciences are taken in a broad sense to include fields categorized as such by the National Science Foundation (e.g., anthropology, sociology, psychology, political science, and economics) as well as fields that deploy similar methods such as marketing and educational research. 2 These platforms include: the Collaborative Institutional Training Initiative at https://www.citiprogram.org/: the National Institutes of Health Protecting Human Research Participants at https://phrp.nihtraining.com/users/login.php; and the FHI360 Research Ethics Training Curriculum at http://www.flri360.org/sites/all/libraries/webpages/fl~ii-retc2/index.html. Washington Academy of Sciences 61 References Anderson, Chris. (2008). “The End of Theory: The Data Deluge Makes the Scientific Method Obsolete” Wired. Accessed online at http://archive.wired.com/science/discoveries/inagazine/16-07/pb_theory/ (original post: 06.23.08). Darrow, Barb. (2015). “Data science is still white hot, but nothing lasts forever” Fortune (online) accessed at http://fortune.com/2015/05/21/data-science-white-hot/ 30 August 2015. Davenport, Thomas H. & D.J. Patil. 2012. Data Scientist: The Sexiest Job of the 21st Century. Harvard Business Review. 90(10): 70-76. De Mauro, Andrea; Marco Greco, and Michele Grimaldi. 2015. "What is big data? A consensual definition and a review of key research topics" A/P Conference Proceedings 1644: 97-104. (http://scitation.aip.org/content/aip/proceeding/aipcp/10. 1063/1.4907823). doi: 1 0. 1 063/1 .4907823 (https://dx.doi.org/ 10. 1 063%2F 1 .4907823). Economist (2010). “Data, Data Everywhere” accessed online, http://www.economist.com/node/15557443, originally posted 25 FE 2010. Groves, Robert. (2011). “Designed Data” and “Organic Data” blog posted May 31, 201 1 accessed at http://directorsblog.blogs.census.gov/201 1/05/3 1/designed-data-and- organic-data (accessed 28 August 2015). Hammerbacher, Jeff. 2009. “Information Platforms and the Rise of the Data Scientist” pp. 73-84 in Segaran, Toby, and Jeff Hammerbacher Beautiful Data : The Stories behind Elegant Data Solutions. Sebastapol, CA: O’Reilly. McAfee, Andrew, and Erik Brynjolfsson. 2012. “Big Data: The Management Revolution” Harvard Business Review. (October 2012). Manyika, James, Michael Chui, Brad Brown, Jacques Bughin, Richard Dobbs, Charles Roxburgh, and Angela Hung Byers. 2011. “Big data: The next frontier for innovation, competition, and productivity.” McKinsey Global Institute. National Research Council. 2014. Training Students to Extract Value from Big Data: Summary of a Workshop. Washington, DC: National Academies. President’s Council of Advisors on Science and Technology (PCAST). 2015. “Big Data and Dala-Intensive Computing.” Accessed online at www.whitehouse.gov/ostp/pcast. (August 2015). . 2014. “Big Data and Privacy: A Technological Perspective.” Accessed online at www.whitehouse.gov/ostp/pcast. (May 2014) . 2010. “Designing a Digital Future: Federally Funded Research and Development in Networking and Information Technology” accessed online at www.whitehouse.gov/ostp/pcast. (December 2010). Fall 2015 62 RJMetrics. 2015. “The State of Data Science” RJMetrics, Benchmark Report Series (October 2015). [https://rjmetrics.com/resources/reports/the-state-of-data-science/ (accessed 8 October 2015)]. U.S. Bureau of Labor Statistics. 2014. “Current Population Survey, AAT-series, Table 11. Employed persons by detailed occupation, sex, race, and Hispanic or Latino ethnicity” [http://www.bls.gov/cps/tables.htm (accessed 8 October 2015)]. . 2004. “Current Population Survey, AAT-series, Table 1 1. Employed persons by detailed occupation, sex, race, and Hispanic or Latino ethnicity” [http://www.bls.gov/cps/tables.htm (accessed 10 February 2014)]. . 2014. “Current Population Survey, AAT-series Table 39. Median weekly earnings of full-time wage and salary workers by detailed occupation and sex.” [http://www.bls.gov/cps/tables.htm (10 February 2014)]. . 2004. “Current Population Survey, AAT-series Table 39. Median weekly earnings of full-time wage and salary workers by detailed occupation and sex.” [http://www.bls.gov/cps/tables.htm (10 February 2014)]. . 2014. “Table 1.2 Employment by detailed occupation, 2012 and projected 2022.” [http://www.bls.gov/cps/tables.htm (10 February 2014)]. Bio Lisa M. Frehill is Senior Analyst and Acting Director of the Analytics Team at the Energetics Technology Center (Waldorf, Maryland, U.S.). She is currently on detail as Organizational Evaluation and Assessment Researcher at the National Science Foundation. Dr. Frehill is an internationally recognized expert on human resources in science and engineering, designing and executing program evaluations, strategic workforce planning, and change management. Washington Academy of Sciences 63 Social Media Analysis for Higher Education Anamaria Berea, William Rand, Kevin Wittmer University of Maryland Gerard Wall vibeffect Abstract The educational system involves a complex set of actors, including learners, parents, teachers, and administrators. However, we now have more data than ever to analyze this system, which could result in a quick understanding and evaluation of public policies in this complex policy area. This paper explores a new area of data about the educational experience, namely social media data. This paper outlines an exploratory analysis of the Twitter discussions regarding higher education in the USA. Based on a collection of more than 1.5 million tweets over a period of 4 months, we identify a few key issues in the current higher education discourse on social media. We also identify the effect of the expressed feelings of the social media users when it comes to college applications, decisions and completion. We conclude that policies in higher education can be better tailored if they are informed by social media discussions. Introduction The increasing amount of data, the decreasing cost of computational power, and the improving state of analytics has revolutionized fields from stock trading to social analytics, but somehow higher education has not received as much attention. The technology that has transformed many for- profit businesses and governments can be applied at various colleges and universities. One obvious place that analytics could be useful is in the classroom, but currently instructors at many universities are using outdated and inefficient methods to grade assignments and compile these scores into self- generated databases. In fact, Darnell West argues that “many of the typical pedagogies provide little immediate feedback to students, require teachers to spend hours grading routine assignments, are not very proactive about showing students how to improve comprehension, and fail to take advantage of digital resources that can improve the learning process” (West 2012). Data mining and analytics provide the capabilities necessary to circumvent the traditionally cumbersome grading processes and glean Fall 2015 64 insights from student data about performance, learning approaches, and other metrics. For example, Leah Macfadyen and Shane Dawson developed an “early warning system” which correctly identified 81% of students who failed an online course by creating a regression model that analyzed such variables as total number of discussion messages posted and total number of assignments completed (Macfadyen and Dawson 2010). Big data analytics within education could also be used to monitor student progression through various course sequences for specific majors, online courses that change activities by measuring everything from individual clicks to aggregate performance and algorithms that suggest courses a student should take by analyzing her past grades in similar courses (Bienkowski etal. 2012). While traditional in-person classrooms may allow for the collection of big data for these applications, Anthony Picciano notes “to move into the more extensive and especially time-sensitive learning analytics applications, it is important that instructional transactions are collected as they occur” (Picciano 2012). This rapid collection of data is most likely to be facilitated by course management/leaming management system architectures and online and blended learning course structures (Worsley 2012). There is little work that has looked at how to use analytics methods outside the classroom to improve the overall educational ecosystem, as well as educational policy. However, insights produced by the previously described learning analytics systems can also be used to inform policy decisions. According to van Bameveld, Arnold, and Campbell (2012), “Like business, higher education is adopting practices to ensure organizational success at all levels by addressing questions about retention, admissions, fund raising, and operational efficiency”. Michael Horn and Katherine Mackey (2011) suggest that education analytics can be used to shift the focus from inputs to outputs when measuring academic institutional success. Instead of using seat-time, faculty-student ratios, and dollars spent as a measure of success, analytics software can provide information on more appropriate metrics such as student performance and retention rates. The biggest obstacles to establishing more such systems are building data sharing networks where these myriad metrics can be aggregated, holistically analyzed, and shared among different institutions (West 2012). A recent paper proposes a model and algorithm that would Washington Academy of Sciences 65 help prospective students make better informed decisions about the best fit and best college eco-system based on their unique personalities and behaviors (Berea et al. 2015). Text mining, social media, or sentiment analysis on the college decision process has generally not been discussed in education analytics literature and therefore presents an interesting opportunity to further advance research in this area. A recent survey by Piper Jaffray found that teens are abandoning Facebook in favor of Instagram; 76% of teens are on Instagram and they are using it to gain an unfiltered look at colleges (Stampler 2015). Data Analysis We collected data for this education analytics project for a period of 4 months, between March 4th and July 1st, 2015. For this collection we used TwEater, an original and proprietary collection tool developed at the University of Maryland (TwEater 2015). Originally, the collection was based on 57 keywords and hashtags, such as: “igotin”, “college”, “campus”, “acceptanceletter”, and many more, and the original data set comprised more than 10 million tweets. Since most of these keywords were not necessarily related to the idea of higher education and college admissions and applications, we selected a list of 25 hashtags pertaining exclusively to college, high school and higher education. Out of these, only 20 rendered more than a tweet, with a minimum of one tweet for the hashtag #choosingacollege and a maximum of 1,153,618 tweets for the hashtag #college followed by 282,139 tweets for the hashtag #campus (see Table 1). On the basis of this collection, we assembled a data set of 1,523,817 tweets where most of them (73%) refer to the general idea of “college”. Many of these tweets are quite general, but some of them focus on specific issues, such as: making college applications friendlier for the LGBT community, businesses supporting campuses, parent-student conflicts in college decision making, and hard college choices between various schools. Text Mining Based on this collection, we built a dictionary of about 470,000 unique words that are specific to the discourse about higher education in the Fall 2015 66 USA. This is a dictionary roughly half the size of the English language (the Oxford English Dictionary has over 600,000 words alone), with the caveat that some of the words in our dictionary are informal or abbreviations or pronouns that may not be currently recognized as being part of the formal English. The most frequent words in the education discourse are “campus” and “college”, but if we leave these obvious terms aside, words such as “highschool”, “acceptance”, “life”, and “met” are highlighted as the most frequent ones that are not directly related to colleges. This gives us an indication that students do talk about college acceptance, life, and college related meetings on Twitter. We also analyzed each of the 20 keywords separately and created a histogram of word frequency for each of the 20 keywords. After “college”, “campus”, “higher education” and “highschool”, the largest corpuses (indicated by the number of tweets) belong to hashtags such as #collegeopportunity, #collegetour and #collegebound. The second most frequent word in most corpuses is “student”. Some interesting words, which are sparse (low frequency) but appear more than once and are associated with the most frequent terms mentioned above, are terms such as: “success”, “community”, “hard”, “chip” and “app”. There is a very large gap between the most frequent words and the second most frequent words (showing the long tail distribution of the words) (see Table 1). Twitter only allows for a fixed number of characters per tweet, therefore we also checked how many unique words are being used in a tweet in our data: #collegedecision, #collegechoice and #backtocollege have the most “rich” tweets (an average of ~7 words per tweet), while #collegeopportunity has the least number of unique words per tweet (an average of 0.2), probably due to a different type of content used in the tweet (z.e., hyperlink or video) (see Table 1). Washington Academy of Sciences 67 Table 1. The summary statistics for higher education Twitter data Number of tweets Highest frequency Second highest frequency Absolute sentiment score Corpus size (no. unique words) Words per tweet Min 6 2 2 -461647 17 0.2072 1st Quart 14.8 10.25 3.75 3.8 91.5 1.2242 Median 198 155.5 50.5 37.5 515 3.1876 Mean 76190.9 20985.15 3399.55 -34069.8 37949.7 3.2889 3 rd Quart 2562.2 838.5 151.25 414.8 2727.2 4.5682 Max 1153618 193802 48383 1945 469748 7.3636 Sentiment Analysis We matched the words in each of the 20 dictionaries with the AFINN standard sentiment dictionary (Nielsen 2011) and calculated the sentiment scores of the tweets in our data (see Figure 1). The AFINN dictionary uses a scale from -5 to +5 to rate the effect of approximately 2000 words. We calculated both the absolute and the weighted scores for each keyword. The absolute scores show that the first largest corpuses (“campus”, “college” and “highschool”) are also strongly negative, while all the rest are positive (with the exception of #backtocollege, where the absolute score is only -1, close to neutral, and #collegedecision, which is 0). However, raw sentiment scores do not take into account the volume of the tweets for each keyword. We therefore examine weighted sentiment scores - on corpus size and on tweet - since the distributions of the corpus sizes and number of tweets are quite skewed. The weighted sentiment scores show that #highschool is the most negative talk on Twitter, while #collegematch and #collegeopportunity are the most positive ones. Fall 2015 68 Absolute and Weighted Sentiment Values for Each Keyword 0) (/) Figure 1. Sentiment values for each keyword. 2.3. ZipPs and power law distributions Zipf s law is a well-known statistical regularity observed in natural language (Zipf 1949) that states that the frequency of any word is inversely proportional with its’ rank in the frequency table. We tested whether the Zipf law holds for each of the 20 corpuses and found that #highschool, #highereducation, and #collegetalk have distributions similar to the Zipf distribution (power of ~ -1), while #backtocollege, #rightcollege, and #collegecompletion show the farthest departures from the Zipf distribution (with power of ~ -0.3) (see Figure 2). Power law distributions in college Twitter talk Log Rank Figure 2. Power law and Zipf distributions of words. Washington Academy of Sciences 69 One way to interpret this result (backed by the size of the corpuses, as well) is that there is more actual discussion involved in general topics about high school and college education as opposed to topics about college completion or matching, where the Twitter activity is more likely to inform with links and other types of information as opposed to offering opinions and personal insights and affect. There is no explanation today for why Zipf s law is characteristic to human language, but some prior research suggests that this distribution is more characteristic to natural language and the human memory of language (Cohen et al. 1997; Piantadosi 2014). Therefore tweets that contain other type of content than words are less likely to exhibit this pattern. Retweets Re tweets in any Twitter data are one way to measure the degree of popularity of certain tweets. In our data the retweets to tweets ratio is quite high. The two keywords with the highest retweet to tweet ratio, “rightcollege” and “collegeopportunity”, had retweeting activity for almost each and every tweet — 0.933 and 0.903 respectively - but this is due to the majority of the tweets with “collegeopportunity” that are initiated by the users of @WhiteHouse and @BarackObama, which are popular and frequently retweeted. Disregarding these outliers, the two keywords with the highest retweet to tweet ratio are “highered” and “collegebound” at 0.484 and 0.468, respectively - almost half of the tweets being retweeted. The high retweet to tweet ratio of “collegebound” provides an interesting insight in the context of this project. It indicates that many high school seniors revert to Twitter to broadcast their accomplishments to friends, who share the congratulatory experience. This conclusion is supported by a reading of the tweets. Many of the keywords with high absolute numbers of tweets also have moderately high retweet to tweet ratios, namely “highschool,” “campus,” and “college” at 0.443, 0.399, and 0.356 respectively. Conclusion Our analysis is constrained to only about 4 months of collection and a short list of keywords. But even so, our findings show the following: there is generally a negative sentiment regarding colleges, campuses, high school Fall 2015 70 and higher education; there is a tension between students and parents with respect to college decisions; campuses and colleges are being judged with respect to their inclusions ( i.e ., LGBT); people are more interested in offering their opinions on general subjects (i.e., “campus”) than on specific ones (i.e., “college tours”, “back to school”). Our current research, although exploratory, points towards a few general conclusions when using social media or Big Data for education research. First, the selection of keywords and hashtags is essential, as these are going to determine the constraints for the data that are going to inform any analysis. Second, while there is considerable discussion on Twitter with respect to higher education, most of this discussion is negative. Third, social media is a great resource of information for education policy, as it gives in real time the opinions of the parents and prospective students when it comes to college applications, college acceptance, or college campuses. Acknowledgements The authors wish to thank Mrs. Elbe Cox for support and partnership in initiating and conducting this research. The work has been entirely supported by vibeffect. References Berea, Anamaria, Maksim Tsvetovat, Nathan Daun-Barnett, Mathew Greenwald, and Elena Cox. 2015. "A new multi-dimensional conceptualization of individual achievement in college." Decision Analytics 2(1): 1-15. Bienkowski, Marie, Mingyu Feng, and Barbara Means. 2012. “Enhancing Teaching and Learning Through Educational Data Mining and Learning Analytics.” United States Department of Education Briefs [https://tech.ed.gov/wp- content/uploads/2014/03/edm-la-brief.pdf (accessed 6 September 2015)] Cohen, Albert, Rosario, N. Mantegna, and Shlomo. Havlin. 1997. “Numerical analysis of word frequencies in artificial and natural language texts”. Fractals, 5(01): 95-104. Horn, Michael B. and Katherine Mackey. 2011. “Moving from Inputs to Outputs to Outcomes.” Innosight Institute, [http://www.christenseninstitute.org/wp- Washington Academy of Sciences 71 content/uploads/20 13/04/Moving-from-Inputs-to-Outpuls-to-Outcomes. pdf (accessed 7 September 2015)] Macfadyen, Leah P. and Shane Dawson. 2010. “Mining LMS data to develop an ‘early warning system’ for educators: A proof of concept.” Computers & Education 54(2): 588-599. Nielsen, Fenn. 2011. “a new ANEW: Evaluation of a word list for sentiment analysis in microblogs” in Proceedings of the ESWC201 1 Workshop on 'Making Sense of Microposts': Big things come in small packages 718 in {CEUR} Workshop Proceedings: 93-98. [http://arxiv.Org/abs/l 103.2903 (accessed 7 September 2015)] Oxford English Dictionary. 2015. [http://public.oed.com/about/, (accessed 7 September 2015)] Piantadosi, Steven T. 2014. “Zipf s word frequency law in natural language: A critical review and future directions”. Psychonomic bulletin & review, 21(5): 1112-1130. Picciano, Anthony G. 2012. “The Evolution of Big Data and Learning Analytics in American Higher Education.” Journal of Asynchronous Learning Networks 16(3): 9- 20. Oxford English Dictionary. 2015. [http://public.oed.com/about/, (accessed 7 September 2015)] Rand, William, David Damron, and Radu Machedon, 2015. Tw Eater. [https://github.coin/CenterForComplexitvInBusiness/ (accessed 7 September 2015)] Stampler, Laura. 2015. “How High School Students Use Instagram to Help Pick a College”, Time Magazine, [http://time.com/3762067/college-acceptance-instagram- high-school/ (accessed 7 September 2015)] Van Bameveld, Angela, Kimberly E. Arnold, and John P. Campbell. 2012. “Analytics in Higher Education: Establishing a Common Language.” Educause Learning Initiative : 2-11. West, Darrell M. 2012. “Big Data for Education: Data Mining, Data Analytics, and Web Dashboards.” Governance Studies at Brookings : 1-10. Worsley, Marcelo. 2012. “Multimodal learning analytics: enabling the future of learning through multimodal data analysis and interfaces.” in Proceedings of the 14th ACM international conference on Multimodal interaction (ICMI '12). ACM, New York, NY, 353-356. DOI=http://dx.doi.org/l 0.1 145/2388676.2388755 Zipf, George K. 1949. Human Behavior and the Principle of Least Effort. Cambridge, Massachusetts: Addison-Wesley. Fall 2015 72 Bios Anamaria Berea is a postdoctoral researcher in the Center for Complexity in Business, Robert H. Smith School of Business, University of Maryland. She is researching social phenomena using various computational methods. William Rand is an assistant professor of Marketing and Computer Science at the University of Maryland. William Rand serves as the director at the Center for Complexity in Business. His work examines the use of computational modeling techniques, like agent-based modeling, geographic information systems, social network analysis, and machine learning to help understand and analyze complex systems, such as the diffusion of innovation, organizational learning, and economic markets. Kevin Wittmer is an undergraduate researcher in the Robert H. Smith School of Business, University of Maryland. He is researching various aspects of qualitative and quantitative methods for data analysis. Gerard Wall is the Solutions Architect at vibeffect, pioneering the investigation of how the Higher Education decision and its impact on families can become more transparent and relevant for the “consumer' as family. Washington Academy of Sciences 73 Privacy in a Networked World: New Challenges and Opportunities for Privacy Research Heng Xu and Haiyan Jia The Pennsylvania State University Abstract In this article, we describe the new threats to information privacy that appear as the result of the emerging Big Data practices and methodologies in today’s networked world. In particular, the collection and analysis of large-scale data from social networking sites challenge the traditional conceptualization of privacy. In response, a new conceptual framework is proposed to encompass three key dimensions of privacy in the Big Data context: information identifiability, information ephemerality, and information linkability. Introduction The “privacy as a right” perspective, first introduced by Warren and Brandeis (1890), has since influenced numerous opinions and court cases on privacy and law enforcement (searches and seizures), privacy and self (abortions and embryos), privacy and the press (private facts exposure and celebrity privacy), privacy in the workplace (psychological testing and lifestyle monitoring), etc. (Alderman and Kennedy, 1997). However, these issues were just a subset of privacy issues which Warren and Brandeis were concerned about when they wrote the “right to privacy.” Their main concern was with the advent of technological developments (instant photography and audio recordings in the late nineteenth century) that were increasingly revealing personal information without individuals’ awareness. Such privacy concerns still exist and remain highly relevant after 125 years. In today’s Big Data Era, many data collectors, data brokers, aggregation services and various companies collect and use personal data without individuals’ awareness, which leads to a dark data ecosystem. It has been estimated that there are 4000 separate companies involved in the dark data market and many dark data brokers make the data available to any buyer willing to pay (Levine 2013). Fall 2015 74 The emerging field of big data analytics is distinguished from traditional data analytics by its three key characteristics (McAfee et al. 2012): (1) Volume - Big Data analysis works with petabytes of data in a single dataset; (2) Velocity- Real-time or nearly real-time information is aggregated and analyzed for agile decision-making; and (3) Variety - Big Data takes all forms of information ranging from sensor readings and GPS signals to messages, updates, and images posted on Social Networking Sites (SNSs). Collecting large amounts of data, especially personal and social data, brings both opportunities and challenges. While many practitioners believe that the rise of Big Data has potential for creating better tools and services, scholars ( e.g ., boyd & Crawford 2012) have already warned about how poor execution may lead to negative social and economic consequences such as intrusion to personal privacy, suppression to speech, and misleading predictions, among many others. Interaction between technological innovations and social ecology usually has consequences far beyond the immediate purposes of the technical devices and practices (Kranzberg 1986). One of the major threats that Big Data analytics posits is privacy, as it seeks to identify at the expense of individual and collective identity (Richards & King 2013). Viewing Big Data as a public good, Acquisti (2014) discusses its critical importance for public decision-making, and how it can reduce inefficiencies and increase welfare when used properly. However, Acquisti (2014) also questions who should bear the economic cost of Big Data practices that use personal information: data subjects (whose data are aggregated and analyzed), data holders (who collect and handle consumer data), or both? To face the increasing costs associated with data storage and analysis, data aggregators, and data holders typically assume that they have rights to the data and exploit user data for profit, overriding the interests of individuals in their privacy and leaving them few mitigating measures (Wigan & Clarke 2013). Big Data practices as such pose significant threats to individual privacy. This paper aims at discussing the following challenges to information privacy with the emergence of Big Data: (1) What are the unique threats of Big Data practices to information privacy? (2) How do these unique threats challenge the conceptualization of privacy? (3) How should we address privacy challenges in today’s networked world? Washington Academy of Sciences 75 Redefining Privacy in a Big Data Context Information Identifiability In today’s networked world, privacy is often shaped and enabled by various features of technologies. For instance, websites often require users to disclose certain types of information in order to obtain the services, and provide certain mechanisms and tools for users to manage their privacy preferences. Personal information, such as name, location, personal interests, and even information of one’s social networks, can be revealed voluntarily by the users to socialize and establish social connections. However, disclosure of private information can have significant consequences, and thus trigger users’ privacy concerns and shape their privacy management behaviors (Fogel & Nehmad 2009). For instance, popular SNSs such as Facebook and Twitter require various levels of information disclosure and information accuracy by design. Facebook requires users to provide real names and work/education email addresses to be added to an affiliation or network. Twitter, on the other hand, does not necessarily require real names, but it sets users’ profiles as public by default, potentially exposing a large amount of personal information to the wide audience and other third parties. Further, many social networking and mobile applications monitor, record and even publish users’ location information, which is susceptible to unauthorized disclosure. In the context of Big Data, we argue that one fundamental dimension of information privacy is information identifiability , which is defined as the amount and the accuracy of personally identifiable information being revealed. Unique to Big Data practices, individuals’ identities can be easily identified or re-identified. For instance, Narayanan & Shmatikov (2009) have demonstrated how to efficiently de-anonymize a large number of Twitter and Flickr users by simply using data of username, location, and “follow” or “contact” relationships. Data mining using vast amounts of identifiable information generate hypotheses and discover general patterns that could actually be stereotypical and misleading, possibly causing both privacy loss and economic loss for data subjects, and posing privacy threats that the existing privacy laws are far behind to define or protect (Brankovic & Estivill-Castro 1999). What is more risky is that Big Data analytics can now gather and extract implicit user data and across different social networking platforms. Beyond user specified data such as usernames and Fall 2015 76 locations, SNS services now can automatically generate user information through mechanisms such as face recognition, geo-tagging, and multi-site uploading, further increasing the amount and the accuracy of personal information (Smith et al. 2012). These data are usually extracted from the uploaded files or generated through metadata. Users are often unaware that these data are stored and can be used for identification. Thus, information identifiability is one key dimension of privacy, and has extensive new meanings in the Big Data context. Big Data tools significantly increase the potential to identify individual users through social data and reveal more user information in increasing quantities and accuracy. The lack of user awareness and regulatory mechanisms to control such information revelation signifies its impact on information privacy. Information Ephemerality However, information identifiability does not fully capture the scope of information privacy, especially in the Big Data era. Palen and Dourish (2003) argue that privacy is not only about the identity boundaries defining self versus others, but also the temporal boundaries between past, present and future. Events of information disclosure are not isolated, but sequentially connected. Therefore, information disclosed at a specific instance becomes contextualized and interpreted in relation to other events and situations, if the latter are available. In our daily life, information tends to be ephemeral; the information that we share and exchange is constrained to a certain physical location and a certain time period before it gets forgotten. While we constantly observe the action of forgetting in our social life (Mayer-Schonberger 2009) and in social norms and policy (Blanchette & Johnson 2002), recent advances in information technologies have offered inexpensive, large-volume digital data storage capacity, making the persistence of information the odd commonplace (Ambrose 2012). The extended information lifespan has significant privacy implications, as the preservation of personal information amplifies and prolongs the effect of any privacy loss. The persistence or the ephemerality of information has not been a major privacy concern in the past few decades, but more recently, the current and new consensus of privacy threat is formed around the fact that information, once online, is there forever. This new realization has brought attention to this new aspect of privacy — “the right to be forgotten” Washington Academy of Sciences 77 (Ambrose 2012). We define information ephemerality as the duration of private information being available, accessible and stable, an increasing significant dimension of information privacy in close relation to big social data. While the traditional forms of data analytic tools may not be able to handle large-scale longitudinal data, Big Data technologies, in particular, can use the persisting records of social data, sometimes beyond a single SNS platform, and change the availability and accessibility of information from the “here and now” to the “everywhere and forever” (Grudin 2002). The accumulated user data on Facebook alone have been used to reveal the evolution of user interactions over three years (Viswanath et al. 2009), the longitudinal changes in privacy and disclosure behaviors in six years (Stutzman et al. 2012), as well as the year-long variation of national happiness levels (Facebook 2010). However, changing the ephemeral nature of information and making longitudinal analysis of such big social data can be damaging. When modeling large datasets over time, many time-sensitive factors may come into play to influence outcomes. Without considering these factors and changes over the course of time, data will be taken out of context, often lose meaning and value, and be interpreted in misleading ways. For instance, boyd and Crawford (2012) point out that the types of social networks derived from mining a longitudinal dataset — “articulated networks” (networks resulted from people specifying contacts through mechanisms such as friend lists or instant messenger lists) and “behavioral networks” (networks derived from communication patterns such as email exchanges and Facebook photo-tagging) — tend to be inequivalent to true personal networks. “False discoveries” like this made out of the large-scale social data not only breach personal privacy, but may have severe real-world consequences affecting the products, bank loans, and health insurance a person receives. Information persistence is a unique “big social data” threat to users’ information privacy. Because the real world is one that is ephemeral rather than permanent, individuals apply the same kind of expectations to their online disclosure, expecting the information that they share online will not be everlasting (Shein 2013). Big Data tools not only serve to document and store longitudinal redords of private information, but also use and analyze them for inferences, knowledge, and trends regarding users’ behavioral Fall 2015 78 intentions and social implications, significantly challenging users’ privacy expectations and violating their privacy rules. Information Linkability User data on SNSs concern not only information about the users themselves, but also information about the users’ colleagues, friends, and others they come into contact with. As SNSs facilitate connectedness across boundaries and in dynamic ways, neither a one-time snapshot nor an over-time trace of a single user’s profile can fully capture the complexity of SNS data (boyd & Crawford 2012). Unique to the social data generated and accumulated online, information privacy is dependent not on one single user, but on a web of users to whom this individual is connected and on the information that they disclose. Xu (2012) proposes the notion of privacy 2.0, describing this phenomenon that information disclosure is co-constructed by users and their social connections, which demands the responsibilities of privacy protection to be distributed through their social networks. Following Xu (2012), we suggest that information linkability as the third key dimension of information privacy in the Big Data context, and define it as the degree to which information is relational and linked through social connections. As privacy scholars (Lampinen et al. 201 1; de Wolf et al. 2014) have recently observed, the connected nature of SNS data and the interpersonal nature of information sharing have made individualistic privacy protection strategies inadequate. Even if a user adopts tight privacy settings, his or her personal information could still be accessed or misused by their friends’ ignorance of privacy and security (Xu, 2012). As a result of such information linkability, SNS data are often gathered and exploited without the consent of the individuals to whom the data relate, and individuals who volunteer such data only have moral responsibility for their actions (Wigan & Clarke 2013). Privacy risks in relation to information linkability become an especially prominent problem with Big Data practices. Many analytic tools are specifically designed for social network analysis to draw patterns and insights from cliques, groups, and even large social networks (Davenport et al. 2013). To address this emerging privacy issue, Troshynski et al. (2008) argue that users, researchers, and practitioners of big social data consider Washington Academy of Sciences 79 not only personal privacy implications but accountability, a broader concept that encompasses the privacy considerations in a multi-directional relationship - accountability for others’ personal privacy, accountability to superiors, colleagues, and to the public. To achieve accountability, considerations need to be given to the control of and power regarding the access and use of linked and relational information and the differences between accessibility and publicness (boyd & Crawford 2012). The Emerging Field of Human-Data Interaction (HDI) The technological advancement on machine learning and automated content analysis continues to improve the strength of today’s big data ecosystem. To address privacy concerns, many privacy scholars suggest that individuals’ awareness of privacy should be enhanced by providing information transparency about what data is collected, how it is used, and whom it is shared with (Wang et al. 2013; Xu et al. 2012). In the privacy literature, researchers have examined multiple ways of enhancing transparency, such as providing explicit textual privacy statements (Pollach 2006), presenting privacy facts in the form of nutrition labels (Kelley et al. 2009), using warning icons to suggest suspicious data use (Lin et al. 2012), and using justification messages to explain information disclosure (Knijnenburg and Kobsa 2013). Transparency is also at heart of existing and proposed regulatory schemes. For instance, the U.S. Consumer Privacy Bill of Rights suggests that “companies should provide clear descriptions of [...] why they need the data, how they will use it” (White House 2012). While empowering individuals with privacy comprehensiveness is a desirable approach to raise awareness, information transparency itself cannot guarantee privacy. If implemented inappropriately, the strategies can even backfire. For instance, practices to enhance information transparency have been criticized for i) burdening users’ cognitive load by having users process long and ambiguous statements, and ii) leading to a “context collapse” where users lack contextual explanations and justifications to aid their real-time privacy decision making (Vitak 2012). Therefore, in raising users’ privacy awareness, it becomes imperative to find an effective way to present and implement transparency. In this article, we argue that privacy researchers who are interested in addressing Big Data privacy challenges are likely to benefit from the Fall 2015 80 emerging field of Human-Data Interaction (HDI) (Mortier et al. 2014). HDI emphasizes on creating a collaborative but sometimes combative data ecosystem around multiple stakeholders engaging in the collection and use of personal data. The HDI approach does not throw out transparency entirely, but gently refocuses this paradigm onto individuals’ privacy awareness that would enable legibility , agency and negotiability. Legibility is concerned with making data space (from collection, use, analysis, to retention) both transparent and comprehensible (Mortier et al. 2014). To achieve the goal of legibility, researchers need to create innovative mechanisms to visualize: who has collected what private data; how the private data are being processed; how their private data are mingled with others’ private data; what is done by the data brokers; and who are using their private data and how. We argue that legibility empowerment is a precursor to an individual’s ability to exercise agency in situations where personal data are being collected and used. Agency is concerned with giving people the capacity to act within the data ecosystem. Consistent with Mortier and associates (2014), we do not believe that all individuals should continually exercise this capacity; but some of them can have the agency whenever they wish to. Thus privacy researchers and technologists need to operationalize agency through intelligent personalized approach by providing individuals with the customized option of expressing concern over certain data use which they do not agree with. Negotiability is concerned with many dynamic relationships that arise around data and data processing (Mortier et al. 2014). This theme requires collaboration and engagement with stakeholders to collaboratively decide what and why data exchanges occur, as well as specify the information flow “from whom,” “to whom,” “for what reasons,” and “under what conditions.” Placing discussions on negotiability empowerment within the relevant contexts does not suggest that there is an agreement about the level of privacy that is appropriate in any given context. However, knowing the relevant dimensions and stakeholders of information flow in the specific context does clarify the discussion. Washington Academy of Sciences 81 Conclusion This article suggests a new approach to conceptualizing privacy with emphasis on emerging privacy threats in terms of information identifiability, information ephemerality, and information linkability. The latter two, in particular, are of growing importance, and pose significant and unique threats to information privacy with the emergence and widespread of Big Data technologies. The reconceptualization of information privacy in these three dimensions provides a unique opportunity for the emerging field of Human-Data Interaction (HDI). It delineates three mechanisms through which big social data analysis may influence users, and serves as a theoretical foundation for future user-centered studies of privacy concerns and privacy decision-making concerning Big Data practices and products. This conceptual framework can further guide privacy research and ethics discussions to draw economic, social and legislative implications of Big Data practices, as well as finding practical solutions to these three privacy challenges. Acknowledgments The authors gratefully acknowledge the financial support of the U.S. National Science Foundation under grant CNS-0953749. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the U.S. National Science Foundation. References Acquisti, A. (2014). “The economics and behavioral economics of privacy”, in Privacy, Big Data, and the Public Good: Frameworks for Engagement , :J. Lane, V. Stodden, S. Bender, and H. Nissenbaum (eds) Cambridge: Cambridge University Press, 98-1 12. Aldennan, E., and Kennedy, C. (1997). The Right to Privacy , New York: Vintage Books. Ambrose, M. L. (2012). “It’s about time: Privacy, information lifecycles, and the right to be forgotten”. Stanford Technology Law Review, 16, 101-154. Fall 2015 82 Blanchette, J. F., and Johnson, D. G. (2002). Data retention and the panoptic society: The social benefits of forgetfulness. The Information Society, 18(1), 33-45. Brankovic, L., and Estivill-Castro, V. (1999). Privacy issues in knowledge discovery and data mining. In Proceedings of Australian Institute of Computer Ethics Conference (pp. 89-99). boyd, d., and Crawford, K. (2012). Critical questions for big data: Provocations for a cultural, technological, and scholarly phenomenon. Information, Communication & Society, 15(5), 662-679. boyd, d., and Ellison, N. B. (2007). Social network sites: Definition, history, and scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230. Davenport, T. H., Barth, P., and Bean, R. (2013). How “Big Data” is different. MIT Sloan Management Review, 54(1), 43-46. de Wolf, R., Willaert, K., and Pierson, J. (2014). Managing privacy boundaries together: Exploring individual and group privacy management strategies in Facebook. Computers in Human Behavior, 35, 444-454. Facebook. (2010). How happy are we? Accessed 30 October 2015: https://www.facebook.com/notes/facebook/how-happy-are-we/ 150162112130. Fogel, J., and[https://cups. cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf (accessed 30 October 2014)] Nehmad, E. (2009). Internet social network communities: Risk taking, trust, and privacy concerns. Computers in Human Behavior, 25(1), 153-160. Grudin, J. (2002). Group dynamics and ubiquitous computing. Communications of the ACM, 45(12), 74-78. ’ Kelley, P. G., Bresee, J., Cranor, L. F., and Reeder, R. W. (2009). A nutrition label for privacy. In Proceedings of the 5 th Symposium on Usable Privacy and Security. [https://cups.cs.cmu.edu/soups/2009/proceedings/a4-kelley.pdf (accessed 30 October 2014)] Knijnenburg, B. P. and Kobsa, A. (2013). Making decisions about privacy: information disclosure in context-aware recommender systems. ACM Transactions on Interactive Intelligent Systems, 3 (3), 20. Kranzberg, M. (1986). “Technology and History: Kranzberg's Laws.” Technology’ and Culture 27(3): 544-560. Lin, J., Amini, S., Hong, J. I., Sadeh, N., Lindqvist, J., and Zhang, J. (2012). Expectation and purpose: understanding users' mental models of mobile app privacy through crowdsourcing. In Proceedings of the 2012 ACM Conference on Ubiquitous Computing, (pp. 501-510). Lampinen, A., Lehtinen, V., Lehmuskallio, A., and Tamminen, S. (201 1). “We're in it together: interpersonal management of disclosure in social network services”. In Proceedings of the SIGCHl Conference on Human Factors in Computing Systems (pp. 3217-3226). Washington Academy of Sciences 83 Levine, Y. (2013). “What surveillance Valley knows about you”. http://pando.com/2013/12/22/a-peek-into-surveillance-valley/ (accessed 30 October 2015) Mayer-Schonberger, V. (2009). Delete: The Virtue of Forgetting in the Digital Age. Princeton, N.J.: Princeton University Press. McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D. J., and Barton, D. (2012). Big data: The management revolution. Harvard Business Review , 90(10), 61-67. Mortier, R., Haddadi, H., Henderson, T., McAuley, D., and Crowcroft, J. (2014). Human-Data Interaction: The Human Face of the Data-Driven Society. Available at SSRN: http://ssrn.com/abstract=250805 1 (accessed 30 October 2015) Narayanan, A., and Shmatikov, V. (2009). “De-anonymizing social networks”. In Proceedings of 30th IEEE Symposium on Security and Privacy, (pp. 173-187). Palen, L., and Dourish, P. (2003).” Unpacking privacy for a networked world”. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 129-136). Pollach, I. (2006). “Privacy statements as a means of uncertainty reduction in WWW interactions”. Journal of Organizational and End User Computing (JOEUC), 18(1), 23-49. Richards, Neil M., and Jonathan H. King. (20 1 3). “Three Paradoxes of Big Data.” Stanford Law Review Online 66 (41). [http://www.stanfordlawreview.org/online/privacy-and-big-data/three-paradoxes-big -data (accessed 30 October 2015] Shein, E. (2013). “Ephemeral data”. Communications of the ACM, 56(9), 20-22. Smith, M., Szongott, C., Henne, B., and von Voigt, G. (2012). “Big data privacy issues in public social media”. In Proceedings of 6th IEEE International Conference on Digital Ecosystems Technologies (DEST), (pp. 1-6). Stutzman, F., Gross, R., and Acquisti, A. (2012). “Silent listeners: The evolution of privacy and disclosure on facebook”. Journal of Privacy and Confidentiality, 4(2), 7-41. Troshynski, E., Lee, C., and Dourish, P. (2008). “Accountabilities of presence: refraining location-based systems”. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 487-496). Viswanath, B., Mislove, A., Cha, M.,and Gummadi, K. P. (2009).” On the evolution of user interaction in facebook”. In Proceedings of the 2nd ACM workshop on Online Social Networks (pp. 37-42). Vitak, J. (2012). The impact of context collapse and privacy on social network site disclosures. Journal of Broadcasting & Electronic Media, 56, 4 (2012), 451-470. Warren, S. D., and Brandeis, L. D. (1890). “The right to privacy”. Harvard Law Review, 4(5), 193-220. Wigan, M. R., and Clarke, R. (2013). “Big data's big unintended consequences”. Computer, 46(6), 46-53. Fall 2015 84 White House. (2012). Consumer data privacy in a networked world: A framework for protecting privacy and promoting innovation in the global digital economy. Washington, DC: White House. Wang, N., Grossklags, J., and Xu, H. (2013) “An online experiment of privacy authorization dialogues for social applications”. In Proceedings of the 2013 conference on Computer supported cooperative work. (pp. 26 1 -272). Xu, H., Teo, H.-H., Tan, B. C., and Agarwal, R. (2012).” Effects of Individual Self-Protection, Industry Self-Regulation, and Government Regulation on Privacy Concerns: A Study of Location-Based Service”s. Information Systems Research , 23(4), 342-1363. Xu, H. (2012). “Reframing Privacy 2.0 in Online Social Network”. University of Pennsylvania Journal of Constitutional Law , 14(4), 1077-1 102. Bios Heng Xu is an associate professor of Information Sciences and Technology at the Pennsylvania State University. Her current research focus is on the interplay between social and technological issues associated with information privacy. She has authored or coauthored over 100 research papers on information privacy, security management, human-computer interaction, and technology innovation adoption. Haiyan Jia is a post-doctoral scholar at the Penn State University in the College of Information Sciences and Technology. Her research interest primarily focuses on the social and psychological effects of communication technology ranging from Web to mobile apps to smart objects. Her current work investigates online privacy in social and collective contexts. Washington Academy of Sciences 85 Washington Academy of Sciences 1200 New York Avenue, NW Room 1 1 3 Washington, DC 20005 Membership Application Please fill in the blanks and send your application to the Washington Academy of Sciences at the address above. We will contact you as soon as your application has been reviewed by the Membership Committee. Thank you for your interest in the Washington Academy of Sciences. (Dr. Mrs. Mr. Ms.) Business Address Home Address Email Phone Cell Please indicate: Preferred mailing address Type of membership Business Regular Home Student Schools of Higher Education Attended Degree Dates Present Occupation or Professional Position Please list memberships in scientific societies - and include office held: Fall 2015 86 Journal of the Washington Academy of Sciences Instructions to Authors The Washington Academy of Sciences publishes its interdisciplinary peer-reviewed Journal of the Washington Academy of Sciences four times a year — Spring, Summer, Fall, and Winter. 1 . Deadlines for quarterly submissions are: Spring - February 1 Fall - August 1 Summer - May 1 Winter - November 1 2. Draft manuscripts using a word processing program (such as MSWord), not PDF. 3. Papers should be 6,000 words or fewer. With 7 or more graphics, reduce the number of words by 500 for each graphic. 4. Include an abstract of 150-200 words. 5. Graphics must be in black & white or greytone. They must be referenced in the text. 6. Use endnotes, not footnotes. The bibliography may be in a style considered standard for the discipline or professional field represented by the paper. 7. Submit papers as email attachments to the editor. 8. Include the author’s name, affiliation, and contact information — including postal address. Membership in an Academy-affiliated society may also be noted. 9. Manuscripts are peer reviewed and become the property of the Washington Academy of Sciences. 10. There are no page charges. Please see the Academy’s web site, www.washacadsci.org, for the library subscription rate, listing of articles dating to 1899, and information on accessing them. Washington Academy of Sciences 87 Washington Academy of Sciences Affiliated Institutions National Institute for Standards & Technology (NIST) Meadowlark Botanical Gardens The John W. Kluge Center of the Library of Congress Potomac Overlook Regional Park Koshland Science Museum American Registry of Pathology Living Oceans Foundation National Rural Electric Cooperative Association (NRECA) Fall 2015 88 Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies Acoustical Society of America, Washington Chapter American Assoc, of Physics Teachers, Chesapeake Section American Astronomical Society American Ceramics Society American Fisheries Society American Institute of Aeronautics and Astronautics American Meteorological Society, Washington, DC Chapter American Nuclear Society, Washington DC Section American Phytopathological Society, Potomac Division American Society for Cybernetics American Society for Metals, Washington Chapter American Society of Civil Engineers, National Capital Section American Society of Mechanical Engineers, Washington Section American Society of Microbiology, Washington Branch American Society of Plant Biologists, Mid-Atlantic Anthropological Society of Washington ASM International Association for Women in Science, DC Metropolitan Chapter Association for Computing Machinery, DC Area Chapter Association for Science, Technology, and Innovation Association of Information Technology Professionals Biological Society of Washington Botanical Society of Washington Capital Area Food Protection Association Chemical Society of Washington District of Columbia Institute of Chemists District of Columbia Psychological Association Eastern Sociological Society Electrochemical Society, National Capital Section Entomological Society of Washington Geological Society of Washington Historical Society of Washington DC Human Factors and Ergonomics Society, Potomac Chapter Institute of Electrical and Electronics Engineers, Northern Virginia Section (continued on next page) Paul Arveson Frank R. Haig, S. J. Sethanne Howard Vacant Lee Benaka David W. Brandt Vacant Charles Martin Vacant Stuart Umpleby Vacant Vacant Daniel J. Vavrick Vacant Mark Holland Vacant Toni Marechaux Jodi Wesemann Alan Ford F. Douglas Witherspoon Chuck Lowe Stephen Gardiner Chris Puttock Keith Lempel Elise Ann Brown Vacant Tony Jimenez Ronald W. Mandersheid Vacant Vacant Jeffrey B. Plescia Jurate Landwehr Vacant Gerald P. Krueger Murty Polavarapu Washington Academy of Sciences Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies (continued from previous page) Institute of Electrical and Electronics Engineers, Washington Section Institute of Food Technologies, Washington DC Section Institute of Industrial Engineers, National Capital Chapter International Association for Dental Research, American Section International Society for the Systems Sciences International Society of Automation, Baltimore Washington Section Instrument Society of America Marine Technology Society Maryland Native Plant Society Mathematical Association of America, Maryland-District of Columbia- Virginia Section Medical Society of the District of Columbia National Capital Area Skeptics National Capital Astronomers National Geographic Society Optical Society of America, National Capital Section Pest Science Society of America Philosophical Society of Washington Society for Experimental Biology and Medicine Society of American Foresters, National Capital Society Society of American Military Engineers, Washington DC Post Society of Manufacturing Engineers, Washington DC Chapter Society of Mining, Metallurgy, and Exploration, Inc., Washington DC Section Soil and Water Conservation Society, National Capital Chapter Technology Transfer Society, Washington Area Chapter Virginia Native Plant Society, Potowmack Chapter Washington DC Chapter of the Institute for Operations Research and the Management Sciences (WINFORMS) Washington Evolutionary Systems Society Washington History of Science Club Washington Paint Technology Group Washington Society of Engineers Washington Society for the History of Medicine Washington Statistical Society World Future Society, National Capital Region Chapter Richard Hill Vacant Neal F. Schmeidler J. Terrell Hoffeld Vacant Vacant Hank Hegner Jake Sobin Vacant D. S. Joseph Vacant Vacant Jay H. Miller Vacant James Cole Vacant Eugenie Mielczarek Vacant Daina Apple Vacant Vacant E. Lee Bray Terrell Erickson Richard Leshuk Vacant Russell Wooten Vacant Albert G. Gluckman Vacant Alvin Reiner Alain Touwaide Michael P. Cohen Jim Honig Washington Academy of Sciences Room 113 1200 New York Ave. NW Washington, DC 20005 Return Postage Guaranteed NONPROFIT ORG US POSTAGE PAID MERRIFIELDVA 22081 PERMIT# 888 4*8******”*1 1 9***«************AUT0**IVIIXED ADC 207 SERIAL RECORDS DIVISION ERNST MAYR LIBRAY MUSEUM COMP ZOOLOGY HARVARD UNIVERSITY 26 OXFORD ST CAMBRIDGE MA 02138-2902 WAS 833^ Volume 101 Number 4 Winter 2015 Journal of the WASHINGTON ACADEMY OF SCIENCES Board of Discipline Editors iv JSTOR 1 Tribute to Julius "Jay" Earl Uhlaner 6. Kruger. 3 A Study of the Primary Granitoid Outcroppings and Sedimentary Rocks 7 N. Bassanganam, Yang Mei Zhen, Prince E. Y. Danguene, M. Wang Candidates for the Source of the 1977 "WOW" Signal A. Paris, E. Davies 25 Affine Geometry, Planck Length and Cosmic Acceleration G. L Murphy 33 2015 membership list 45 Membership Application 53 Instruction to Authors 54 Affiliated Institutions 55 Affiliated Societies and Delegates 56 ISSN 0043-0439 Issued Quarterly at Washington DC MCZ LIBRARY FEB 1 6 2018 HARVARD UNIVERSITY Washington Academy of Sciences Founded in 1898 Board of Managers Elected Officers President Mina Izadjoo President Elect Mike Coble Treasurer Ronald Hietala Secretary John Kaufhold Vice President, Administration Terry Longstreth Vice President, Membership Sue Cross Vice President, Junior Academy Vice President, Affiliated Societies Gene Williams Members at Large Paul Arveson Michael Cohen Frank Haig, S.J. Neal Schmeidler Mary Snieckus Past President Terrell Erickson Affiliated Society Delegates Shown on back cover Editor of the Journal Sethanne Howard Journal of the Washington Academy of Sciences (ISSN 0043-0439) Published by the Washington Academy of Sciences email: wasiournal(5)washacadsci.orq website: www.washacadsci.org The Journal of the Washington Academy of Sciences The Journal is the official organ of the Academy. It publishes articles on science policy, the history of science, critical reviews, original science research, proceedings of scholarly meetings of its Affiliated Societies, and other items of interest to its members. It is published quarterly. The last issue of the year contains a directory of the current membership of the Academy. Subscription Rates Members, fellows, and life members in good standing receive the Journal free of charge. Subscriptions are available on a calendar year basis, payable in advance. Payment must be made in US currency at the following rates. US and Canada $30.00 Other Countries $35.00 Single Copies (when available) $1 5.00 Claims for Missing Issues Claims must be received within 65 days of mailing. Claims will not be allowed if non- delivery was the result of failure to notify the Academy of a change of address. Notification of Change of Address Address changes should be sent promptly to the Academy Office. Notification should contain both old and new addresses and zip codes. POSTMASTER: Send address changes to WAS, Rm 113, 1200 New York Ave. NW Washington, DC 20005 Academy Office Washington Academy of Sciences Room 113 1200 New York Ave. NW Washington, DC 20005 Phone:(202) 326-8975 1 200 New York Ave. Suite 113 Washington DC 20005 wwwwashacadsci.org Journal of the WASHINGTON ACADEMY OF SCIENCES Volume 101 Number 4 Winter 2015 Contents Editorial Remarks S. Howard ii Board of Discipline Editors iii JSTOR 1 Tribute to Julius “Jay” Earl Uhlaner G. Kruger 3 A Study of the Primary Granitoid Outcroppings and Sedimentary Rocks 7 N. Bassanganam, Yang Mei Zhen, Prince E. Y. Danguene, M. Wang Candidates for the Source of the 1977 “WOW” Signal A. Paris, E. Davies 25 Affine Geometry, Planck Length and Cosmic Acceleration G. L. Murphy 33 Membership List for 2015 45 Membership List 53 Instructions to Authors 54 Affiliated Institutions 55 Affiliated Societies and Delegates 56 ISSN 0043-0439 Issued Quarterly at Washington DC MCZ LIBRARY t i A K V ARD UNIVERSITY Winter 2015 ii Editorial Comments This Winter issue of the Journal has some interesting, eclectic things to share. The Academy has signed a contract with JSTOR - a company that digitizes science journals for online access. We start with a description of that process. Eventually they will have digitized all our back issues. Following that is a tribute to WAS Fellow Jay Uhlaner reprinted with permission from the Human Factors & Ergonomics Society Bulletin and written by G. Kruger. Next is a graduate student paper that describes geological studies done in the Central African Republic around the region of Boali. Noted for its waterfalls Les Chutes de la Mbi is a 656-foot cascade where the Upper M’poko River meets the Oubangui River. The natural beauty of the site has earned it a place on the tentative UNESCO World Fleritage Site list. The Falls of Boali are 250 m wide and 50 m high, and are a popular tourist destination. I do not usually put photos in the editorial comments, but this is an exception. The Falls of Boali are shown below. Washington Academy of Sciences iii This is the first paper we have received from there. The author is a student at the China University of Geosciences in Wuhan China. He speaks French and Chinese but not English, so it was a challenge to get the paper first into English and second into the structure of a technical and publishable paper. It took several months of work, and I am proud to present the work for his Master’s thesis. Then comes a paper on the “wow” signal. This one has a back story. In 1977 a radio telescope in Ohio received an intense, short signal. At that time data were recorded on a chart recorder - a bit like a lie detector setup. A sheet of paper rolls out with a moving pen recording the data. The signal the telescope received was so large that the telescope operator wrote “wow” on the paper - hence the name the “wow” signal. Talk to any radio astronomer, mention the “wow” signal, and they will know the story. To date it has not been explained. This paper offers a possibility for the signal. Last up is a paper that discusses the metric based general theory of relativity and dark energy. The author provides a view of the problem of dark energy based on Schrodinger’s affine field theory. Put on your serious math hats for this one. It may explain that odd “cosmological constant”, A. As usual the Winter issue ends with a list of members of the Academy. Editor Sethanne Howard Winter 2015 IV Journal of the Washington Academy of Sciences Editor Sethanne Howard sethanneh@msn.com Board of Discipline Editors The Journal of the Washington Academy of Sciences has a 12-member Board of Discipline Editors representing many scientific and technical fields. The members of the Board of Discipline Editors are affiliated with a variety of scientific institutions in the Washington area and beyond — government agencies such as the National Institute of Standards and Technology (NIST); universities such as Georgetown; and professional associations such as the Institute of Electrical and Electronics Engineers (IEEE). Anthropology Emanuela Appetiti eappetitiOTotmail.com Astronomy Sethanne Howard sethannehOlmsn.com Biology/Biophysics Eugenie Mielczarek mielczarOlphysics. gmu.edu Botany Mark Holland mahollandOlsalisburv.edu Chemistry Deana Jaber diaberO, marvmount.edu Environmental Natural Sciences Terrell Erickson lerrell.erickson 1 Olwdc.nsda.gov Health Robin Stombler rstomblerO.aubumstrat.com History of Medicine Alain Touwaide atouwaideOhotmail.com Operations Research Michael Katehakis mnkOrci.rutgers.edu Physics Katharine Gebbie katharine.gebbieOnist.gov Science Education Jim Egenrieder i imOdeepwater.org Systems Science Elizabeth Corona elizabethcoronaOlgmail.com Washington Academy of Sciences The Journal of the Washington Academy of Sciences joins the JSTOR Archive i The Washington Academy of Sciences has signed an agreement with the JSTOR archive dedicated to preserving scholarly literature. The complete back run of the Journal of the Washington Academy of Sciences (JWAS), which dates to 1899, will be digitized and made available via the JSTOR online platform. In addition to the Washington Academy, more than 1,050 publishers, including scholarly societies and publishing academies of sciences, are currently part of the JSTOR archive which hosts some 2,200 digitized journals comprised of 9 million digitized articles in various collections. For example, the oldest journal in the JSTOR collections is the Proceedings and Transactions of the Royal Society of London , which dates back to 1665. More than 8,000 institutions from 175 countries make use of JSTOR, including universities, secondary schools, government and non- profit organizations, community colleges, museums, and public libraries. The Academy’s former president Terrell Erickson says, “This is a terrific opportunity for the Washington Academy of Sciences as it expands our Journal’s reach beyond our current subscriber base to a much larger audience.” Several programs help to make sure that the archive’s contents are widely-available at a reasonable cost to users such as students and other science professionals. For instance, in addition to the 8,000+ subscribing libraries, JSTOR is also available to individual unaffiliated researchers who can access single articles through various JSTOR accessibility programs like “Register & Read” and “JPASS.” JSTOR further supports the African Access Initiative (AAI) and Developing Nations Access Initiative (DNAI), and these initiatives waive or reduce fees for 1,279 not- for-profit and academic institutions in developing countries. The not-for-profit JSTOR archive was conceived to help libraries and publishers respond to the rising costs associated with the storage of printed journal literature and to ensure that this material would not be “lost” as academic research became increasingly electronic. Through the digitization of complete journal runs, JSTOR makes it possible for Winter 2015 2 subscribing libraries to share the costs associated with storage and maintenance of journal literature, as the non-destructive digitization process will be done at no cost to the Academy. Furthermore, JSTOR is offering the Academy a modest revenue-sharing arrangement based upon access to JSTOR by its users. JSTOR will work from print copies of JWAS to create image files that are exact replicas of the original Journal pages and text files that enable searching. Upon completion of this process, users will be able to conduct full-text searches back to the first volume and issue in 1 899 when JWAS was called the Washington Academy of Sciences Proceedings. Scholars will then be able to browse, search, view, and print JWAS and the earlier Proceedings directly from their desktops. The Academy will retain the copyright to the material published in its Journal , as the JSTOR license agreement is non-exclusive. The Academy is planning to convert the current JWAS hard-copy format to an online version, and is exploring hosting JWAS electronically at its own website for its members and numerous paid individual and institutional subscribers who will be the only ones to have access the current and recent issues of JWAS. There will be a 3-year gap between the most-recently published issue of the Journal and the last issue available in JSTOR. This window of time is being designated for the puipose of separating these paid subscribers and members from the older issues which will be available via the archive. Questions about the Academy’s Journal can be directed to JWAS editor Sethanne Howard, sethanneh@msn.com . Washington Academy of Sciences 3 MEMBER MILESTONES Julius “Jay” Earl Uhlaner ( 1 9 1 7 - 20 1 5)' By Gerald P. Krueger Human Factors and Ergonomics Society (HFES) Fellow Julius “Jay” Earl Uhlaner was born in Vienna, Austria, in 1917. In 1928, he immigrated to the United States, where he became a naturalized citizen and left a lasting legacy through his leadership and research achievements, especially in applying psychology to military problems. , .. „ , .... Jay graduated from City College of New York in 1938 Julius Earl Uhlaner with a BS in science. He worked in human engineering at Ford Motor Company in Michigan from 1939 to 1940 and established a driver research lab. In his early human factors work, he focused on driver vision, training, and safety issues. These interests led to his thesis work for his MS in psychology and statistics from Iowa State University in 1941 . His contributions to highway safety included significant research on the visibility and interpretability of roadway signs with different types of lettering (e.g., height/width ratios of letters). He served on the Highway Safety Research Board in Lansing, Michigan, and dealt with human factors issues. While serving as a psychologist in the Army Air Coips during World War II from 1943 to 1946, Jay was involved with developing criteria for selecting pilots. From 1946 to 1947, he was assistant director for research and training for the New York State Division of Veteran Affairs. Combining his bent for human factors and personnel selection, he earned a PhD in industrial and organizational psychology at New York University in 1947. Jay then joined the Army Personnel Research Branch as a research psychologist. As the organization grew, it eventually became the Behavior Systems Research Lab (BSRL). In 1969, Jay became BSRL technical 1 Human Factors & Ergonomics Society Bulletin , 58. No. 10, October 2015 Winter 2015 4 director. Two years later, he also took on the title of chief psychologist of the U.S. Army, which is still worn today by the director of BSRL’s even broader-based successor organization: The Army Research Institute (ARI) for the Behavioral and Social Sciences. Under Uhlaner’s visionary guidance, ARI gradually took on missions to develop and improve the performance of people in the Army through behavioral sciences research on personnel selection, classification, job placement, training systems, and human factors in systems design. With Uhlaner at its helm from 1969 to 1978, ARI grew to employ more than 400 research psychologists, many of them well steeped in and practicing classical human factors methods and attaining many noteworthy accomplishments. Jay was best known for some of his innovative contributions to the Army. He foresaw early on the movement toward reliance on computers and automation and had ARI focus on “person-in-the-loop” approaches to examining soldier-system interface situations wherein the infusion of new technologies could enhance soldier performance, training systems, and equipment system testing. He spearheaded development of the first psychological military qualifications test legislated by Congress; introduced computers as major tools and partners in behavioral science research; pioneered research on night-vision testing and driver performance; introduced the first classification system based on psychological aptitude testing in the military services; pioneered the “system measurement bed,” a methodology that influenced industrial psychology; and fostered an interdisciplinary approach to ARBs research. During his career, Jay published close to 200 articles in scientific journals and books on the subjects of industrial psychology, military psychology, and related topics. In 1976, President Gerald R. Ford awarded him the U.S. Presidential Award for Management Improvement for his commanding role in the development and implementation of the Army Classification Battery and Aptitude Jay Ulilaner as director of the U.S. Army Behavior & System Research Laboratory Area System, representing major Washington Academy of Sciences 5 advances in the field of soldier performance prediction. In 1995, the American Psychological Association’s (APA) Division 19 (Military Psychology) recognized Jay with the Lifetime Achievement Award in Military Psychology for his many accomplishments in the application of behavioral science research to military problems. In 2011, Division 19 initiated an award in his name: the Julius E. Uhlaner Award for Distinguished Contributions to Research on Military Selection and Recruitment. In addition, Jay was a Fellow of FIFES, APA, and the Washington Academy of Sciences (WAS). In 1976, WAS granted him the first award “for scientific work of high merit in behavioral sciences” (see below). After retiring from the Army in 1978, Jay was senior vice president at Perceptronics, Inc., a human performance modeling, simulation, and training consulting firm in California (at that time). One of the more notable programs he fostered as part of a consortium for Defense Advanced Research Projects Agency (DARPA) was SIMNET, which offered a tank battle 3-D virtual simulation training network that permitted dozens, if not hundreds of operators in tanks, helicopters, close support aircraft, and other battlefield entities to interact with one another during war game training. At Perceptronics, Jay also did extensive work in mining safety for the Department of Commerce. He retired in 2000 but continued as a member of the board of directors. Subsequently, he carried out his own part-time behavioral sciences consulting work for another decade. Having watched him from a short distance, I can say that Jay Uhlaner continually demonstrated significant political and scientific savvy in dealing with bureaucracy and in getting things done. He was particularly adept at obtaining buy-in to build up human factors research psychology in the military by having his staff seek to provide what the country’s leaders and soldiers needed most. Jay’s family can be contacted through his beloved wife of 66 years, Vera Uhlaner, at P.O. Box 967, Corona del Mar, CA 92625-9998. Winter 2015 6 Award by the Washington Academy of Sciences The presentation was made at the Annual Awards Dinner meeting of the Academy on Thursday, March 18, 1976, at the Cosmos Club. Dr. Julius E. Uhlaner, Chief Psychologist of the U. S. Army and Technical Director of the Army Research Institute for the Behavioral and Social Sciences, and Adjunct Professor of Psychology at George Washington University, was cited for “his outstanding technical direction and leadership in Applied Psychology.” As a psychologist, he is best known for contributions to military psychology, having spent the major part of his career as a civilian research psychologist in the Army. However, he also kept closely in touch with academia and industry. He is best known for some of his innovative contributions to the Army, having developed the first psychological military qualifications test legislated by Congress; introduced the use of the computer as a major tool and partner in Behavioral Science research; pioneered night vision testing research and driver research; introduced the first differential classification system based on psychological aptitude testing anywhere in the military services; pioneered the “system measurement bed,” a methodology which influenced the field of industrial psychology; and fostered the interdisciplinary approach to much of his research. Also, he has exhibited very active professionalism, including the holding of elective offices in divisions of the American Psychological Association. His awards in the Federal service include the Citation for Meritorious Civilian Service, 1960; Citation for Exceptional Civilian Service, 1969; and Citation for Outstanding Performance, 1972. His combination of experience and education led to his trademark for the conduct of research in the Behavioral Sciences — an interdisciplinary approach, systems oriented, and the use of research products. He was elected a Fellow of the Washington Academy of Sciences in 1963; he was also a Fellow of the American Psychological Association. He was a Fellow of the Human Factors Society and the Iowa Academy of Sciences. Other societies of which he is a member are the Operations Research Society of America, International Association of Applied Psychology, Psychonomics Society, and District of Columbia Psychological Association. Washington Academy of Sciences A Study of the Primary Granitoid Outcroppings and Sedimentary Rocks in the Boali Region of the Central African Republic 7 Narcisse Bassanganam, Yang Mei Zhen, Prince E. Yedidya Danguene, Minfang Wang Earth Resource, China University of Geosciences, Wuhan, China Earth Faculty, University of Bangui, Central African Republic Abstract Located in the Central African Republic, the region of Boali is noted for its waterfalls and for the nearby hydroelectric projects. The waterfalls of Boali are 250 m wide and 50 m high, and are a popular tourist destination. The Central African Republic (CAR) has large reserves of Granitoids that remain largely untapped. That is why these rocks, which outcrop and which constitute the base of the Boali region and its surroundings, caught our attention. Previous studies by Bowen (1915) explained the order of appearance of various minerals as a function of the temperature and initial magma (SiCE) content. According to Bowen’s diagram, we can say that the magma underwent a magmatic differentiation giving rocks that are poor in silica (Diorite), followed by rocks rich in silica (Granodiorite and Granite). Knowing the absolute age of the Granitoids on the edge of the craton of Mbomou (2.1 Ga, Moloto et al., 2008, and Toteu et al, 1994), we can deduce the chronology of other formations. Initially there was the formation of the metamorphic formations and sandstones of Boali. This was followed by a slow intrusion of magma which crystallized in depth to give grainy rock (granitoids and pegmatite) in the region of Boali. This intrusion had metamorphosed the pre- existing formations through an orthogneiss. Introduction Boali is a town located in the Ombella M’poko prefecture of the Central African Republic (CAR) (See Figure 1). It is located 100 km northwest of Bangui, the capital of CAR. Boali is between 18°7'0"E longitude, and 4°48'0"N latitude. Access is through the National Road 1 (RNl). Boali is a sub-prefecture in the CAR. The CAR is divided into 16 administrative prefectures, two of which are economic prefectures, and one an autonomous commune; the prefectures are further divided into 71 sub- prefectures. The prefectures are Bamingui-Bangoran, Basse-Kotto, Haute- Kotto, Haut-Mbomou, Kemo, Lobaye, Mambere-Kadei', Mbomou, Nana- Winter 2015 8 Mambere, Ombella-M'Poko, Ouaka, Ouham, Ouham-Pende and Vakaga. The economic prefectures are Nana-Grebizi and Sangha-Mbaere, while the commune is the capital city of Bangui. Fig. 1: Map of CAR showing Ombella M’poko and Boali study area The Central African Republic is a country rich in mineral resources with an important reserve of Granitoids. Granitoid or granitic rock is a variety of coarse grained plutonic rock similar to granite which is composed predominantly of feldspar and quartz. These rocks outcropped and constitute the base of the Boali region, but unfortunately are not exploited. Geologically Boali is very interesting because of its Granitoids. We will identify and define the importance and usefulness of the Granitoids not only to geology, but also for the economy and social development in the CAR. We note that a school was built at Crossing-Boali in 1953 by the priest Alosiste Gezst, and recently, in 2001, the College of General Education (C.G.E) was built. Washington Academy of Sciences 9 Early geologic studies by Bowen (1915) defined the order of appearance of various minerals. Cornacchia et al. (1989) described the geologic formations of Boali. They mentioned that the greenstone belt of Bogoin-Boali rocks represents a succession of structures with a narrow synclinal appearance drawing a large half circle. These structures end in the east in the Bogoin area and to the north in the Boali sector as the outcroppings observed north of Boali. Poidevin (1979) defined the geochemistry of Precambrian basaltic rocks from the CAR; at Mbi not far from the river M’poko there are three types of petrographs: Schist sencitic, chlorite schist, and quartzite. Cornacchia and Dars (1983) showed that a corridor of faults cut north of the CAR existed. Cornacchia et al. (1985) found in the sandstone quartz veins containing crystals of rocks. Poidevin and Pin (1986) showed that the outcropping is plural-kilometric with an intrusion of dolerite and granites. Lithological studies of the Boali-Bogoin-Mbi region by Cornacchia and Giorgi (1986) defined a vast area ranging from the border of the republic of the Congo to south of the Lobaye Subit-Possel road including the Boda area. Their work was earned out south of the M’poko River and continued from the town of Bogoin to Yangana up to the Yasi series in the area of Bangui. Biandja (1988) earned out his work largely in the northern region of the Bogoin. Biandja (2000) pointed out that the southern part of the Boali region is characterized by a series of “Mbi” (waterfalls) incorporated from the bottom upwards. The series contains amphibolites of Mbali and Mbi and pillow basalts. All the intruded granite is in the lower course of the river Mandjo. North of the Bako village on the Mbi, this succession of granite becomes abnormal when it contacts the red sandstone and the red shale of the base of the sandstone shale set. In the northern region of the Bogoin there is a succession of chloritized migmatite and amphibolites that include some biotite in the faults area. There is also migmatized ferruginous quartzite. The sub-horizontal schistose sandstone does not conform to the christallophylliennes formations. However, the whole region of Boali does show some similarities between the north and south. According to studies done by Poidevin (1991), Biandja (1988); and Cornacchia et al. (1985-1989), the Boali region forms the southern part of a greenstone belt that represents the northwestern part of Bogoin-Boali. The Winter 2015 10 orientation of this greenstone belt runs east-west ending at the eastern edge and marries with an intrusive granite border at the western edge. According to the report from the meteorological station of Mbali covering 1993 to 2000, the average annual rainfall generally ranges from 1900 mm in February to 2630 mm in December, with an average maximum of 2868 mm. The Granitoids are on the edge of the craton of Mbomou (age 2.1 Ga) (Moloto et a/., 2008; Toteu et a/, 1994). A craton is a large, stable block of the Earth’s crust forming the nucleus of a continent. Recent studies by Rolin (1992) focused in the Central African Republic area of pan-African strike- slip of the Oubanguides. In general, Djebebe-Ndjiguim (2013) found that the density of the vegetation made it very difficult to search for significant outcrops. We continue their work to include not only new information on the geologic formation of the Boali region, but also to note the effect that non- exploitation of the granitoids in the area has on the region. It is a complex issue. Consequently the granitoids have not contributed to the social development in the Boali area in particular and to the CAR in general. Techniques Used to Gather the Data Boali is located 100 km northwest of Bangui, the capital of CAR. This field study was done on 24/25 June 2015. We used the basic tools of the geologist: a compass, camera, hammer, bag, notebook, and pencil. Out- general approach is based on the work of Cornacchia and Giorgi (1986). As noted by Djebebe-Ndjiguim (2013) the amount and density of the vegetation made it very difficult to search for significant outcrops. The authors followed two protocols set by previous researchers. The first protocol we followed was that of Biandja (1988). His work was carried out largely in the Bogoin northern region. In his lithological description he was able to list petrographic features consisting of lateritic, ferruginous, and conglomeratic blocks for recent formations. They contained, on average, quartzite, white quartzite, sandstone quartzite for covering the proterozoic formation; meta-volcano sedimentary, ferruginous quartzite, gneiss, Amphibolites, meta-volcanic basic to ultra-basic schist, and Metarhyolitoids (meta-volcanic acid) for the base formations of metamorphic rocks. For the intrusions Biandja distinguished many Washington Academy of Sciences characteristics of crystalline Granitoid intruding porphyroides granites from the base. The second protocol we followed was that of Poidevin (1991). His work was also earned out north of Bogoin. He identified different petrographic characteristics and classified them by stratigraphic unit (as U/? where n is a number 1 to 4). His four classifications are: Andesite in pillow- lavas and chlorite amphibolites for the main basalt unit (Ul); Para amphibolites, meta-rhyolites, with greywacke, feldspathic quartzite to amphiboles for the intermediate unit (U2); the greenstone and many pillow- lavas for the upper unit (U3); and Itabirite (U4). In addition to his four stratigraphic units, he also revealed the existence of geological formations of regional importance such as the granitoids and the series of schisto- quartzitic rocks. The Geologic Data We studied a variety of rocks types: plutonic; sedimentary; metamorphic; and deformations of rocks. In general, the extent and density of the local vegetation made it very difficult to search for significant outcroppings. (Djebebe-Ndjiguim 2013). We will consider the variety of rocks type by type. Plutonic rocks A pluton is a body of intrusive igneous rock (called plutonic rock) that is crystallized from magma slowly cooling below the surface of the Earth. In this category we studied two types: quartz veins and Granitoids. Quartz vein (lode): There are two types of quartz veins in the study sector: metamorphic formations; and rock crystal veins located in the sandstone. Quartz veins are not barren of mineralized rock crystals. And so in these veins we noted the presence of some minerals, such as emerald and gold, due to the movement of warm waters (Comacchia et al. 1985). In the greenstone belt toward the vein wall there are altered Amphibolites in the chlorite-schist. According to Cornacchia et al. (1985), quartz veins containing rock crystals are found in the sandstone. These veins continue through to the quartz veins found in the metamorphic rocks. They originate in the emanation from granite and are mineralized rock crystal. The veins occur from 30° N to the south with a thickness of 15 centimeters Winter 2015 12 to 5 meters and orient North 90° to 105°. They include some geodes in which beautiful quartz crystals have developed. The crystals have a thickness of at least 30 millimeters. They can reach 1.5 m thick in some geodes. Across the rock outcroppings in the direction N 30° they form solid blocks of milky-white appearance and are poly fractured. During our field observations we spotted four levels of implementation veins in the quartz downstream of the dam at the Mbi, and even more implementation veins next to the road to Bossemmbele. These are the extension of those downstream of the dam. The seams are flush to both sides of the hill overlooking the dam. Some veins fold into a semicircle under the mast of the town’s police station and also in the stone quarry. Granitoids: Granitoids are plutonic rocks that are poor in silicon dioxide (SiCh). They are designated in the upper part of the table of the international classification of streckeizen. In our region of Boali there are diorites, granodiorites, granites of Mbi, and granites of Bolen. We observed that granitoid outcroppings in the region cover a very large area. Although grey in appearance these rocks sometimes have alternating beds of dark ferromagnesian amphibole and biotite and clear beds (quartz, feldspar, and muscovite). The granitoids of Mbi orient 30° N dipping 70° W. They are traversed by quartz aplite and pegmatite veins. These formations are subdivided into granite, matching granite, and orthogneiss. Granite of Mbi - Granite is a fully crystalline rock. Minerals are on average 2 to 5 mm in size about the size of a grain of wheat (granite comes from the Latin granurn = grain). They contain three essential minerals: quartz, alkali feldspar (orthoclase and microcline), and plagioclase combined with mica (biotite and muscovite). The quartz comes in a grayish color surrounding other crystals. Its appearance is that of salt but with a bold loamy appearance as if it burst out of the rock. In the region of Mbi the quartz has a conchoidal fracture. The alkali feldspars have variable colors (white, pink, red) and are twin Karlsbad (the crystal is alternately brilliant and dull). Biotite occurs in black strips some with a golden luster, with cleavages or cleavage lines. This intrusive massif has lagged behind the plate tectonics. It is late granite, very marked, and located on the left bank of Mbali in our study area. This massif has a grainy central facies with large elements of alkaline feldspar, very rich in feldspar, and with very fine grain borders. It manifests itself in Washington Academy of Sciences 13 the landscape by significant outcroppings and can be observed in the stone quarry without major difficulty. Granodiorites: Granodiorites have a constitution nearly that of granite; their silica content can be as strong as that in granite but contains more plagioclase feldspar than orthoclase feldspar. Common rock “granite” can be distinguished from granodiorites by carefully considering their feldspar. Granodiorites of the Boali region have micro-fractures that allow the circulation of fluids. There is a possibility of finding gold and pyrite. The presence of the epidote gives the rock its green color. This epidotisation is due to the alteration of the potassium in the feldspar. The outcropping is plural-kilometric with an intrusion of the dolerite and granites. They are dated to 2.1 Ga. (Poidevin and Pin 1986) Diorites: Diorite is an intrusive igneous grainy rock with a silica deficiency (less than 20%); therefore, it does not contain free quartz. It is principally composed of the minerals plagioclase feldspar (typically andesine), biotite, hornblende, and/or pyroxene. Feldspar, generally grayish, helps to give the rock a dark color. Diorites are intruding amphibolites and are contiguous with the granodiorites of Mbi. Dolerite: Dolerites are intermediate rocks that fall between grainy gabbros and basalts with microlitic grain that is visible under a microscope and shows sub-hedral plagioclase laths molded by interstitial pyroxene. They are generally massive and compact with a color ranging from black to grey but more often dark green. We saw three hills of dolerite in the intruding granodiorites in a nearby outcropping of granite. Contact areas: The Bangui-Boali section shows several contact areas characterized by vein crates between sedimentary rocks and metamorphic rocks. About 1 00 km from the first dam to the north of Boali we find a contact between Amphibolites and sandstones. The contact is characterized by a puckered quartz hill. At 123 km another contact is characterized by a type of vein that is favored by the hydrothermalism between granitoids and massive Amphibolites; there is another contact with upright schistosity (sub-vertical N 60°). Winter 2015 14 Sedimentary rocks Sedimentary rocks are rocks that are formed by the deposition of material at the Earth’s surface and within bodies of water. Sedimentation is the collective name for processes that cause mineral and/or organic particles (detritus) to settle and accumulate or minerals to precipitate from a solution. Sandstone is a sedimentary rock. It is a consolidated rock that belongs to the class of arenite rocks that have a grain size between 0.0625 and 1 mm. Thus we can distinguish between quartz sandstones, where a microcrystalline material persists between the grains of quartz, and quartzite sandstone, where grains are linked to each other following a secondary pathway that depends on the cement. They are located south of the Kassango area and belong to the Oolitic sandstone of the Boali series (see Photo 1C which shows the sandstones of Boali falls). The corresponding features are homogeneous fine-grained quartzose sandstones (with clay cement in the south-west that changes to siliceous cement to the south and south-east). Quartzite occurs on the road to the city in the stratigraphic extension falls downstream from the third Boali falls. The sandstones are grayish and friable rocks whose diamante detrital minerals are amorphous quartz grains often recrystallized as anhedral feldspar. Observation with a microscope reveals rare biotite lamellae and a few fine flakes of muscovite. Boali sandstones are the equivalent of those of Fatima, a district located in the Bangui capital of the CAR. Metamorphic rocks Metamorphic rocks arise from the transformation of existing rocks, in a process called metamorphism, which means “change in form”. We found four types of metamorphic rocks in our study area: Schist, Amphibolites, Itabirite, and Gneiss. Schist: Schists are characterized by medium to large, flat, sheet-like grains with a preferred orientation. The outcroppings form in slabs on the bed of the Kassango at the roadside and are often interstratified with the sandstones. Of greenish hue this rock has a sub-vacuolar structure throughout. It presents numerous vacuoles and therefore it is strongly schistose. It fits into beds that are clear of recrystallized quartz and chlorite and has dark beds of rare sericite altered biotite. We find these in the region of Boali, and we find this same shale on the road to Damara. These are rich Washington Academy of Sciences 15 in mica and nodules and are very crumpled; this is the schist of Boali, the equivalent of the Fatima shale that belongs to the series of Bangui, which are above the Yangana shale. Photo 1 : sandstones of Boali falls Amphibolites: Amphibolites are dark green rocks consisting mainly of amphibole crystals more or less ordered along the planes of schistosity. We can distinguish laterized amphibolites, layered amphibolites, and massive amphibolites. Highly altered and chloritized laterized amphibolites are found in the area of the lakes of the crocodiles in a stone quarry about 100 km away Winter 2015 16 from the laterites on the highway. Layered amphibolites are amphibolites that have a banded texture characterized by alternating feldspathic quartz beds and detachment beds. The hyper-fractured veins come within 1 km of the Bogoin village (village Bobissa). Massive amphibolites are mottled with Granoblastic massive rocks. The fine-grained rocks show a discontinuity in their arrangement. The dark minerals are dominant with a cleavage of amphibole and biotite. There are also some rare glitters of muscovite. The massive amphibolites stretch from the Bogoin village to where they make contact with the granodiorites. Itabirite : The itabirites are quartzite ferruginous rubane. The outcropping is in a kilometer wide band. These are generally quartz-rich rocks occurring with magnetite and often oligiste. This last mineral concentrates in massive structures around the quartz veins crossing the Banded Iron Formation (BIF) where the banding is very marked; there is a layout of dark magnetite beds alternating with beds of clear quartzo-feldsparthic rock. Gneiss: The gneisses are medium-grained or coarse rocks about 1 mm to 1 cm in size. They often have net foliation characterized by beds of generally dark hue, rich in minerals (mica, amphibole), and alternate with clear beds of ferromagnesian (white grey, pink) quartz and feldspar visible to the naked eye. We noted the presence of the orthogneisses, which are rocks that form a contact between amphibolites and granodiorites on one side and form a contact between the granite and granodiorites on the other side. Other metamorphic rocks The quartzite and muscovites rocks occupy the eastern part of the region. Shale appears in slabs on the bed of the Ngalou. Chloritoschistes and schist outcroppings occur in the region of Bogoin. Orthogneisses occupy the southern part of the region of Bogoin. The southern region of the Kolango is characterized by a lower relief that is very soft sided in its uppermost part. On the sides of the rocky massive benches, block elements, and fractured outcroppings we can distinguish massive metabasalts in the pillow lavas and metabasalts in the stringers of intruding quartz. Washington Academy of Sciences 17 Deformations Deformation takes an object from its initial state to its final state by mass transport (translation, displacement, rotation, and by internal deformation). The deformed object is defined by its dimensions. Stratification is one form of deformation. Bedding planes illustrate the style of the planar structural element. These were initially roughly flat, horizontal surfaces. Their characteristics and variations are an imprint of deformations that have been imposed by the sedimentary terrain since their deposition. This stratification is observed in the sandstone outcropping. At the entrance to the falls of Boali we observed the stratification cross the sandstone (See Photo IE). Geological foliation (metamorphic arrangement in layers) with medium to large grained flakes in a preferred sheetlike or planar orientation is called schistosity. The plane of the schistosity is called S. In formations containing more competent levels, stretching leads to socking which is to say leads to the segmentation of the most competent object into fragments and socks. Photo ID illustrates deformations characterized by boudinage, folds, faults, and shears. • Boudinage is a term used in geology to indicate structures formed by extension (where a rigid body is deformed often into a sausage or boudin like shape). • A fold is a permanent waveform deformation in layered rock (the rocks bend or twist). It occurs when one or a stack of originally flat surfaces (such as sedimentary rock) are permanently bent or curved. • A fault is a fracture in the bedrock. They are breaks accompanied by the relative movement of two components. The movement can be vertical (vertical, oblique, fault normal or reverse) or horizontal (strike-slip or shear). • A shear is the response of a rock to deformation usually by compression. The shear can be emphasized by certain minerals. We essentially observed the schist as shears and lineaments. These are break planes that are accompanied by the relative movement of two components which show the hang of the faults. Lineaments are mineral lineations that occur during metamorphic crystallization. Winter 2015 18 In our study area there is a boudin fold ply spilled to the NW. It consists essentially of anhedral quartz crystals a centimeter in size. It is molded into the clay schist (shale) and found at the roadside in Kassango. Quartz flanges are located in the clay schist of Kassango. They can be found at the level of the boudin folds. On the road to the town’s police station we find a crease spilled in sandstone. The itabirite are also very creased. The wrinkles are crooked with a very upright fold axis. Under the mast of the police station is a surrounding concentric fold with a diameter of 40 cm. The fold shown in photo ID is observed in the clay schist (shale) of Kassango. Finally, we see a deformation characterized by a fold slumping downward. Accompanying the schistosity and the boudin is a tangential tectonic surface with direction S-SE toward N-NW. A second deformation is a tangential tectonic surface contrary to the first. It runs NW-SE. This tectonic surface is confined by the mega fold conic running N-NW. A tectonic surface relates to the structure of the Earth’s crust and the large- scale processes which take place within it. Shears : Sinistral and dextral shears were observed at the stone quarry (S2, See Figure 2). They form a corridor of sinistral shear 5 m wide for the S2 shears and fall 155-45° SW. The basal formation of the stone quarry shows deformation bands approximately 60 m wide. We found that the dextral shear (S2, Figure. 2) was hardly observable. On the other hand, the S2 shears are very representative of the class. Figure 2: The center insert shows the dextral Shears in the area. The left and right sides of the figure show the sinistral shear, S2 Washington Academy of Sciences 19 Lineaments: The lineaments are mineral lineations during metamorphic crystallization. We observed lineaments in the sandstone towards the falls. The lineament runs N 120-35° S. It crosses all fractures. We observed in the dolerites two families of lineaments: one in direction N 135° sub-vertical; and the other, south-facing N 25° with a dip of 60°. There are small intercalations of gneiss in the outcroppings of dolerite. The thickness of these dolerites can reach 40 m. Diabase dykes continue to the top of the hill. All these formations in the sector are affected by brittle deformation which appears here as faults. The faults are the fractures in the bedrock. They are breaks accompanied by the relative movement of two components. The movement can be vertical (vertical, oblique, fault normal or reverse) or horizontal (strike-slip or shear). The fault of Boali is a normal fault (see Photo 1 E. F) corresponding to Figure 3, which shows a normal fault. These faults have been found in the sandstone in front of the police station (in the main city). They include three (3) series of fracturing. FI and F2 in direction N 0+/-100 they have an embedding of 60 -75° E, sub-vertical; and F3 in direction N 145 +/- 10 they are sub-verticals. See Table 1 (in front of the Internet service provider for Boali, in the main city). In addition, the brittle tectonic of the Mbi sector highlights four major series of faults: These series of faults run: N45° - N 50°; N 80° - N 100°;N130°-N 140°, corresponding to F3; and N160° -N 175° (see Table 1 (Mbi sector)). Fig. 3: the shear dextral and normal fault corresponding to the study areas observed Winter 2015 20 Locality Outcropping Fault Direction and dip In front of the Internet service provider for Boali sandstone faults F 1 and F2 NO -75 E N 10. sub-vertical N 145 Sector of Mbi fault N45-N50 N80-N100 N130-N140 N160-N175 Table 1: faults family observed in the study areas. Photo 2: The basic faults and veins Timeline for the Faults: Photo 2 shows the various fault lines we observed. First FI was established with a filling. Then a second fault F2 parallel to FI and included in FI was established with a new filling of a quartz vein. A third fault F3 (also found in the area of Mbi) oblique to FI and F2 in the direction N 145 sub-vertical, and joins with FI and F2, accompanied by its vein filling. Finally a mineral lineament of direction and dip, N 120 - 35 S (see Photo 2), has complex features that indicate it is recent. Structurally, D1 and D2 deformations having contrary motion show two tectonic movements, namely: tectonics of the Ebumean age (2.1 Ga) responsible for thrusting from S-E to N-W and a second pan-African age tectonics. We therefore suggest: a dating of metamorphic and sedimentary rocks to coincide with the chronology of events; an elemental geochemistry Washington Academy of Sciences 21 to trace for a Concordia diagram to have both the age of the formation and the age of metamorphism. On the regional level, there is a corridor of grid cut faults towards the north of the CAR (Cornacchia and Dars, 1983) in the direction N 70° and N 40°. This is the area of strike-slip faults of the Oubanguiides. Other setback faults with a direction N 130° to N 160° towards the sinistral fault affect all the structural units (Poidevin, 1991). These major setbacks date from the pan-African phase. As we get closer to our study area we find different faults in the major setback of the pan-African phase described by Rolin (1992). There are two families of faults (N 45° and N 80°) that correspond to the dextral grid N 70° and N 40° of the pan-African phase. There are flaws running N 130° - N 140° corresponding to the sinistral transcurrent N 130° of pan- African water. Similarly, faults running N 160° - 175° can be classified within the family of sinistral offsets at N 160° of the pan-African phase. We found two families of shear flaws including the first sinistral flaw (N 130°) and the second dextral flaw (N 45°) as we have described. These two flaws affect both formations of Mbi. Conclusion The CAR is a landlocked country in Central Africa. It is divided into 16 administrative prefectures, two of which are economic prefectures, and one an autonomous commune; the prefectures are further divided into 71 sub-prefectures. Geologically CAR is a country rich in mineral resources. Our study is located in the region of Boali, which is a town located in the prefecture of Ombella M’poko. Boali is on the National Road 1 (RN1) about 100 km northwest of the Bangui capital of the CAR. For this work we focused on the protocols set by previous studies of the geology formation in the Boali region. We also considered studies of the region by others. We include not only new information on the geologic formation of the Boali region, but also discuss the effect that non- exploitation of the granitoids in the area has on the region. The Boali region has an important reserve of granitoids, which form outcroppings and constitute the base of the region. Geologically granitoids consist essentially of quartz and feldspar (a ferromagnetic material). In region of Boali the Winter 2015 22 crystals form in veins which intrude into the sedimentary and metamorphic formation. These rocks are important and useful for economic and social development. In the region of Boali most mining is done by artisanal gold miners. Granitoids have never interested the people in the Boali region. We note that in the region of Boali, none of the local residences are made with material from the granitoids. Yet granitoids are needed for the infrastructure. In general for the CAR and in particular for the Boali region, granitoids are wealth ignored and abandoned. If the granitoids are exploited in the region of Boali, then they can contribute to the buildings; for example, tiles can be made of granitoid. Most products made with granitoids are the hardest of materials, which offer a luxury in comparison to marble tiles. In the petrographic plane the sedimentary rocks are quite fractured and mingle with quartz veins, metamorphic and magmatic rocks. Magmatic differentiation which led to the establishment of the Diorites, the Granodiorites and Granites as well as intrusion of intermediate rocks (Dolerite) shows a bimodal magmatism confirmed by the presence of Granitoids and Ultrabasites. The quartz from quartz veins can be made into glass for the manufacture of laboratory equipment such as: burettes, beakers, and test tubes, which are urgently needed in the Central African Republic. Quartz veins are an asset for developing jewelry workshops. We use the quartz from quartz veins in the manufacture of silicon pads, integrated circuits for audio and video devices, microprocessors for computers, solar panels, and electric watches. It is also used in gas and electronic lighters. Furthermore it can be used in construction, for coating houses, pavements, and layering of load-bearing seats. Unfortunately for the CAR in general and for the region in particular these rocks have not been exploited. We do note, however, that it is a complex issue. Consequently the granitoids have not contributed to the social development in the CAR and Boali. In Boali unemployed youth might find productive work by artisanal mining of these reserves. The government might consider implementing a policy for medium and small business development in the areas mentioned above that might lower the high percentage of unemployment with all its consequences, not least of which are prostitution and violence. In our investigation in the region of Washington Academy of Sciences 23 Boali, we conclude that currently the most important human activities are the individual artisanal agriculture production, fishing, and hunting to satisfy the daily needs. References 1915. N.L. Bowen. “The later stages of the evolution of the igneous rocks” Journal of Geology, 23, pp. 1 - 89. 1979. J.L. Poidevin. “Stratigraphic Precambrian formations of Central Africa Republic” In: 10th symposium Geol. Afric., Montpellier, 1979-1904 / 25-27. S.L.: S.N, p. 12 (Summary). 1983. M. Comacchia, R. Dars. “A major structural feature of Africa continent. The Central Africa Republic and Cameroon lineaments to the Gulf of Aden” Bull. Soc. Geol. France, XIV- 1, pp. 101-109. 1985. M. Cornacchia, L. Giorgi, J.C. Lachaud. “Preliminary note on the hydrogeology of the Bangui region. Central African Republic” 10th Congress Nat. Soc. Sav, (Montpellier) Sciences. Fasc. VI. pp. 331-342. 1986. M. Cornacchia, L. Giorgi. “The series Precambrian sedimentary and volcano- sedimentary Central African Republic” Ann. Mus. Roy. Afr. Central, Tervuren, Belgium, ser. in-8°, Sci. geol., 93, p. 51. 1986. J.L. Poidevin, C. Pin. “2 GA U-Pb Zircon dating of Mbi granodiorite (Central African Republic) and its bearing on the chronology of the Proterozoic of Central Africa.” London: Journal of African Earth Sciences , 5, 6: pp. 581-587. 1988. J. Biandja. “Metallogenic approach of Greenstone Belt Bogoin (RCA). His gold mineralization”. Earth Sciences. University Pierre et Marie Curie - Paris VI. 88-7. p. 345. 1989. M. Comacchia, L. Giorgi "Discrepancies and major Precambrian magmatic series of Bogoin region. (West-Center of the Central African Republic)" , J. Afr. Earth Sci., vol. 9, pp. 221-226. 1991. J.L. Poidevin. “ Greenstone belts of the Central African Republic (Bandas, Boufoyo, and Mbomou Bogoin). Contribution to the knowledge of the Precambrian craton northern Congo ” Thesis: University Blaise Pascal Clermont-Ferrand, 458 p. multigr. 1992. P. Rolin. “Presence of a major ductile overlap pan African age in the central part of the Central African Republic. Preliminary Results” C.R. Acad. Sci. Paris, 315, 467- 470. Winter 2015 24 1994. S.F. Toteu, W.R. van Schmus, J. Penaye, J.B. Nyobe. “U-Pb and Sm-Nd evidence for Eburnian and Pan-African high-grade metamorphism in cratonic rocks of southern Cameroon. Precambrian ’ Elsevier. Res vol. 67: pp. 321-347. 2000. J. Biandja. Private communication. 2008. G.R. Moloto-A-Kanguemba, R.I.F. Trindade, P. Monie, A. Nedelec, R. Siqueira. “A late Neoproterozoic paleomagnetic pole for the Congo craton: Tectonic setting, paleomagnetism and geochronology of the Nola dike swarm (Central African Republic). Precambrian ” Elsevier. Res vol: 164: pp. 214-226. 2013. C.L. Djebebe-Ndjiguim. “Characterization of the aquifers of the Bangui urban area, Central African Republic, as an alternative drinking water supply resource” Hydrological Sciences Journal. 58 (8), 1760-1778. Bios Narcisse Bassanganam is a Master’s Student in Mineral Resource Prospecting and Exploration at the China University of Geosciences (Wuhan). He also received a BS in the Exploration Engineering of Mineral and Resources from there. He has conducted field research in both China and the Central African Republic. Yang Mei Zhen is on the faculty of the China University of Geosciences (Wuhan). She teaches in the Earth Resources area and directed Mr. Bassanganam ’s work. Prince Emilien Yedidya Danguene is an Educator Researcher at the University of Bangui and President of the Christian Community of the Central Africa Republic. He is the former Minister of Development of Mining and Energy Projects. Minfang Wang is an Associate Professor at the China University of Geosciences (Wuhan). She has worked on many projects, including research at the Institute of Mineralogy and Technology at Freiberg. Washington Academy of Sciences 25 Hydrogen Clouds from Comets 266/P Christensen and P/2008 Y2 (Gibbs) are Candidates for the Source of the 1977 “WOW” Signal Antonio Paris St. Petersburg College, FL Evan Davies The Explorers Club, 46 East 70th St, New York, NY Abstract On 1977 August 15, the Ohio State University Radio Observatory detected a strong narrowband signal northwest of the globular star cluster M55 in the constellation Sagittarius (Sgr). The frequency of the signal, which closely matched the hydrogen line (1420.40575 1 77 MHz), peaked at approximately 23:16:01 EDT. Since then, several investigations into the “Wow” signal have ruled out the source as terrestrial in origin or other objects such as satellites, planets and asteroids. From 1977 July 27 to 1977 August 15, comets 266P/Christensen and P/2008 Y2 (Gibbs) were transiting in the neighborhood of the Chi Sagittarii star group. Ephemerides for both comets during this orbital period placed them at the vicinity of the “Wow” signal. Surrounding every active comet, such as 266P/Christensen and P/2008 Y2 (Gibbs), is a large hydrogen cloud with a radius of several million kilometers around their nucleus. These two comets were not detected until after 2006, therefore, the comets and/or their hydrogen clouds were not accounted for during the “Wow” signal emission. Because the frequency for the “Wow” signal fell close to the hydrogen line, and the hydrogen clouds of 266P/Christensen and P/2008 Y2 (Gibbs) were in the proximity of the right ascension and declination values of the “Wow” signal, the comet(s) and/or their hydrogen clouds are strong candidates for the source of the 1977 “Wow” signal. Introduction On 1977 August 15 at approximately 23:16:01 EDT, the Big Ear Radio Telescope at The Ohio State University detected an intermittent narrowband radio signal (<10 KHz) northwest of the globular star cluster M55 in the constellation of Sagittarius (Sgr) [ 1 ] [2] and approximately 2.5° south of the Chi Sagittarii star group [5]. Determining the exact location where the 72- second signal originated from in the sky was problematic because the telescope used two separate feed horns to search for radio signals [5], The data from the signal, moreover, were processed in such a way that it was Winter 2015 26 difficult to establish which of the two horns detected the signal [2]. There are, therefore, two possible right ascension values for the source of the alleged extraterrestrial intelligence signal: 19h22m24.64s ± 10s and 19h25m17.01s± 10s and the declination was determined to be -27°03' ± 20 (Table 1) [2]. Two similar values for the signal’s frequency were assigned: 1420.356 MHz and 1420.4556 MHz. These two frequencies fall close to the hydrogen line, which is 1420.40575177 MHz [6], Table 1 : Right Ascension and Declination Equinox Conversions; and Galactic Coordinates for the “Wow” Signal (Source: Ohio State University Big Horn Report) Declination Positive Horn Negative Horn B1950.0 Equinox -27°03'± 20’ 19h22m24.64s ± 10s 19h25m17.01s± 10s J2000.0 Equinox -26°57'± 20’ 19h25m31s± 10s 19h28m22s± 10s Galactic Latitude N/A -18d53.4m± 2.1m -19d28.8m± 2.1m Galactic Longitude N/A 1 ld39.0m± 0.91TI 1 ld54.0m± Q.9m Previous Investigations by the Astronomical Community Subsequent research to re-detect and identify the “Wow” signal by The Ohio State University, the Very Large Array, and The University of Tasmania’s Mount Pleasant Radio Observatory were null. After a search of the area where the “Wow” signal was detected (Table 2), the Very Large Array and The Ohio State University Radio Observatory concluded there was strong evidence against the origin of the source as terrestrial in nature or objects such as planets, man-made spacecraft, artificial satellites, and radio transmissions emanating from Earth. Furthermore, the Very Large Array proposed the intermittent “Wow” signal matched the signature of a transiting celestial source [5], while The University of Tasmania suggested the signal was moving with the source of the hydrogen line [7], Anatomy of a Comet and Its Hydrogen Cloud The distinctive parts of a comet include the nucleus, coma, dust tail, ion tail, and a hydrogen cloud. Moderately active comets are surrounded by a widespread cloud of neutral hydrogen atoms [4], The hydrogen is released from the comet when ultraviolet radiation from the Sun splits water vapor molecules released from the nucleus of the comet into the constituent components oxygen and hydrogen [8], The size of the hydrogen cloud is determined by the size of the comet and can extend over 100 million km in Washington Academy of Sciences 27 width, such as the hydrogen cloud of comet Hale Bopp [9]. As a comet approaches the Sun, its hydrogen cloud increases significantly. Since the rate of hydrogen production from the comet’s nucleus and coma has been calculated at 5 x 102g atoms of hydrogen every second, the hydrogen cloud is the largest part of the comet [9]. Moreover, due to two closely spaced energy levels in the ground state of the hydrogen atom, the neutral hydrogen cloud enveloping the comet will release photons and emit electromagnetic radiation at a frequency along the hydrogen line (1420.40575177 MHz) [10]. Date of Search RA DEC VLA 25 SEP 1995 19h21m28.1s to 19h25m48s -27°41 to -26° 18 07 MAY 1996 19h21m28.1s to 19h25m48s -27°41 to -26° 18 Ohio State U. 05 OCT 1998 09 OCT 1998 9-10 APR 1999 17-18 MAR 1999 20-21 MAR 1999 22-23 MAR 1999 19h22m22s -27°03 1 9h25m 12s -27°03 1 9h25m 1 2s -26°48 19h22m22s -27° 18 1 9h25m 1 2s -27° 18 1 9h22m22s -26°48 Table 2: Right Ascension and Declination Observations Grid by the VLA and Ohio State University (Source: VLA and Ohio State) Comets 266P/Christensen and P/2008 Y2 (Gibbs) From 1977 July 27 to 1977 August 15, Jupiter-family comets 266P/Christensen and P/2008 Y2 (Gibbs) were transiting in the vicinity of the Chi Sagittarii star group and significantly close to the source of the “Wow” signal (Figure 1) [3][ 1 1 ]. Of significance to this investigation, the purported source of the “Wow” signal was fixed between the right ascension and declination values (Table 3) of comets 266P/Christensen and P/2008 Y2 (Gibbs). On their orbital plane, moreover, 266P/Christensen was 3.8055 AU from Earth and moving at a radial velocity of +13.379 km/s; and P/2008 Y2 (Gibbs) was 4.406 AU from Earth and moving at a radial velocity of + 19.641 km/s (Figure 2) [3]. Winter 2015 28 Figure 1: Location of Comets 266P and P/2008 from 1977 July 27 to 1977 August 15. (Source: The Minor Planet Center and NASA JPL Small Body Database) [11]. 19h 30m . •' 19h 20m 19h 10 8.3” X- 4.9” . ’ * -24‘ 1 Chil Sgr — • — I 178P/HUG-BELL .266B/CHRISTENSEN 2P6P/CHRISTENSEN ' * . — ,(g)07AUG77 qisausT? • NEGATIVE HORN POSITIVE HORN 15 AUG 77 15 AUG 77 J2000 Equinox J2000 Equinox ' ®-®- . • ■ © ■ . © NEGATIVE HORN POSITIVE HORN • 15 AUG 77. 15 AUG 77 B1950 Equinox' B1950 Equinox ORBITAL PATH Jr • ^ W ® P/2008 Y2 (GIBBS) p/2008 Y2 (GIBBS) ' R/2008-Y2 (GIBBS) ' 27 JULY 77 oi AUG 77 . •' 05 AUG 77 P/2008 Y2 (GIBBS) 15 AUG 77 ' * . • ‘ ■/ 1 19h 30m 19h 20rrt / 19h 10m r © Table 3: Right Ascension and Declination Values for Comets P/2008 Y2 (Gibbs) and 266P/Christensen (Source: Minor Planet Center) Date RA DEC P/2008 Y2 (Gibbs) 27 JUL 1977 19h28m12s± 10s -27°3 1 01 AUG 1977 19h25ml 7s ± 10s -27°33 05 AUG 1977 19h22m23s± 10s -27°35 15 AUG 1977 19h16m37s± 10s -27°36 266P/Christensen 07 AUG 1977 19h29m47s± 10s -25°53 15 AUG 1977 19h25m17s± 10s -25°58 Washington Academy of Sciences 29 Figure 2: On 1977 August 15, comet 266P/Chrislensen was 3.8055 AU from Earth and comet P/2008 Y2 (Gibbs) was 4.406 AU from Earth (Source: JPL Solar System Dynamics Database) [12] The data regarding cornets 266P/Christensen and P/2008 Y2 (Gibbs), therefore, strongly suggest either comet, or both, could be the source of the hydrogen line signal detected by the Ohio State University on 1977 August 15. Chemicals in comets emit radio waves. The hydrogen radio waves from a comet, such as from 266P/Christensen and P/2008 Y2 (Gibbs), travel through space akin to light. Therefore, radio telescopes, including the Big Ear Radio Telescope at The Ohio State University, could have intercepted them. It is noteworthy to comment, moreover, during observations of the area by the Very Large Array and The Ohio State University Radio Observatory (from 1995 to 1999), comet 266P/Christensen and P/2008 Y2 (Gibbs) were not in the neighborhood of the right ascension and declination values of the “Wow” signal (Table 4) [5], thus the hydrogen cloud from these two comets would not have been detected. Additionally, because the period for comet 266P/Christensen is 6.63 years and P/2008 Y2 (Gibbs) is 6.8 years [3], their orbital period could account for why the “Wow” signal was intermittent and not detected during subsequent searches of the area. Winter 20 15 30 Conclusions There is noteworthy data to propose that the hydrogen signal detected by the Big Ear Radio Telescope at The Ohio State University, specifically 1420.356 MHz and 1420.4556 MHz, emanated from the neutral hydrogen clouds of comets 266P/Christensen and/or P/2008 Y2 (Gibbs). There are, conversely, many unknowns the astronomical community will need to address to confirm the hydrogen clouds from these comets were the source of the 1977 “Wow” signal. To date, no observations have acquired and measured the size, mass and spectral signature, most critically, of these two comets. Additionally, in 1977 the Big Ear Radio Telescope was operating in drift scan mode. Consequently, if a comet (or any celestial object) was the source of the “Wow” signal, it should have been detected in the trailing beam after detection in the leading beam [13]. Comet 266P/Christensen will transit the neighborhood of the “Wow” signal again on 2017 January 25 and can be located at 19h25m 15.00s and declination -24°50' at a magnitude of +23 [3]. On 2018 January 07, comet P/2008 Y2 (Gibbs) will also transit the neighborhood of the “Wow” signal. Comet P/2008 Y2 (Gibbs) can be located at right ascension 19h25m 17.6s and declination -26°05' at a magnitude of +26.9 [3]. During this period, the astronomical community will have an opportunity to direct radio telescopes toward this phenomenon, analyze the hydrogen spectra of these two comets, and test the authors’ hypothesis. Washington Academy of Sciences 31 Table 4: Location of Comets 266P/Christensen and P/2008 Y2 (Gibbs) During VLA and Ohio State University Observations (Source: The Minor Planet Center) Pate RA PEC P/2008 Y2 (Gibbs) 25 SEP 1995 (VLA) 1 1 h42m +00°22' 07 MAY 1996 (VLA) 1 6h 1 1 m -32°0T 05 OCT 1998 (Ohio) 20h12m -22°45' 09 OCT 1998 (Ohio) 20h15m -22°4T 9-10 APR 1999 (Ohio) 22h02m -13° 19' 17-18 MAR 1999 (Ohio) 21h48m - 1 4°42' 20-21 MAR 1999 (Ohio) 21h50m -14°31' 22-23 MAR 1999 (Ohio) 2 1 h5 1 m -14°24' 266P/Christensen 25 SEP 1995 (VLA) 15h12m -20°31' 07 MAY 1996 (VLA) 18h03m -27°28' 05 OCT 1998 (Ohio) 22h01m -14°09' 09 OCT 1998 (Ohio) 22h00m -14° 12' 9-10 APR 1999 (Ohio) 00h24m +02°35' 17-18 MAR 1999 (Ohio) 23h56m -00°42' 20-21 MAR 1999 (Ohio) 23h59m -00°17' 22-23 MAR 1999 (Ohio) 00h03m +00°08' References 1 . Shostak, Seth (2002). “Interstellar Signal from the 70s Continues to Puzzle Researchers”, http://archive.seti.org/epo/news/features/interstellar-signal-from-the- 70s.php accessed on 01 Oct. 2015. 2. Ehman, Jerry R. (2010). “Wow! Signal 30th Anniversary Report.” North American Astrophysical Observatory http://www.bigear.org/Wow30th/wow30th.htm accessed on 14 Oct. 2015. 3. The International Astronomical Union Minor Planet Center, Database: MPEC 2009- A03 P/2008 Y2 (Gibbs); MPEC 2008-U27 266P/Christensen. http://www.minorplanetcenter.net/ accessed on 2 1 Nov. 2015. 4. Centre for Astrophysics and Supercomputing. Cometary Hydrogen Cloud, COSMOS, Swinburne Astronomy, http://astronomy.swin.edu.au/cosmos/C/Cometarv+Hvdrogen+Cloud accessed on 12 Oct. 2015. 5. Gray, Robert; Marvel, Kevin (2001). “A VLA Search for the Ohio State ‘Wow’”. ApJ, 546 (2001) pp. 1171-1177 http://www.bigear.org/Gray-Marvel.pdf accessed on 4 Nov. 2015. 6. Chaisson, Eric, and McMillan, Steve. (2005) Astronomy Today. 7th edition. Upper Saddle River, NJ. Pearson/Prenlice Hall pp. 458-459. Winter 2015 32 7. Gray, Robert H.; Ellingsen, Simon (2002). “A Search for Periodic Emissions at the Wow Locale”. ApJ, 578 (2002) pp. 967-971. 8. Palen, Stacy. (2012) Understanding Our Universe, 2nd edition, New York, W.W. Norton pp 228-230. 9. Lang, Kenneth R (2010). Hydrogen Cloud of a Comet. NASA ’s Cosmos. https://ase.tufts.edu/cosmos/view picture.asp?id=1291 accessed on 01 Sept. 2015. 10. Term, Joe. (2015) “Hendrik C. Van De Hulst. The Bruce Medalists”. http://phys- astro.sonoma.edu/brucemedalists/vandeHulst accessed on 01 Sept. 2015. 1 1. Comet Base Observations Catalogue for P/2008 Y2 (Gibbs); 266P/Christensen. http://cometbase.net/en/observation/index accessed 13 Sept. 2015. 12. Jet Propulsion Laboratory Small Bodies Database. Ephemerides and Orbital Solutions for P/2008 Y2 (Gibbs); 266P/Christensen http://ssd.ipl.nasa.gov/sbdb.cgi accessed on 13 Sept. 2015. 13. Private Communication, (2015) Childers, Russ, Chief Observer at the OSU Radio Observatory, 1989-1997. Bios Antonio Paris is a Professor of Astronomy at St. Petersburg College, FL; the Director of Planetarium and Space Programs at the Museum of Science and Industry in Tampa, FL; and the Chief Scientist at the Center for Planetary Science - a science outreach program promoting astronomy, planetary science, and astrophysics to the next generation of space explorers. He is a member of the Washington Academy of Sciences, the American Astronomical Society, the St. Petersburg Astronomy Club, FL; and the author of two books, Aerial Phenomena and Space Science. Evan Davies is a fellow of both the Royal Geographical Society and The Explorers Club, and his popular space science writing has appeared in Wiley publications as well as Archaeology and Spaceflight magazines. He is the author of Emigrating Beyond Earth: Human Adaptation and Space Colonization and has held a lifelong interest in space exploration. Washington Academy of Sciences 33 Affine Geometry, Planck Length and Cosmic Acceleration George L. Murphy Tallmadge, OH Abstract In Ihe 1940s Schrodinger developed a generalization of Einstein’s metric gravitational theory based on a purely affine geometry. Today there are some reasons to give this theory renewed attention. First, it is another step along the path that Einstein pioneered in abandoning a priori assumptions about the geometry of the world. Second, Schrodinger’ s theory offers the prospect of dealing with the breakdown of the metric concept at the Planck scale while retaining the continuum. And third, the requirement that the cosmological constant cannot vanish in this theory means that the cosmic acceleration which has recently been discovered can be included in a natural way with this approach, and that the problem of a large vacuum energy can be resolved. Introduction A scientific theory that does not predict novel phenomena or correlate known facts better than its competitors will be relegated to the history of science museum. It may, however, return to active duty if it helps to explain new data. Today Schrodinger’ s affine field theory from the 1940s deserves such reconsideration. 1 It may help to explain cosmic acceleration and provide a basis for quantized gravitation as a result of an advance beyond metric-based general relativity. This theory received inadequate attention, or the wrong type of attention, when it was proposed. Many physicists considered it to be only a variant of a theory Einstein was then developing in which a non-symmetric part was added to the metric tensor.2 Today we can see that a theory in which metric is a secondary concept can explore territory that is closed to a theory in which it is fundamental. When Schrodinger proposed his theory many relativists viewed the cosmological term negatively. Pauli, for example, rejected it because it required a non-vanishing cosmological constant. 3 Now we know that cosmic expansion is accelerating in a way that is compatible with a cosmological term in the gravitational field equations.4 W inter 20 1 5 34 Attempts to extend general relativity like those of Schrodinger and Einstein were also burdened by expectations that they could be unified field theories encompassing all physical phenomena. Einstein’s hope that a successful theory of this type would eliminate what he saw as unattractive features of quantum mechanics gave many physicists the impression that the whole line of work was essentially reactionary.5 Here we eschew any expectation that an affine theory can, by itself, provide a unified explanation of physical interactions. While it does generalize Einstein’s theory of gravitation, its main interest is that it provides a broader geometric framework for further work. We will begin by considering reasons for generalizing the geometry that Einstein used in his 1915 theory, and then explore the possibilities connected with quantum gravity and dark energy. Is Riemannian Geometry Necessary? The name “geometry”, from Greek words for “earth” and “measure”, shows the discipline’s origin in practical concerns. But geometry also became the theoretical system of Euclid. It was long assumed that this system described the world correctly, and Kant’s view that our minds must perceive the world that way put a sophisticated seal on the idea.6 But failures to prove Euclid’s parallel postulate finally moved mathematicians to realize that it could be replaced by another assumption, and thus to develop non-Euclidean geometries.7 Further progress resulted in the Riemannian differential geometry that Einstein used in his gravitational theory. He did not impose a global geometry a priori but made the local character of space-time something to be determined by physical measurements. Geometry and physics were united, an achievement that Weyl symbolized in the equation “Pythagoras + Newton = Einstein.”8 General relativity uses a metric geometry: The local properties of space-time are completely specified by the metric tensor gov. It is, however, possible to consider more general differential geometries, as Weyl did within three years of the introduction of Einstein’s theory in an attempt to include electromagnetism.9 Other generalizations then followed.10 Attempts to extend Einstein’s unification of physics and geometry are viewed most helpfully in the spirit of Klein’s Erlanger Programme Here different geometries are considered in terms of the transformation Washington Academy of Sciences 35 groups that they allow. We can begin with a topological space whose meaningful properties are invariant under all continuous transformations, and then specialize by limiting this group, allowing new properties to emerge. The concepts of lines and points as their intersections are invariant under projective transformations and are therefore meaningful in projective geometry. A structure of parallelism can be added to yield affine geometry, and then metric properties can be introduced. By assuming a metric geometry of space-time we add concepts and specialize. We need not use a geometry more complex than necessary for physics, but insistence that the geometry of the world must be Riemannian is in the same spirit as the old idea that the geometry of the world must be Euclid’s. These arguments will appeal to those who believe that Einstein’s geometric view of gravitation was fundamental. Physicists who see it as just one way of interpreting his field equations will question the value of, for example, allowing torsion, a skew-symmetric part of the affine connection. This seems to be, as Weinberg argues, “just a tensor,” just one more field.12 But in generalizing the geometry of Einstein’s gravitational theory we are not adding torsion but removing conditions on the geometry that imply the vanishing of torsion. The fundamental question for physics, however, is whether a geometric approach will facilitate our understanding of our observations of the world. When we consider that possibility we will see some of the concrete advantages offered by an affine theory. Schrodinger’s Affine Theory The basic assumption of this theory is a four-dimensional manifold with an affine connection T^v which is not assumed to be symmetric and which gives the change in a vector A tt when displaced by a coordinate increment dxv: 8 A A = -r“vAA dxv . In the usual way, the covariant derivative of a vector field A A can be defined as Af'a = AMa + T^aAA and a curvature tensor can be constructed as shown by Equation (1): Winter 2015 36 dP - pA _ p/2 i p// p« _ p// p« ovp 1 o\\p 1 opy ' 1 apv ov avl op' (i) This has two independent contractions. The first, given by Equation (2), Ra* = Kp r7' CJV,fl rM op,v + va — ' 1 apl o\’ A av 1 op '■ (2) is a generalization of the Ricci tensor of Riemannian geometry. The second, given by Equation (3): o = bu =rw -r;/ (3) pov poy pv,o"> V ' which vanishes in a Riemannian space, is completely antisymmetric. Metric concepts have not yet been introduced and there is no way to compare lengths along different curves or to measure angles. We know, however, that at some scales the concepts of length and time are meaningful in our universe. In order to represent them we need to have a metric tensor, a symmetric second rank tensor gm/ , to define a magnitude ds associated with a displacement dx A via the generalization of the Pythagorean theorem ds 2 = gfJvdxfJdxv . If g is not to be a foreign body within the theory then we must use a symmetric second rank tensor that has already been defined, and the only possibility at this stage is the symmetric part of Rm,. Thus gm/ must be proportional to R(av). We can write this relationship more suggestively as Equation (4): ^«n.) = Agov W with A a constant. If ds is to have the same dimension as dx , dimensional analysis with powers of dxA shows that R(av) has dimension -2 so that A must also have dimension -2. In metric language it must have the dimensions of an inverse length squared. The similarity between Eq. (4) and Einstein’s vacuum equations with a cosmological term can hardly be missed. (The symbol A was not chosen at random.) To see the significance of this more fully, we need to look briefly at Schrodinger’s formulation of what he called “the final affine field laws.” Washington Academy of Sciences 37 To obtain dynamical equations for the field variables T^v we use Hamilton’s principle with an appropriate Lagrange density. At this point we have no way to raise and lower indices, so the simplest density can be written as Eq. (5): L = (2 / A) (5) where A is a constant which will give the action the proper dimensions. (It is also possible to use R + (l/4)£? in Eq. (5) to achieve projective invariance.)13 Following Schrodinger, we define the tensor density g"1 by 9L / 9R„„ = g"1' (6) and form the covariant and contravariant tensors g and gftv associated with it. The Euler-Lagrange equations of the variational principle then give Eq. (7): g^-*r^g«v-*r^,9^ = °. (7) where *r" , Schrodinger’s “star affinity”, is an abbreviation for C + (2 / 3)S“Yp and Y „ is (1/2) PP torsion. T^a T^a pa ap a contraction of the The defining Eq. (5) is equivalent to Eq. (8) so that Eq. (6) can be written as an equation involving only the affinity and the contracted curvature tensor. In the limiting case in which is symmetric, Eq. (6) is the equation satisfied by the metric tensor in Riemannian geometry. It can be solved to give the as functions of g and its derivatives, the Christoffel brackets. R is then the symmetric Ricci tensor of general relativity, and Eq. (4) and Eq. (7) are seen to be identical, the vacuum Einstein equations with a cosmological term. Winter 2015 38 The Demise of Metric The metric concept is meaningful in our universe only at sufficiently large scales. The uncertainty principle of quantum theory and the effect of gravitation on clock rates, together with limited resolution of a clock due to its size, show that below certain limits time intervals cannot be measured.14 To measure a time interval we must use a clock. The time-energy uncertainty relation tells us that the time t taken for the measurement and the uncertainty in the clock’s energy must satisfy tAE > fi . On the other hand, the gravitational field of the clock’s energy of a distance R will result in a change in a time interval t given by St / 1 ~ GE / c4R , where E includes both the original rest energy of the clock and AE. t can be no larger than R/c if the parts of the clock are to communicate with one another. When these results are combined we find that St > GJi / c5t . Since the measurement is of no value unless St 1 + 3ti/2.20 If the connection coefficients were represented by such functions then their behavior in their function space would be somewhat like Brownian motion or turbulence. Suppose that this were the case for sufficiently small regions of space-time. (“Small” can only be defined from the outside, since there is no metric inside these regions.) The metric concept would break down if attempts were made to explore such regions, and we have already seen that that is actually the case for time and length scales below the Planck values. On the other hand, scale invariance would be broken by the presence of the length |A| ' “ . We will consider question of the spatio-temporal scale at which the metric concept fails when we discuss cosmology in the next section. Palmer has proposed a new approach to quantum theory based on the ideas “that states of physical reality belong to, and are governed by, a non-computable fractal subset I of state space, invariant under the action of some subordinate deterministic causal dynamics D” and that “gravity plays a key role in generating the fractal geometry off”21 An affine theory seems to have the potential to provide such a geometry. But whether that proposal is pursued, it seems clear that an affine approach can ensure that one of the basic requirements for an adequate theory of quantum gravity is satisfied. Winter 2015 40 The Necessity of a Cosmological Term The way in which we have led up to Eq. (8) brings out the inevitable appearance of the cosmological term. It is instructive to proceed this way because that term has been the object of controversy. Einstein added this term to his original equations to make a static universe possible and then dropped it when it was found that the universe is expanding. Between Hubble’s discovery of cosmic expansion in 1929 and the realization in the late 1990s that this expansion was indeed speeding up, many workers in general relativity and cosmology dismissed the cosmological term. A number of theoretical models of dark energy have been proposed to account for the acceleration of cosmic expansion, but present data are compatible with Einstein's 1917 cosmological term.22 The simplest solution of Eq. (8) is the well-known de Sitter metric which can be written as Eq. (9): ds2 =-dt2 + exp(2Ht)dl2 , (9) where c = 1 now (so that length and time scales are identical), dl is the Euclidean spatial line element and H = (A/3) “ is the Hubble constant. The fact that the purely affine theory requires a cosmological term must count in its favor as prediction of a “novel fact,” one that was not assumed in the formulation of the theory. Most other dark energy models have been introduced precisely to explain cosmic acceleration and cannot therefore count it as a prediction. Quantum theory, however, presents the problem of a cosmological constant that is far too large. The cosmological term in the gravitational field equations has the form of a stress-energy tensor for a fluid whose pressure is negative and equal in magnitude to its energy density, and the vacuum energy of quantum fields has just this form. This effective cosmological constant obtained from quantum field theory cannot be reconciled with observations. Zel’dovich’s23 calculation of the vacuum stress-energy tensor for an assortment of boson and fermion fields with a cutoff on the order of the proton’s Compton wavelength gave a value 44 orders of magnitude larger than what observations at that time would allow. A more fundamental cutoff is defined by the Planck length. Washington Academy of Sciences 41 The resulting vacuum energy in Einstein’s equations gives a model universe that, according to Eq. (9), expands exponentially with a characteristic time on the order of T* ~ 10 43 s. Of course this is unacceptable. But things could be different when we consider this vacuum energy in the context of affine theory. The vacuum energy density p gives rise to an effective cosmological constant 87 ip, and if we were combine this with A to give an effective cosmological constant we get Eq. (10): A' = A + 8 np (10) then we could replace Eq. (8) with Eq. (11): R0V=A'SoV (11) We could then choose the value of A to “renormalize” the effective cosmological constant to a value A' in accord with observations, as Shifflett has suggested.24 Since p has a large positive value, A would have to have a negative value of nearly the same magnitude. Choosing such a value for A is not completely arbitrary. The cosmological constant defines a fundamental length and time |A| ' ~ which would break the scale invariance of the fractal regime discussed in the previous section. If that is close to the cutoff for calculation of vacuum energy then A and A' would be of the same order of magnitude. Affine theory has a universal standard of length, |A| 1 “ . In the 1930s Eddington gave this as a reason to retain “the cosmical constant” even after the original motive for it had disappeared.25 It seemed obvious then that this length would be of cosmological size. The idea that there are two fundamental lengths, one provided by A and the other defined by the fundamental constants h,c, and G, might have raised suspicions if the physical significance of the Planck length had been given more attention at that time. We can see now the possibility that the two fundamental lengths are approximately the same. Winter 20 15 42 Prospects for Further Progress Eqs. (7) and (8) are the field equations of Schrodinger’s theory, which reduce to Einstein’s vacuum equations with a cosmological term in the limiting case of a symmetric connection. Since we are not pursuing a unified field theory we can describe other fields and particles by introducing non-geometric variables A and a Lagrange density Lm which will depend on the ®A, their derivatives and g which, in turn, is a function of T^v and its derivatives via Eq. (8).26 If Lm is added to Eq. (5) and the procedures of Hamilton’s principle are earned through then the full gravitational field equations with an energy-momentum tensor defined in the standard way can be obtained if that tensor is small in comparison with the cosmological term. The energy densities of baryonic and dark matter are much smaller than the magnitude of the A that we have hypothesized to renormalize quantum vacuum energy, so this approximation makes sense for those forms of matter. However, the vacuum energy itself is comparable in magnitude with A. So while this approximation method has some value, it does not enable us to shed any light on features of quantum field theory such as ultraviolet divergences. We have taken advantage of the possibility of a connection that is not differentiable to suggest that the metric concept might break down below some space-time scale, and suggest that this could be correlated with the implication of the uncertainty principle and the gravitational effect on clock rates that intervals below the Planck scale cannot be measured. However, this would also mean that the Ricci tensor, which is used to form the Lagrangian, involves derivatives of the connection and would not be defined. The classical Hamilton’s principle could no longer be used to derive field equations. This is not surprising if the metric concept indeed fails below the Planck scale. One way to proceed would be to look for an algebraic expression which approximates the classical action Eq. (5) on scales for which the latter is meaningful. We could then use this action in Feynman’s “sum over histories” approach to quantum theory in order to explore the implications of the affine theory further. Washington Academy of Sciences 43 For now we have barely a hint of ways in which an adequate quantum theory of matter and non-gravitational phenomena might be developed on the basis of affine geometry. This approach does, however, seem capable of showing in a natural way how the metric concept may fail below the Planck scale without abandoning the continuum concept. In addition, the requirement that the cosmological constant not vanish and the possibility that the theory can deal with the problem of a huge quantum vacuum energy means that the observed cosmic acceleration can be explained in an unforced manner. The long-dormant theory that Schrodinger proposed in the 1940s seems today to have some promise. References 1 E. Schrodinger, Proc. R. Irish Acad. 51 A, 163 (1947); 51 A, 205 (1948); 52 A, 1(1948). These are summarized in E. Schrodinger, Space-Time Structure, Cambridge, (1963), Chapter XII. 2 A. Einstein, The Meaning of Relativity, 5th ed., Princeton, (1956), Appendix II, is the final version of Einstein’s mixed affine-metric theory. 3 W. Pauli, Theoiy of Relativity, Pergamon (1958), p.225. The dismissive footnote in L.D. Landau and E.M Lifshitz, The Classical Theoiy of Fields, revised 2d ed., Addison- Wesley, (1962) may also be noted. 4 S.W. Allen et al., Mon. Not. R. Astron. Soc. 383, 879 (2008); Vikhlinin, A. et a!.: Astrophys. J. 692, 1060 (2009). 5 Schrodinger’s attitude toward this issue was more nuanced than Einstein’s. See, e.g., his comment on p. 160 of Space-Time Structure. 6 P. Guyer, Kant and the Claims of Knowledge, Cambridge (1987), pp. 359-360. 7 C.B. Boyer and U.C. Merzbach, A Histoiy of Mathematics, 2d ed., John Wiley, New York (1989), pp.580-583. 8 H. Weyl, Space-Time-Matter, 4th ed., Dover, (1950), p.228. Weyl represented Einstein as a bracketing of Pythagoras and Newton but I have written this as an equation to simplify typography. 9 Ibid., pp.282-312. 10 A.S. Eddington, The Mathematical Theoiy of Relativity, 2d ed., Cambridge, (1924). Chapter VII and Supplementary Notes 13 and 14 provide a survey of some of this work. 1 1 Boyer and Merzbach, A History of Mathematics, pp.6 11-613. Winter 2015 44 12 S. Weinberg, Physics Today 59.4, 16 (2006). For Weinberg’s more extended statement on the geometric approach to general relativity see his Gravitation and Cosmology >, John Wiley (1972), pp.vi - viii and 147. 13 G.L. Murphy, Phys. Rev. Dll, 2752(1975). 14 G.L. Murphy, Am. J. Phys. 42, 958(1974). 13 M. Planck, Sitzungsber. Dent. Akad. Wiss. Berlin. Kl. Math. - Phys. 440-480 (1899). 16 B. Riemann, Abhandhmgen d. K. Gesells. zu Gottingen 13, 133 (1854). 17 R. Penrose, The Road to Reality: A Guide to the Laws of the Universe, Alfred A. Knopf, (2005), pp. 958-962. The negative result of one search for such structure is reported at http://arxiv.org/abs/1512.01216 . Is B.S. DeWitt, ’’The Quantization of Geometry”, In L. Witten, (ed.) Gravitation: An Introduction to Current Research, John Wiley, (1962), pp. 372-373. 19 B.J. West, M. Bologna, P. Grigolini, Physics of Fractal Operators, Springer, (2003), Chapter 1. 20 K. Weierstrass, Abhandhmgen aus der Functionenlehre, Springer, (1886), p.97; http://planetmath.org/encyclopedia/WeierstrassFunction.html . 21 T.N. Palmer, Proc. Roy. Soc. A465, 3165 (2110). 22 Cf. endnote 4. 23 Y.B. Zel’dovich, Sov. Phys. Usp. 11, 381 (1968). 24 J.A. Shifflett Gen. Rel. & Grav. 40, 1745 (2008); 41, 1865 (2009). 25 A. Eddington, The Expanding Universe, Macmillan (1933), p. 148. 26 G.L. Murphy, “An Affine Approach to General Relativity”, presented at the Third Australasian Conference on General Relativity, University of Adelaide, Adelaide, South Australia, 30 January 1976. Bio George L. Murphy received a Ph.D. in physics from Johns Hopkins University in 1972 and has published physics papers primarily on aspects of general relativity and cosmology. After teaching in colleges for 12 years he began theological study, receiving a M.Div. from Wartburg Seminary, and served as a parish pastor until retirement. He has written extensively on science-theology relationships. Washington Academy of Sciences 45 Membership List for 2015 M : Member, LM — Life Member, F = Fellow, LF = Life Fellow, EM = Emeritus Member, EF = Emeritus Fellow ABEL, DAVID (Dr.) 1 13 Hedgewood Drive, Greenbelt MD 20770-1610 (LM) ADKINS, MICHAEL K (Mr.) 4143, Elizabeth Lane, Annandale VA 22003 (M) ANTMAN, STUART (Dr.) University of Maryland, 2309 Mathematics Building, College Park MD 20742-4015 (EF) APPETITI, EMANUELA PO Box 25499, Washington DC 20027 (LM) APPLE, DA1NA DRAVNIEKS National Capital Society of American Foresters, PO Box 9288, Arlington VA 22219 (M) ARLE, KATHY (Ms.) 3049 Heather Lane, Falls Church VA 22044 (M) ARSEM, COLLINS (Mr.) 3144 Gracefield Rd Apt 117, Silver Spring MD 20904- 5878 (EM) ARVESON, PAUL T. (Mr.) 6902 Breezewood Terrace, Rockville MD 20852-4324 (F) BARBOUR, LARRY L. (Mr.) Pequest Valley Farm, 585 Townsbury Road, Great Meadows NJ 07838 (M) BARWICK, W. ALLEN (Dr.) 13620 Maidstone Lane, Potomac MD 20854-1008 (F) BEAM, WALTER R. (Dr.) 4804 Wellington Farms Drive, Chester VA 23831 (F) BECKER, EDWIN D. (Dr.) Bldg. 5, Rm. 128, Nat. Institutes Of Health, Bethesda MD 20892-0520 (EF) BEMENT, ARDEN (Dr.) National Science Foundation, 4201 Wilson Boulevard, Arlington VA 22230 (EF) BERLEANT, DANIEL (Dr.) 12473 Rivercrest Dr., Little Rock AR 72212 (M) BERRY, JESSE F. (Mr.) 2601 Oakenshield Drive, Rockville MD 20854 (M) BIGLARI, HAIK (Dr.) Sr. Director of Engineering, Fairchild Controls, 540 Highland Street, Frederick MD 21701-7672 (M) BIONDO, SAMUEL J. (Dr.) 10144 Nightingale St., Gaithersburg MD 20882 (EF) BODSON, DENNIS (Dr.) 233 N. Columbus Street, Arlington VA 22203 (F) BOISVERT, RONALD F. (Dr.) Mail Stop 8910, Nat. Institute of Standards and Technology, 100 Bureau Drive, Gaithersburg MD 20899-8910 (F) BOYCE, GEORGE (Mr.) 7 Greenway PL, Greenbelt MD 20070 (M) BRISKMAN, ROBERT D. (Mr.) 61 Valerian Court, North Bethesda MD 20852 (F) BUFORD, MARILYN (Dr.) 3073 White Birch Court, Fairfax VA 22031 (F) CASTILLO, YOLANDA F. (Ms.) 1717 Pennsylvania Ave NW Ste 1025, Washington DC 20008-3951 (M) CAWS, PETER J (Dr.) Apt. 230, 2475 Virginia Ave. N.W., Washington DC 20037 (M) CHRISTMAN, GERARD (Mr.) 6109 Berlee Drive, Alexandria VA 22312 (F) CIORNEIU, BORIS (Dr.) 20069 Great Falls Forest Dr., Great Falls VA 22066 (M) CIUCA, LIVIU BOGDEN (Mr.) Aleea Scolii, no.2, Galati, Bucharest, Romania (M) CLINE, THOMAS LYTTON (Dr.) 13708 Sherwood Forest Drive, Silver Spring MD 20904 (F) Winter 2015 46 COFFEY, TIMOTHY P. (Dr.) 976 Spencer Rd., McLean VA 22102 (F) COHEN, MICHAEL P. (Dr.) 1615 Q. St. NW T-l, Washington DC 20009-6310 (LF) COLE, JAMES H. (Mr.) 9709 Katie Leigh Ct, Great Falls VA 22066-3800 (F) CURRIE, S.J., CHARLES L. (Rev.) Jesuit Community, Georgetown University, Washington DC 20057 (EF) DANCKWERTH, DANIEL 419 Beach Drive, Annapolis MD 21403-3906 (M) DANNER, DAVID L. (Dr.) 1364, Suite 101, Beverly Road, McLean VA 22101 (F) DAVIS, ROBERT E. (Dr.) 1793 Rochester Street, Crofton MD 21114 (F) DEAN, DONNA (Dr.) 367 Mound Builder Loop, Hedgesville WV 25427-7211 (EF) DEDRICK, ROBERT L. (Dr.) 21 Green Pond Rd, Saranac Lake NY 12983 (EF) DIMITOGLOU, GEORGE (Dr.) 11053 Seven Hill Lane, Potomac MD 20854 (M) DONALDSON, JOHANNA B. (Mrs.) 3020 North Edison Street, Arlington VA 22207 (EF) DURRANI, SAJJAD (Dr.) 17513 Lafayette Dr, Olney MD 20832 (EF) EDINGER, STANLEY EVAN (Dr.) Apt #1016, 5801 Nicholson Lane, North Bethesda MD 20852 (EM) EPHRATH, ARYE R. (Dr.) 5467 Ashleigh Rd., Fairfax VA 22030 (M) ERICKSON, TERRELL A. (Ms.) 4806 Cherokee St., College Park MD 20740-1865 (M) ETTER, PAUL C. (Mr.) 16609 Bethayres Road, Rockville MD 20855 (F) FASANELLI, FLORENCE (Dr.) 471 1 Davenport Street, Washington DC 20016 (EF) FAULKNER, JOSEPH A. (Mr.) 2 Bay Drive, Lewes DE 19958 (F) FRANKLIN, JUDE E. (Dr.) 7616 Carteret Road, Bethesda MD 20817-2021 (F) FRASER, GERALD (Dr.) 581 1 Cromwell Drive, Bethesda MD 20816 (M) FREEMAN, ERNEST R. (Mr.) 5357 Strathmore Avenue, Kensington MD 20895- 1160 (ELF) GAGE, DOUGLAS W. (Dr.) XPM Technologies, 1020 N. Quincy Street, Arlington VA 22201-4637 (M) GAITHER, VERLESE P. (Ms.) 10301 Musket Court, Fort Washington MD 20744 (M) GAUNAURD, GUILLERMO C. (Dr.) 4807 Macon Road, Rockville MD 20852- 2348 (EF) GIBBON, JOROME (Mr.) 3 1 1 Pennsylvania Avenue, Falls Church VA 22046 (F) GIBBONS, JOHN H. (Dr.) Resource Strategies, P.O. Box 379, The Plains VA 20198 (EF) GIFFORD, PROSSER (Dr.) 59 Penzance Rd, Woods Hole MA 02543-1043 (F) GIORDANO, JAMES (Dr.) Neuroethics Studies Program, Pellegrino Center for Clinical Bioethics, Georgetown University Med Center, Washington DC 20057 (M) GLUCKMAN, ALBERT G. (Mr.) 18123 Homeland Drive, Olney MD 20832-1792 (EF) GOLDSTON, DAVID (Mr.) 816 N. Highland St., Arlington VA 22201 (M) Washington Academy of Sciences 47 GRAY, MARY (Prof.) Department of Mathematics, Statistics, and Computer Science, American University, 4400 Massachusetts Avenue NW, Washington DC 20016-8050 (F) GRIER, REBECCA (Dr.) 6300 Stevenson Ave #501, Alexandria VA 22304 (M) HACK, HARVEY (Dr.) 176, Via Dante, Arnold MD 21012-1315 (F) HAIG, SJ, FRANK R. (Rev.) Loyola University Maryland, 4501 North Charles St, Baltimore MD 21210-2699 (F) HARR, JAMES W. (Mr.) 180 Strawberry Lane, Centreville MD 21617 (EF) HAZAN, PAUL 14528 Chesterfield Rd, Rockville MD 20853 (F) HEANEY, JAMES B. 6 Olive Ct, Greenbelt MD 20770 (M) HENDEE, JAMES (Dr.) 51 1 SE 13th Court, Pompano Beach FL 33060 (M) HERBST, ROBERT L. (Mr.) 4109 Wynnwood Drive, Annadale VA 22003 (LF) H1BBS, EUTHYMIA D. (Dr.) 7302 Durbin Terrace, Bethesda MD 20817 (M) HIETALA, RONALD (Dr.) 6351 Waterway Drive, Falls Church VA 22044-1322 (M) HILL II, RICHARD E. (Mr.) 4360 Lee Hwy #204, Arlington VA 22207 (M) HOFFELD, J. TERRELL (Dr.) 1 1307 Ashley Drive, Rockville MD 20852-2403 (F) HOLLAND, PH.D., MARK A. (De.) 201 Oakdale Rd., Salisbury MD 21801 (M) HOLLINSHEAD, ARIEL (Dr.) 23465 Harbor View Rd. #622, Punta Gorda FL 33980-2162 (EF) HONIG, JOHN G. (Dr.) 7701 Glenmore Spring Way, Bethesda MD 20817 (LF) HORLICK, JEFFREY (Mr.) 8 Duvall Lane, Gaithersburg MD 20877-1838 (F) HOROWITZ, EMANUEL (Dr.) Apt 618, 3100 N. Leisure World Blvd, Silver Spring MD 20906 (EF) HOROWITZ, SHARYN (Ms.) 217 Katie Court, Falls Church VA 22046 (M) HOWARD, SETH ANNE (Dr.) 5526 Green Dory Lane, Columbia MD 21044 (LF) HOWARD-PEEBLES, PATRICIA (Dr.) 7501 Virginia Parkway 2312, McKinney TX 75071 (EF) HURDLE, BURTON G. (Dr.) 3440 South Jefferson St., Apt. 356, Falls Church VA 22041 (F) IKOSSI, KIKI (Dr.) 6275 Gentle LN, Alexandria VA 22310 (F) JACOX, MARILYN E. (Dr.) 16665 Concord St Ste 206, Kenington MD 20895-2525 (F) JARRELL, H. JUDITH (Dr.) 9617 Alta Vista Ter., Bethesda MD 20814 (F) JAYARAO, ARUNDHATI (Dr.) 8811 Trafalgar Ct, Springfield V A 22151 (M) JOHNSON, EDGAR M. (Dr.) 1384 Mission San Carlos Drive, Amelia Island FL 32034 (LF) JONG, SHUNG-CHANG (Dr.) 8892 Whitechurch Ct, Bristow VA 20136 (LF) JORDANA, ROMAN DE VICENTE (Dr.) Batalla De Garellano, 15, Aravaca, 28023, Madrid, Spain (EF) KADTKE, JAMES (Dr.) Apt. 824, 1701 16th St. NW, Washington DC 20009-3131 (M) KAHN, ROBERT E. (Dr.) 909 Lynton Place, Mclean VA 22102 (F) KAPETANAKOS, C. A. (Dr.) 443 1 MacArthur Blvd, Washington DC 20007 (EF) KATZ, ROBERT (Dr.) 3310 N. Leisure Blvd #530, Silver Spring MD 20906 (EF) Winter 2015 48 KAUFHOLD, JOHN (Dr.) Suite 1200, 4601 N. Fairfax Dr, Arlington VA 22203 (M) KAY, PEG (Ms.) 6111 Wooten Drive, Falls Church VA 22044 (LF) KEEFER, LARRY (Dr.) 7016 River Road, Bethesda MD 20817 (F) KE1SER, BERNHARD E. (Dr.) 2046 Carrhill Road, Vienna VA 22181-2917 (LF) KLINGSBERG, CYRUS (Dr.) Apt. LI 84, 500 E. Marylyn Ave, State College PA 16801-6225 (EF) KLOPFENSTEIN, REX C. (Mr.) 4224 Worcester Dr., Fairfax VA 22032-1 140 (LF) KRUEGER, GERALD P. (Dr.) Krueger Ergonomics Consultants, 4105 Komes Court, Alexandria VA 22306-1252 (EM) KUO, CHUN-HUNG (Dr.) 4637 Knight Place, Alexandria VA 223 1 1 (M) LAWSON, ROGER H. (Dr.) 10613 Steamboat Landing, Columbia MD 21044 (EF) LEDGER, SAM (Mr.) 420 7th Street NW, Apt 903, Washington DC 20004 (M) LE1BOWITZ, LAWRENCE M. (Dr.) 3903 Laro Court, Fairfax VA 2203 1 (LF) LEMKIN, PETER (Dr.) 148 Keeneland Circle, North Potomac MD 20878 (EM) LESHUK, RICHARD (Mr.) 9004 Paddock Lane, Potomac MD 20854 (M) LEWIS, DAVID C. (Dr.) 27 Bolling Circle, Palmyra VA 22963 (F) LEWIS, E. NEIL (Dr.) Malvern Instruments, Suite 300, 7221 Lee Deforest Dr, Columbia MD 21046 (M) LIBELO, LOUIS F. (Dr.) 9413 Bulls Run Parkway, Bethesda MD 20817 (LF) LONDON, MARILYN (Ms.) 3520 Nimitz Rd, Kensington MD 20895 (F) LONGSTRETH, III, WALLACE I (Mr.) 8709 Humming Bird Court, Laurel MD 207231254 (M) LOOMIS, TOM H. W. (Mr.) 11502 Allview Dr., Beltsville MD 20705 (EM) LUTZ, ROBERT J. (Dr.) 6031 Willow Glen Dr, Wilminton NC 28412 (EF) LYON, HARRY B. (Mr.) 7722 Northdown Road, Alexandria VA 22308-1329 (M) LYONS, JOHN W. (Dr.) 7430 Woodville Road, Mt. Airy MD 21771 (EF) MALCOM, SHIRLEY M. (Dr.) 12901 Wexford Park, Clarksville MD 21029-1401 (F) MANDERSCHEID, RONALD W. (Dr.) 10837 Admirals Way, Potomac MD 20854- 1232 (LF) MASON, LANCE (Dr.) 1212 Calla Cerrito, Santa Barbara CA 93101 (M) MCFADDEN, GEOFFREY B (Dr.) 20117 Darlington Drive, Montgomery Village MD 20886 (M) MENZER, ROBERT E. (Dr.) 90 Highpoint Dr, Gulf Breeze FL 32561-4014 (EF) MESSINA, CARLA G. (Mrs.) 9800 Marquette Drive, Bethesda MD 20817 (F) METAILIE, GEORGES C. (Dr.) 18 Rue Liancourt, 75014 Paris, FRANCE (F) MIELCZAREK, EUGENIE A. (Dr.) 3181 Readsborough Ct, Fairfax VA 2203 1 - 2625 (F) MILLER, JAY H. (Mr.) 8924 Ridge Place, Bethesda MD 20817-3364 (M) MILLER II, ROBERT D. (Dr.) The Catholic University of America, 10918 Dresden Drive, Beltsville MD 20705 (M) MORGOUNOV, ALEXEY (Dr.) CIMMYT, P.K. 39, Emek, Ankara 0651 1, Turkey (M) MORRIS, JOSEPH (Mr.) PO Box 3005, Oakton VA 22124-9005 (M) Washington Academy of Sciences 49 MORRIS, P.E., ALAN (Dr.) 4550 N. Park Ave. #104, Chevy Chase MD 20815 (EF) MOUNTAIN, RAYMOND D. (Dr.) 701 King Farm Blvd #327, Rockville MD 20850 (F) MUMMA, MICF1AEL J. (Dr.) 210 Glen Oban Drive, Arnold MD 21012 (F) MURDOCH, WALLACE P. (Dr.) 65 Magaw Avenue, Carlisle PA 17015 (EF) NORRIS, KARL H. (Mr.) 1 1204 Montgomery Road, Beltsville MD 20705 (EF) O'HARE, JOHN J. (Dr.) 108 Rutland Blvd, West Palm Beach FL 33405-5057 (EF) OHRINGER, LEE (Mr.) 5014 Rodman Road, Bethesda MD 20816 (EF) ORDWAY, FRED (Dr.) 5205 Elsmere Avenue, Bethesda MD 20814-5732 (EF) OTT, WILLIAM R (Dr.) 19125 N. Pike Creek Place, Montgomery Village MD 20886 (EF) PAJER, BERNADETTE (Mrs.) 25116 143rd St. SE, Monroe WA 98272 (M) PARIS, ANTONIO (Mr.) 25066 12th Ave N, St Petersburg FL 33713-5510 (M) PARR, ALBERT C (Dr.) 2656 SW Eastwood Avenue, Gresham OR 97080-9477 (F) PATEL, D. G. (Dr.) 1 1403 Crownwood Lane, Rockville MD 20850 (F) PAULONIS, JOHN (Mr.) Yonkers NY 10710 (M) PAULONIS, JOHN J (Mr.) P.O. Box 335, Yonkers NY 10710 (M) PAZ, ELVIRA L. (Dr.) 172 Cook Hill Road, Wallingford CT 06492 (ELF) PICKFJOLTZ, RAYMOND L. (Dr.) 3613 Glenbrook Road, Fairfax VA 22031-3210 (EF) PLESCIA, JEFFREY (Dr.) Applied Physics Laboratory, The Johns Hopkins University, MS 200-W230 11100 Johns Hopkins Road, Laurel MD 20723-6099 (M) POLAVARAPU, MURTY 10416 Hunter Ridge Dr., Oakton VA 22124 (LF) POLINSKi, ROMUALD (Mr.) 01-201 WARSZAWA, UL., Wolska 43, Poland (M) PRZYTYCKI, JOZEF M. (Prof.) 10005 Broad St, Bethesda MD 20814 (F) PYKE, JR, THOMAS N. (Mr.) 4887 N. 35th Road, Arlington VA 22207 (F) RADER, CHARLES A. (Mr.) 1101 Paca Drive, Edgewater MD 21037 (EF) RAMAKER, DAVID E. (Dr.) 6943 Essex Avenue, Springfield VA 22150 (EF) RAVITSKY, CHARLES (Mr.) 37129 Village 37, Camarillo CA 93012 (EF) READER, JOSEPH (Dr.) National Institute of Standards and Technology, 100 Bureau Drive, MS 8422, Gaithersburg MD 20899-8422 (F) REAGAN, ANN M. (Dr.) PO Box 22, Lusby Maryland 20657 (M) REDISH, EDWARD F. (Prof.) 6820 Winterberry Lane, Bethesda MD 20817 (F) REISCHAUER, ROBERT (Dr.) 5509 Mohican Rd, Bethesda MD 20816 (EF) RENAUD, PHILIP (Capt.) Living Oceans Foundation, 8181 Professional Place Suite 215, Landover MD 20785 (M) RICH, PAUL (Dr.) 1527 New Hampshire Avenue, NW, Washington DC 20036 (M) RICKER, RICHARD (Dr.) 12809 Talley Ln, Darnestown MD 20878-6108 (F) RIDGELL, MARY P.O. Box 133, 48073 Mattapany Road, St. Mary's City MD 20686-0133 (LM) ROBERTS, SUSAN (Dr.) Ocean Studies Board, Keck 607, National Research Council, 500 Fifth Street, NW, Washington DC 20001 (F) ROGERS, KENNETH (Dr.) 355 Fellowship Circle, Gaithersburg MD 20877 (LM) Winter 2015 50 ROMAN, NANCY GRACE (Dr.) 4620 North Park Avenue Apt. 306W, Chevy Chase MD 20815 (F) ROOD, SALLY A (Dr.) PO Box 12093, Arlington VA 22219 (F) ROSENBLATT, JOAN R. (Dr.) 701 King Farm Blvd, Apt 630, Rockville MD 20850 (EF) SAUBERMAN, P.E., HARRY R (Mr.) 8810 Sandy Ridge Ct., Fairfax VA 22031 (M) SCHINDLER, ALBERT I. (Dr.) 6615 Sulky Lane, Rockville MD 20852 (EF) SCHMEIDLER, NEAL F. (Mr.) 7218 Hadlow Drive, Springfield VA 22152 (F) SCHNEPFE, MARIAN M. (Dr.) Potomac Towers, Apt. 640, 2001 N. Adams Street, Arlington VA 22201 (EF) SEVERINSKY, ALEX J. (Dr.) 4707 Foxhall Cres NW, Washington DC 20007-1064 (EM) SHAFRIN, ELAINE G. (Mrs.) 8168 Connecticut Ave NW Apt 2014, Washington DC 20815-817 (EF) SHETLER, STANWYN G. (Dr.) 142 E Meadowland Ln, Sterling VA 20164-1144 (EF) SHIELDS, EDWARD (Dr.) PO Box 165, Grand Portage Minnesota 55605 (M) SHROPSHIRE, JR, W. (Dr.) Apt. 426, 300 Westminster Canterbury Dr., Winchester VA 22603 (LF) SILBER, RONNIE (Dr.) 13710 Colgate Way #1338, Silver Spring MD 20904 (M) SIMMS, JAMES ROBERT (Mr.) 9405 Elizabeth Ct., Fulton MD 20759 (M) SMITH, THOMAS E. (Dr.) 3148 Gracefield Rd Apt 215, Silver Spring MD 20904- 5863 (LF) SNIECKUS, MARY (Ms.) 1700, Dublin Dr., Silver Spring MD 20902 (M) SODERBERG, DAVID L. (Mr.) 403 West Side Dr. Apt. 102, Gaithersburg MD 20878 (M) SOLAND, RICHARD M. (Dr.) 2516 Arizona Av Apt 6, Santa Monica CA 90404- 1426 (LF) SOZER, AMANDA (Dr.) 4707B Eisenhower Ave, Alexandria VA 22304 (M) SPILHAUS, JR, A.F. (Dr.) 10900 Picasso Lane, Potomac MD 20854 (EM) SRIRAM, RAM DUVVURU (Dr.) 1030 Castlefield Street, Ellicott City MD 21042 (LF) STEIN, DAVID E. (Mr.) P.O. Box 571433, Las Vegas NV 89157 (M) STERN, KURT H. (Dr.) 103 Grant Avenue, Takoma Park MD 20912-4328 (EF) STEWART, PETER (Prof.) 417 7th Street NE, Washington DC 20002 (M) STIEF, LOUIS J. (Dr.) 332 N St., SW., Washington DC 20024-2904 (EF) STOMBLER, ROBIN (Ms.) Auburn Health Strategies, 3519 South Four Mile Run Dr., Arlington VA 22206 (M) STRAUSS, SIMON W. (Dr.) 3160 Bracefield Rd. RC 1305, Silver Spring MD 20904 (LF) SYED, ALI (Dr) 603, H Street SW, Washington DC 20024 (M) TABOR, HERBERT (Dr.) NIDDK, LBP, Bldg 8, Rm 223, National Institutes of Health, Bethesda MD 20892-0830 (M) Washington Academy of Sciences 51 TAYSING-LARA, MONICA (Ms.) 3343 Deni Place NW, Washington DC 20007 (M) TEICH, ALBERT H. (Dr.) PO Box 309, Garrett Park MD 20896 (EF) THEOFANOS, MARY FRANCES (Ms.) 7241 Antares Drive, Gaithersburg MD 20879 (M) THOMPSON, DENNIS (Mr.) 325 Sandy Ridge Rd, Fredricksburg VA 22405 (M) THOMPSON, CHRISTIAN F. (Dr.) 278 Palm Island Way, Ponte Vedra FL 32081 (LF) TIMASHEV, SVIATOSLAV (SLAVA) A. (Dr. Prof.) 3306 Potterton Dr., Falls Church V A 22044-1603 (F) TOUWAIDE, ALAIN PO Box 25499, Washington DC 20027 (LF) TOWNSEND, MARJORIE R. (Mrs.) 3529 Tilden Street, NW, Washington DC 20008-3194 (LF) TROXLER, G.W. (Dr.) PO Box 1 144, Chincoteague VA 23336-9144 (F) TYLER, PAUL E. (Dr.) 1023 Rocky Point Ct. N.E., Albuquerque NM 87123-1944 (EF) UBELAKER, DOUGLAS H. (Dr.) Dept, of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington DC 20560-01 12 (F) UMPLEBY, STUART (Professor) The George Washington University, Apt 1207 4141 N Henderson Rd, Arlington VA 22203 (F) VAISHNAV, MARIANNE P. (Ms.) P.O. Box 2129, Gaithersburg MD 20879 (LF) VAN TUYL, ANDREW (Dr.) 3618 Littledale Road, Apt. 203, Kensington MD 20895-3434 (EF) VARADI, PETER F. (Dr.) Apartment 1606W, 4620 North Park Avenue, Chevy Chase MD 20815-7507 (EF) VAVRICK, DANIEL J. (Dr.) 10314 Kupperton Court, Fredericksburg VA 22408 (F) VIZAS, CHRISTOPHER (Dr.) 504 East Capitol Street, NE, Washington DC 20003 (M) WALDMANN, THOMAS A. (Dr.) 3910 Rickover Road, Silver Spring MD 20902 (F) WALLER, JOHN D. (Dr.) 5943 Kelley Court, Alexandria VA 22312-3032 (M) WEAR, DOUGLAS (Dr.) 8014 Barron Street, Takoma Park MD 20912 (M) WEBB, RALPH E. (Dr.) 21-P Ridge Road, Greenbelt MD 20770 (EF) WEIL, TIMOTHY (Mr.) SECURITYFEEDS, PO Box 18385, Denver CO 80218 (M) WEISS, ARM AND B. (Dr.) 6516 Truman Lane, Falls Church VA 22043 (LF) WERGIN, WILLIAM P. (Dr.) 1 Arch Place #322, Gaithersburg MD 20878 (EF) WHITE, CARTER (Dr.) 12160 forest Hill Rd, Waynesboro PA 17268 (F) WIESE, WOLFGANG L. (Dr.) 8229 Stone Trail Drive, Bethesda MD 20817 (EF) WILCOX, ROBIN J. (Ms.) 8601 Park Avenue, Bowie MD 20720 (M) WILLIAMS, CARL (Dr.) 2272 Dunster Lane, Potomac MD 29854 (F) WILLIAMS, E. EUGENE (Dr.) Dept, of Biological Sciences, Salisbury University, 1101 Camden Ave, Salisbury MD 21801 (M) Winter 2015 52 WITHERSPOON, F. DOUGLAS ASTI, 11316 Smoke Rise Ct., Fairfax Station VA 22039 (M) WU, KLELI (Mr.) 360, Suite 48, Swift Ave, South San Francisco CA 94080 (M) Washington Academy of Sciences 53 Washington Academy of Sciences 1200 New York Avenue Rm 113 Washington, DC 20005 Please fill in the blanks and send your application to the address above. We will contact you as soon as your application has been reviewed by the Membership Committee. Thank you for your interest in the Washington Academy of Sciences. (Dr. Mrs. Mr. Ms) Business Address Home Address Email Phone ===__ Cell Phone _^__________=======_===__^=__= preferred mailing address Type of membership Business Home Regular Student Schools of Higher Education attended Degrees Dates Present Occupation or Professional Position Please list memberships in scientific societies - include office held: Winter 2015 54 Instructions to Authors 1. Deadlines for quarterly submissions are: Spring - February 1 Fall - August 1 Summer - May 1 Winter - November 1 2. Draft Manuscripts using a word processing program (such as MSWord), not PDF. We do not accept PDF manuscripts. 3. Papers should be 6,000 words or fewer. If there are 7 or more graphics, reduce the number of words by 500 for each graphic. 4. Include an abstract of 150-200 words. 5. Include a two to three sentence bio of the authors. 6. Graphics must be in graytone, and be easily resizable by the editors to fit the Journal’s page size. Reference the graphic in the text. 7. Use endnotes or footnotes. The bibliography may be in a style considered standard for the discipline or professional field represented by the paper. 8. Submit papers as email attachments to the editor or to wasioumal@washacadsci.org . 9. Include the author’s name, affiliation, and contact information - including postal address. Membership in an Academy-affiliated society may also be noted. It is not required. 10. Manuscripts are peer reviewed and become the property of the Washington Academy of Sciences. 1 1 . There are no page charges. 12. Manuscripts can be accepted by any of the Board of Discipline Editors. Washington Academy of Sciences Washington Academy of Sciences Affiliated Institutions National Institute for Standards & Technology (NIST) Meadowlark Botanical Gardens The John W. Kluge Center of the Library of Congress Potomac Overlook Regional Park Koshland Science Museum American Registry of Pathology Living Oceans Foundation National Rural Electric Cooperative Association (NRECA) Winter 2015 56 Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies Acoustical Society of America, Washington Chapter American Assoc, of Physics Teachers, Chesapeake Section American Astronomical Society American Ceramics Society American Fisheries Society American Institute of Aeronautics and Astronautics American Meteorological Society, Washington, DC Chapter American Nuclear Society, Washington DC Section American Phytopathological Society, Potomac Division American Society for Cybernetics American Society for Metals, Washington Chapter American Society of Civil Engineers, National Capital Section American Society of Mechanical Engineers, Washington Section American Society of Microbiology, Washington Branch American Society of Plant Biologists, Mid-Atlantic Anthropological Society of Washington ASM International Association for Women in Science, DC Metropolitan Chapter Association for Computing Machinery, DC Area Chapter Association for Science, Technology, and Innovation Association of Information Technology Professionals Biological Society of Washington Botanical Society of Washington Capital Area Food Protection Association Chemical Society of Washington District of Columbia Institute of Chemists District of Columbia Psychological Association Eastern Sociological Society Electrochemical Society, National Capital Section Entomological Society of Washington Geological Society of Washington Flistorical Society of Washington DC Human Factors and Ergonomics Society, Potomac Chapter Institute of Electrical and Electronics Engineers, Northern Virginia Section (continued on next page) Paul Arveson Frank R. Haig, S. J. Sethanne Howard Vacant Lee Benaka David W. Brandt Vacant Charles Martin Vacant Stuart Umpleby Vacant Vacant Daniel J. Vavrick Vacant Mark Holland Vacant Toni Marechaux Jodi Wesemann Alan Ford F. Douglas Witherspoon Chuck Lowe Stephen Gardiner Chris Puttock Keith Lempel Elise Ann Brown Vacant Tony Jimenez Ronald W. Mandersheid Vacant Vacant Jeffrey B. Plescia Jurate Landwehr Vacant Gerald P. Krueger Murty Polavarapu Washington Academy of Sciences Delegates to the Washington Academy of Sciences Representing Affiliated Scientific Societies (continued from previous page) Institute of Electrical and Electronics Engineers, Washington Section Institute of Food Technologies, Washington DC Section Institute of Industrial Engineers, National Capital Chapter International Association for Dental Research, American Section International Society for the Systems Sciences International Society of Automation, Baltimore Washington Section Instrument Society of America Marine Technology Society Maryland Native Plant Society Mathematical Association of America, Maryland-District of Columbia- Virginia Section Medical Society of the District of Columbia National Capital Area Skeptics National Capital Astronomers National Geographic Society Optical Society of America, National Capital Section Pest Science Society of America Philosophical Society of Washington Society for Experimental Biology and Medicine Society of American Foresters, National Capital Society Society of American Military Engineers, Washington DC Post Society of Manufacturing Engineers, Washington DC Chapter Society of Mining, Metallurgy, and Exploration, Inc., Washington DC Section Soil and Water Conservation Society, National Capital Chapter Technology Transfer Society, Washington Area Chapter Virginia Native Plant Society, Potowmack Chapter Washington DC Chapter of the Institute for Operations Research and the Management Sciences (WINFORMS) Washington Evolutionary Systems Society Washington Elistory of Science Club Washington Paint Technology Group Washington Society of Engineers Washington Society for the History of Medicine Washington Statistical Society World Future Society, National Capital Region Chapter Richard Hill Vacant Neal F. Schmeidler J. Terrell Hoffeld Vacant Vacant Hank Hegner Jake Sobin Vacant D. S. Joseph Vacant Vacant Jay H. Miller Vacant James Cole Vacant Eugenie Mielczarek Vacant Daina Apple Vacant Vacant E. Lee Bray Terrell Erickson Richard Leshuk Vacant Russell Wooten Vacant Albert G. Gluckman Vacant Alvin Reiner Alain Touwaide Michael P. Cohen Jim Honig Washington Academy of Sciences Room 113 1200 New York Ave. NW Washington, DC 20005 Return Postage Guaranteed NONPROFIT ORG US POSTAGE PAID MERRIFIELD VA 22081 PERMIT# 888 tuuu I,lllllllll,llll,,llllllllll1l1l1,1l,ll1lllllllllll„l1111ll,ll,l ZZZ n* 3.7.™^ *«AUT0**M|xed ADC 207 SERIAL RECORDS DIVISION ' ERNST MAYR LIBFRAY MUSEUM COMP ZOOLOGY HARVARD UNIVERSITY 26 OXFORD ST mlitt CAMBRIDGE MA 02138-2902 — ™ * CMA'/ft/- : •' ! C? O 1 \ / J