A METHOD OF ASSESSING

JUNGIAN PSYCHOLOGICAL TYPE DEVELOPMENT

IN A HIGH SCHOOL STUDENT SAMPLE

BY
JANIE DARLENE SWEET

A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL

OF THE UNIVERSITY OF FLORIDA IN

PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA
1981

Copyright 1981

by

Janie Darlene Sweet

ACKNOWLEDGEMENTS

Words cannot express adequately my appreciation and
heart- felt thanks to my committee, Richard Anderson, Wilson
Guertin, and especially to my chairman, Donald Avila, for
putting up with me all these years, for too many phone calls
during meetings, and for my showing up on trips from South
Carolina to Florida with very little notice. Don has served
me well as an advisor and was always there when I needed
someone to listen and be a friend.

To Drs . Margaret K. Morgan and Mary H. McCaulley, my
love and appreciation for long years of support, encourage-
ment, and guidance. My feelings for them are equally shared
by my husband, George.

Thanks to the staff of the Center for Applications of
Psychological Type for their assistance and patience in
helping me get ready to conduct the testing phase of the
study.

To Richard Kainz , doctoral student in clinical psy-
chology and long-time member of the CAPT staff, I am
indebted in a multitude of ways-- for his part in data
collection, for fearlessly helping me tackle the computer,
for playing devil's advocate, and for being a true friend.

To Claudette Connolly, my dearest and closest friend,
there is not room to express my appreciation for all the

years of encouragement and support. My only regret that
this dissertation is finished is in no longer having a valid
excuse to visit you often.

To Dr. Drew Barrett, Assistant Principal at Sarasota
High School at the time this study was conducted, my deepest
and warmest thanks to you and to the kids, for making the
study possible and for all your valuable efforts and time in
providing the needed information.

My thanks to my family, George, Scott, and George, Jr. ,
for hanging in there during the disappointments and the
sacrifices, and for sharing in the joys. Anything that I
have accomplished would not have been possible without their
patience, understanding, and love.

Thanks to my parents, my aunt, and my grandmother for
their love, support, and all the babysitting so I could work
knowing my children were in good hands.

To the spirit of Isabel Briggs Myers--I regret I could
not finish this in time for her reading- -her work and her
being have inspired so many. This is dedicated in memorium
to her.

CABLE OF CONTENTS

PAGE

ACKNOWLEDGEMENTS iii

LIST OF TABLES vii

LIST OF FIGURES ix

ABSTRACT x

CHAPTER

I INTRODUCTION 1

Purpose of the Study 1

Statement of the Problem 1

Theoretical Basis for the Study 4

Focus of the Present Study 7

Hypotheses 9

II RELEVANT LITERATURE 10

Jung's Theory of Psychological Types 10

The Developmental Nature of the Theory. . 10
Extraversion, Introversion and the

Four Psychological Functions 14

The Origin of Psychological Type

Preferences 17

An Interpretation of the Dominant and

the Auxiliary Function 25

Evidence Related to the Value of Clarity
and Strength of Psychological Type

Preferences 30

An Assessment Problem 44

III DESIGN OF THE STUDY 47

Subjects 47

Procedure 49

Instrumentation Employed in the Study 51

The Myers -Briggs Type Indicator 51

Description 51

Validity 52

Reliability 55

CHAPTER

TABLE OF CONTENTS
(Continued)

PAGE

The Moral Judgment Questionnaire 58

Description 58

Validity 59

Reliability 62

Demographic and School Performance

Data 63

Demographic 63

School Performance 63

The Discrepancy Index 64

Statement of Specific Hypotheses 68

Data Analysis 70

IV RESULTS 73

One-way Analysis of Variance 73

Post Hoc Comparisons 75

Selection Ratio Type Table Analyses 75

Frequencies 76

Results of Hypotheses Testing 76

V DISCUSSION AND CONCLUSIONS 85

Discussion of Results 85

Summary 93

Implications for Education 94

Limitations and Suggestions for

Further Research 98

APPENDICES

A SUPPLEMENTARY ANALYSES OF TEACHER ESTIMATES

OF READING LEVEL 103

B DESCRIPTIVE STATISTICS FOR DISCREPANCY

INDICES 105

C TYPE DISTRIBUTION IN STUDY SAMPLE Ill

D RESULTS AND DISCUSSION OF SELECTION RATIO

TYPE TABLE (SRTT) ANALYSES 117

BIBLIOGRAPHY 124

BIOGRAPHICAL SKETCH 131

LIST OF TABLES
TABLE PAGE

1 Subjects broken down by grade level 48

2 Confidence ranges for reported preferences 65

3 Continuous score ranges differing significantly

from chance 66

4 Significant findings for H 77

2 1

5 Significant findings for H 79

6 Significant findings for H2-2 and H2 ' 4 81

7 Significant findings for H 84

APPENDICES

Al Significant findings for teacher estimates of

reading level 104

Bl Descriptive statistics for DI's reflecting

strength of preference 106

B2 Descriptive statistics for DI's reflecting

word-pair/phrase discrepancies 107

B3 Descriptive statistics for DI's reflecting

split-half differences 108

B4 Descriptive statistics for total discrepancies

for each MBTI scale 109

B5 Descriptive statistics for combined

discrepancies on all 4 MBTI scales 110

CI Psychological type distribution for 485 high

school students 114

C2 Myers' students in Pennsylvania high schools 115

C3 Comparison of 485 high school students with

Myers' Pennsylvania high school student sample... 116

vii

LIST OF TABLES
(Continued)

TABLE PAGE

Dl SRTT analysis for D1EI groups 1 and 3 121

D2 SRTT analysis for D1TF groups 2 and 4 122

D3 SRTT analysis for D4SN groups 1 and 3 123

LIST OF FIGURES
FIGURE PAGE

1 Diagram illustrating the basic psychological

type preferences 29

2 Myers' schema for determining dominant

function 30

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

A METHOD OF ASSESSING

JUNGIAN PSYCHOLOGICAL TYPE DEVELOPMENT

IN A HIGH SCHOOL STUDENT SAMPLE

By

JANIE DARLENE SWEET

August 1981

Chairman: Donald Avila

Major Department: Foundations of Education

A Discrepancy Index designed to indicate the level of
psychological type development at which a person is func-
tioning was constructed and evaluated. The index was
derived from the following three categories of discrepant or
inconsistent responses to the Myers-Briggs Type Indicator
(MBTI) : (1) when a person has a very low overall score in
general for a particular scale of the MBTI; (2) a person
gives answers favoring one pole of a scale on the word-pair
items and then chooses the opposite pole on the phrase
questions, or vice versa; and/or (3) a person gives answers
favoring one pole or an MBTI scale on the x or y-half of the
MBTI and then favors the opposite pole on the other half.

The focus of the study was to determine, in a large
sample of metropolitan Florida high school students
(N = 485) , whether the Discrepancy Index has the capability
to differentiate between students high and low in level of
moral reasoning and high and low on academic variables
(overall and academic grade point average and class rank) .

It appears that the Discrepancy Index dealing with
strength of preference has discriminating power. As pre-
dicted, students expressing high preference for their per-
ceptive function, whether it is for Sensing or for
Intuition, fared better academically than students with low
preferences for their perceptive function. A similar rela-
tionship existed for high Moral Judgment Scale score and
high preference for their judging function, either Thinking
or Feeling.

The Discrepancy Indices for word-pair/phrase discrepancy
and x-y half differences need further adjustment and
re- evaluation. As constructed for the present study they
predicted very little. More sensitive dependent measures
than grade point average should be utilized.

Results of the study do provide evidence that assessment
of "within type" differences is reasonable and possible, and
that strong, clear preference appears to be an important and
measurable component of good type development. The results
obtained provide further support for the theoretical
assumption that it is the clarity and degree of development
of a preference, and not which preference is favored, that
accounts for competence in perception and judgment.

xi

CHAPTER I
INTRODUCTION

Purpose of the Study

The purpose of this study was to evaluate the validity
of a Discrepancy Index designed to indicate the level of
psychological type development at which a person is func-
tioning. The Discrepancy Index was constructed from three
categories of discrepant or inconsistent responses to the
Myers-Briggs Type Indicator (MBTI) .

Statement of the Problem

Some of the problems facing the public school system
have reached or are reaching a crisis point. The push for
accountability in schools is gaining momentum. When large
proportions of students cannot pass functional literacy
tests it is no surprise that the schools bear the burden of
blame.

Part of the total problem is that there is no con-
sensus among members of the American public as to what the
schools should, in fact, be teaching. Once goals are
clearly delineated by school administrators and curriculum
planners, the important question then becomes how to teach
that which is supposed to be taught (a matter of methods
rather than of goals) . One method which has surfaced is the
attempt to individualize the educational process. A problem

2
here is to find ways to individualize instruction while
achieving cost efficiency. The successful individualized
approach will be one which provides a straightforward and
inexpensive way to discover and deal with individual differ-
ences of students in their own approaches to learning situa-
tions. A great deal of research on individualizing
instruction is taking place all over the country. It is
this researcher's belief that if we cannot very soon, and on
a large scale, satisfactorily implement in the classroom the
fruits of these researches, the movement to individualize
will go the route of so many other short-lived educational
innovations .

Carl G. Jung (1921) proposed a theory of individual
differences which is currently being examined and evaluated
in educational settings as a means of individualizing
instruction. Research has clearly established that people
do differ in the basic preferences for the use of the psy-
chological functions and attitudes proposed by Jung, and
that these preferences make a difference in motivation,
aptitude and achievement. In addition, students differ in
the command that they have over their preferred functions.
Teachers and counselors dealing with psychological type
concepts in the classroom have made similar observations.
In describing the psychological type preferences Jung (1921) ,
Myers (1962) and others (McCaulley, 1978; Quenk, 1978; Van
der Hoop, 1939; von Franz & Hillman, 1971) comment upon dif-
ferences between persons with excellent type development and those

3
with problems in type development. In other words, it is
expected that some individuals will have developed their
preferred process or function to a greater extent than
others preferring use of the same function. The degree to
which students have developed their preferred functions has
implications for how well they perform in day-to-day situa-
tions which call upon their powers of perception and their
powers of judgment. In this respect, the present study is
at a far more individual and micro- level than that of
dealing with basic type differences. The focus is on
individual differences within types as much as with the
differences between types which have been the concern of
earlier research.

Knowledge of psychological type preferences of students
is beneficial to educators and to students. In elementary
level classrooms, however, it may not be necessary or even
advisable to determine the type preferences of particular
students. At this age level students are gaining experience
and expertise in their own approaches to learning situations.
Many elementary level teachers have found it useful, however,
to estimate type through observation (Nuernberger & Lawrence,
1974) . Knowledge of activities that appeal to students with
different psychological type preferences allows the instruc-
tor to offer a "smorgaasbord"--where, indeed, there is
something for everybody and students can gain experiences
which develop all their powers of perception and judgment.

4
By the time a student reaches the high school level, an
indication of the type preferences of the individual becomes
relevant for more than understanding communication between
teachers and students. It becomes relevant for decisions
about lifelong goals, including career decisions. It is
particularly important for students experiencing academic
difficulty, and for their teachers, to assess not only their
type preferences but to determine whether or not they really
have command over the psychological functions they prefer to
use. It is for these students that the educational dollar
to individualize instruction can have the highest return.
Public Law 94-142 has mandated the development of highly
individualized educational programs for all handicapped
learners. Learning disabled students are a target group for
highly individualized instruction. Use of the MBIT to
assess type preferences of learning disabled students is
being explored (Metts, 1979).

At this point in time we can assess the basic type
preferences with a considerable degree of reliability. The
pressing problem is to discover a reliable means to appraise
"the potent but as yet unmeasurable variable of 'type
development ' --i. e. , the extent to which the person actually
has developed the processes and the attitudes which he pre-
fers" (Myers, 1962, p. 19).

Theoretical Basis for the Study
The approach to individualizing instruction being
utilized in the present study is based on Carl G. Jung's

5
theory of psychological types (1921) . The theory is highly
relevant to education because it focuses on the individual's
conscious use of his/her powers of perception and his/her
powers of judgment. The theory begins at the point where
the students focus their attention to take in information or
know something (perceive) and then to decide or come to con-
clusions (judge) about it. Some students will naturally be
better at perceiving than judging and vice versa, but all
individuals need to develop their full potential for both
processes .

An instrument, the Myers-Briggs Type Indicator (MBTI) ,
designed to implement Jung's theory of psychological types,
was developed by Isabel Myers in the 1940' s. After careful
study and evaluation, the MBTI was published as a research
instrument by Educational Testing Service in 1962 and as a
tool for general professional use by Consulting Psycholo-
gists Press in 1975. A substantial body of research using
this instrument in education is available (Bibliography,
Center for Applications of Psychological Type, 1981).

Psychological type differences have been fruitfully
studied. Attempts to individualize instruction based on
these findings are meeting with success. Gifted students
and others with special talents have been studied from the
perspective of psychological type differences. Increased
knowledge of student preferences for educational media based
on type differences is emerging (Anast, 1966; Golanty-Koel ,
1977; Williams, 1973). A substantial proportion of the

literature on psychological type differences in students
deals with academic aptitude and performance. With so much
current interest in cognitive structures, perceptual organi-
zation, and teaching and learning "styles" it is not sur-
prising to find a large proportion of studies in those areas
taking psychological type variables into account in settings
that range from middle school through graduate and profes-
sional training. Research over more than three decades has
established the reliability of the MBIT (Carlyn, 1977; Coan,
1978; Myers, 1962) and the practical utility of Jung's
theory in the educational process (Neurnberger & Lawrence,
1974; McCaulley, 1974, 1976a, 1976b; von Franz & Hillman,
1971; Wickes, 1966) .

Myers (1980) considers a major goal of education to be
the fostering of psychological type development in order to
improve and enhance learning and to prevent or head off
future problems. In this writer's opinion we have reached a
major roadblock in accomplishing this goal, and an obstacle
for psychological type research in the educational process
in general. We have looked extensively at basic type
differences between students in all sorts of fields,
programs and settings. What we do not yet know is the
amount of variance in our studies that might potentially be
accounted for if we could identify persons who are at a high
level of type development (regardless of which preferences
they like and use best) from those who are less differenti-
ated and at low 1 eve 1 s of type development.

7
Before we can study this issue further, we must have
a way to measure type development. Research related to
education employing the MBTI has not yet dealt centrally or
systematically with the issue of individual differences in
type development. There are several clues that such studies
will be productive. Myers (1962) found underachieving stu-
dents in her sample to have lower internal consistency
measures on the MBTI than high achievers. She also found
aptitude and achievement scores for high school and college
students in her sample to be lower for those potentially
less clear about their type preferences.

Focus of the Present Study
In day-to-day situations, inconsistent or discrepant
responses to an MBTI scale are being interpreted as
suggestive of inadequate development of the psychological
process that the scale measures. Before any lack of
development can be assumed to exist, more direct evidence
regarding the meaning of inconsistent response patterns must
be established.

It was the focus of the present study to construct and
evaluate one possible index of level of psychological type
development. The focus was three- fold. First, a Dis-
crepancy Index was created which operationalizes the con-
cepts of inconsistent response and strength of type
preference. The inconsistencies or discrepancies upon which
the Discrepancy Index was built are: (1) when a person
responds or gives answers favoring one pole of an MBTI scale

on the X or Y-half of the MBTI and then favors the opposite
pole on the other half; (2) a person gives answers favoring
one pole of a scale on the word-pair items of the MBTI and
then chooses the opposite pole on the phrase-question items,
or vice versa; and/or (3) a person has a very low overall
score in general for a particular scale of the MBTI.

Secondly, it was determined whether, in a large sample
of high school students (N = 485) , a relationship exists
between discrepant responding on the MBTI as measured by the
Discrepancy Index and students' responses to a Moral Judg-
ment Questionnaire. The Moral Judgment Scale was developed
by Maitland and Goldman (1974) to assess the level of develop-
ment of moral reasoning (judgment) that a person employs.

Finally, the academic and overall gradepoint average
and class rank of students with low Discrepancy Index scores
were compared with those of students with high Discrepancy
Index scores to determine whether a relationship exists
between discrepant response pattern and measures of academic
performance .

It was also expected that students with a high Dis-
crepancy Index score on the Sensing-Intuition (S-N) scale of
the MBTI would have lower reading levels than students with
low Discrepancy Index scores on the S-N scale. Reading test
scores could be made available from school records for so
few students that this relationship was not tested formally
in the present study. English teachers did provide the
investigator with estimates of students' reading level.

9
Supplemental analyses of these estimates of reading level
with Discrepancy Index scores were performed and are
reported in Appendix A.

Hypotheses
H Students who respond in a consistent and non- discrepant
pattern on the MBIT will employ a higher level of moral
reasoning than students who respond in an inconsistent and

discrepant pattern on the MBTI.

2
H Students who respond in a consistent and non-discrepant

pattern on the MBTI will have demonstrated higher academic

performance than students who respond in an inconsistent

and discrepant pattern on the MBTI.

3
H Students with low Discrepancy Index scores on the

Thinking-Feeling scale of the MBTI when it represents their

theorized Dominant Function will employ a higher level of

moral reasoning than students with high Discrepancy Index

scores on the T-F scale when it represents their theorized

Dominant Function.

4
H Students with low Discrepancy Index scores on the

Sensing- Intuition scale of the MBTI when it represents their

theorized Dominant Function will have attained a higher

academic and overall grade point average, or higher academic

and overall class rank than students with high Discrepancy

Index scores on the S-N scale when it represents their

theorized Dominant Function.

CHAPTER II
RELEVANT LITERATURE

In the first portion of this chapter the basic concepts

of Jung's theory of psychological types are reviewed and the

developmental aspects of the theory explicated. In part two

research and clinical observations related to the value of

an individual's holding clear and consistent psychological

type preferences are examined. Part three delineates an

assessment problem in psychological type research.

Jung's Theory of Psychological Types

The Developmental Nature
of the Theory

Increasing recognition of individual differences and
the need to "individualize" the educational experience have
led many educators to the study of psychological type
differences. Many of these educators are attracted by the
underlying assumption that the differences found between
types represent alternative, but equally valid paths to the
achievement of excellence.

Jung (1921) postulated individual differences in a
person's preferences for mode of perception. Within the
realm of perception, he proposed that some are sensors , pre-
ferring to glean knowledge as to what goes on around them
from the physical senses, i.e., sight, hearing, touch, smell,
and taste. These sensors strongly trust what their

10

LI
experiences have told them. Jung stated that others are
intuitives , preferring to perceive the world in terms of
relationships, meanings and abstractions. These intuitors
are often more concerned with what something means or might
be than with what it is.

Jung proposed that individuals also differ in the type
of j udgment they rely upon. Some are thinkers , preferring
to draw on their powers of logic and objectivity in making
most decisions. Thinking types like best to utilize an
impersonal approach in decision-making. The opposite type
of person, Jung called feeling types, preferring to weigh
all evidence within their own subjective value system as a
basis for making decisions. Feeling types like best and
trust most decisions where they have employed their own sub-
jective criteria as to what is of highest value to them-
selves and to others in the situation.

The third individual difference in psychological type
is in the basic orientation people take towards their
environment. Extraverts are oriented toward the environ-
ment, incorporating it into their own self -system and
drawing energy from it. Introverts are oriented away from
the surrounding environment. The environment is more alien
to them. It takes energy away from them. In contrast to
the extravert, they draw their energy from their own inner
world of thought and ideas. Jung's analysis of these differ-
ences in perception (Sensing and Intuition) and judgment
(Thinking and Feeling) as they interact with basic

12
orientation towards the environment (Extraversion and
Introversion) have been seen empirically to organize, within
a powerful theoretical framework, many valuable observations
of the educational process.

In psychological type theory each of the four functions
(Sensing, Intuition, Thinking, and Feeling) and orientations
(Extraversion and Introversion, sometimes called attitudes)
described by Jung is essential, and all of them are used by
every individual. The functions and attitudes are not
equally preferred, however, by all individuals. These basic
differences in preferences lead to differing priorities of
development, with the most preferred functions being differ-
entiated earlier than the less preferred ones.

Jung's psychological theory places great emphasis on
consciousness. As precisely stated by Fordham,

In attempting to divide human beings into
recognizable types, Jung is dealing mainly with
the psychology of consciousness; when a person
is described as either extraverted or intro-
verted, it means that his habitual conscious
attitude is either the one or the other. A
balanced attitude would include equally both
extraversion and introversion, but it frequently
happens that one attitude is developed and the
other remains unconscious. No one, however,
lives completely as one or the other, but mani-
fests the unconscious attitude at times, though
in an inferior way. (1966, p. 31)

Thus, Jung's theoretical position was that the process of

becoming a mature person, termed individuation, is one of

life- long development taking different forms that are

related to the person's preferences in the use of perception

and judgment. Jung considered each psychological type to be

13

a "normal" process of maturation. Quenk (1978) very clearly

emphasized this position,

Thus Jung was quite explicit in stating that the
attitude and function types are not "pure" but
orienting structures which reflect the habitual
way in which a person behaves. In energic,
libido terms, the attitude and function type are
the vehicles for the flow and amplitude of
psychic energy. As a psychology of character,
then, the Typology is not a static, classifi-
catory scheme, but a dynamic process through
which adaptation to the world occurs. (p. 5)

Central to Jung's concept of maturation is "the process
of becoming whole" (Fordham, 1966, p. 140). The key to this
individuation process is the concept of making conscious the
unconscious. Here Jung meant raising a person's level of
awareness as to the existence of the differing approaches,
which potentially increases the person's behavior repertoire
and self -understanding. Most people use or rely heavily on
one function; some use two. A highly differentiated or
developed person uses three, and in rare cases, people exist
who use all four functions (Sensing, Intuition, Thinking,
and Feeling) . The use of all four functions is the end-
point or goal of the individuation process. This bears
similarity to Maslow's construct called self-actualization.
According to Fordham' s interpretation (1966), the utiliza-
tion of all four functions involves both "the individuation
process, and the reconciliation of the opposing trends of
one's nature" (p. 46). Thus, as the separate components of
psychological type are explained in following sections it is
important to bear in mind the dynamic, developmental under-
pinnings of Jung's theory.

14

Extraversion, Introversion, and the
Four Psychological Functions

Jung considered the psychological type preferences to
be dichotomous in nature--as "either-or" preferences. One
either prefers the Introverted (I) or Extraverted (E) atti-
tude, either Sensing (S) or Intuitive (N) perception, and
either Thinking (T) or Feeling (F) judgments. There is a
very logical ordering of the preferences from E-I to S-N
to T-F. We direct our energy (E or I) to perceive a situa-
tion (S or N) , and then to decide about it (T or F) . Jung
called Sensing and Intuition the "irrational" functions
because they are processes or modes of perception. They
are not decision-making functions. They give us first the
percepts or information that later we may or may not use in
making decisions or judgments'. Perception is the process of
cognition. We have utilized it when we say "I know" or "I
see." Conversely, Jung called Thinking and Feeling the
"rational" functions because they form the foundation of the
decision-making process . They are two different avenues or
approaches to coming to conclusions or judging about some-
thing.

It is necessary to acknowledge again that while Jung

postulated the preference for E or I , S or N, and T or F to

be "either-or" in nature, he did not mean that we utilize

one preference to the total exclusion of the other. Fordham

(1966) re-emphasizes this point,

Since human nature is by no means simple, one
rarely finds the absolutely pure type; often the
main function is sufficiently clear to dub the

15

person a thinker, an intuitive and so on, but it
is seconded by another function which modifies
and blurs the picture. Jung in fact refers to
his description of types as "somewhat Galtonesque
family portraits," for human nature refuses to
be classified in a precise and simple way. All
the same, the concept of types has great practi-
cal value as an aid to understanding in personal
relationships and in education. (p. 45)

In more specific terms, extraversion or introversion
indicates the direction of the individual's flow of energy.
A person who prefers the extraverted orientation is most "at
home" when energy and attention are focused on the outer
world of people and objects. Persons preferring intro-
version are more comfortable in the inner world of their
own thoughts and ideas. Kagan's "impulsive" and "reflec-
tive" dimensions reflect the Extravert- Introvert difference
in orientation. Jung (1921) states,

The introvert's attitude is an abstracting one; at
bottom, he is always intent on withdrawing libido
from the object, as though he had to prevent the
object from gaining power over him. The extravert,
on the contrary, has a positive relation to the
object. He affirms its importance to such an
extent that his subjective attitude is constantly
related to and oriented by the object. (p. 329)

Next comes the perceptive function which is divided

into the Sensing preference and the Intuitive preference

(Sensing and Intuition are two of the four functions Jung

refers to) . The person who prefers sensing perception likes

best to take in information (i.e., perceive) through direct

use of and reliance upon the physical senses. This type of

perception is that which is most trusted by the individual.

Alternately, the person who prefers intuitive perception

16
likes best to "see" or perceive the possibilities extant in
a situation rather than the concrete realities of it. For
example, Intuitive types are often described as people with
great powers of imagination. Likewise, Sensing types are
often described as being superbly practical. This is no
surprise, as the intuitive person relishes in seeing "what
might be" and the sensing person delights in clear percep-
tion of "what is ."

Finally, having perceived a situation, we come to con-
clusions or decide about it. For this Jung said we utilize
the judging function. The judging function is characterized
by the preferences for Feeling judgments or for Thinking
judgments (Feeling and Thinking are the other two of the
four functions referred to by Jung) . While every individual
has the potential for both kinds of judgment, each person
tends to prefer or favor one kind over the other. This
leads to greater and greater development of the favored way.
Persons preferring Feeling judgment like best to rely on
their own value system to guide their behavior. The con-
scious and subjective weighing of things important to them
and to others involved in a situation is a cornerstone of
Feeling judgment. This process sheds light on the solution
which can be most trusted and satisfying. Alternately,
persons who prefer Thinking judgment trust and like best to
decide based on what they believe to be the logic of the
situation. Much emphasis is placed on the logical
consequences of one decision as opposed to another (cause

17
and effect). The attempt at impersonal objectivity is a
cornerstone of Thinking judgment. William James was
probably noting type differences in the Thinking and Feeling
functions when he postulated his tough-minded and tender-
minded types .

The Origin of Psychological
Type Preferences

The origin of the attitude and function types is not
certain. It remains the object of scientific inquiry. Jung
(1921) stated that at first we might be erroneously
inclined to consider such differences "as mere idiosyn-
crasies of character peculiar to individuals"; however, the
differences in behavior are consistent enough to be highly
predictable. He proceeded to state that these consistent
differences are found across all ranks of society--

Sex makes no difference either; one finds the
same contrast among women of all classes. Such
a widespread distribution could hardly have come
about if it were merely a question of conscious
and deliberate choice of attitude. In that case,
one would surely find one particular attitude in
one particular class of people linked together by
a common education and background and localized
accordingly. But that is not so at all; on the
contrary, the types seem to be distributed quite
at random. (pp. 330-333)

The type preferences show themselves in behavior very

early. Von Franz and Hillman (1971) say in the crib, and Jun^

says on the first day of life. Jung seems to have ruled out

Jung's highly detailed descriptions of the attitudes
(E or I) and the four functions appear throughout
Psychological Types (1921) , and in more condensed form
in Chapter Eight of The Portable Jung, J. Campbell, 1971

18
any strict Mendelian inheritance pattern from parents as the
origin of type preferences. In other words, he could
observe no pattern wherein two extraverted parents produce
an extraverted child, or two sensing parents a sensing
child. Indeed, present-day data reveal no inheritance pat-
terns either. Jung hints, however, that biological and
physiological factors probably play a primary role--

In the same family one child is introverted, the
other extraverted. Since the facts show that
the attitude- type is a general phenomenon having
an apparently random distribution, it cannot be
a matter of conscious judgment or conscious
intention, but must be due to some unconscious,
instinctive cause. As a general psychological
phenomenon, therefore, the type antithesis must
have some kind of biological foundation . . .
there are obviously individuals who have a
greater capacity, or to whom it is more con-
genial, to adapt in one way and not in another.
It may well be that physiological causes of which
we have no knowledge play a part in this. I do
not think it improbable, in view of one's
experience that a reversal of type often proves
exceedingly harmful to the physiological well-
being of the organism, usually causing acute
exhaustion. (pp. 330-333)

One avenue of inquiry which may eventually shed light
on the origin of type preferences lies in investigation of
the functioning of the right and left hemispheres of the
brain. There may, for example, be a relationship between
hemispheric dominance and preferred mode of perception
(Sensing or Intuition) . It would seem possible that
preference for sensing could be correlated with left-
hemispheric dominance and preference for intuition with
right-hemispheric dominance. Whatever the explanation,
Jung clearly suspected a biological basis for the type

19
preferences; an inborn predisposition to develop in one
direction rather than another.

Both biology and environment play a role in the develop-
ment of most human characteristics. Jung (1921) discussed
how environmental influences (especially parents) can
affect drastically what would be the normal type development
of the child. But, he says these are extreme and abnormal
cases and that

As a rule, whenever such a falsification of type
takes place as a result of parental influence,
the individual becomes neurotic later, and can be
cured only by developing the attitude consonant
with his nature. (p. 332)

Jung thus thought environment very important , with

possibilities existing both for falsification of type and

for helping develop a basic predisposition. In the Inner

World of Childhood, Wickes (1966) provides a further

example of the forces affecting the personality type of a

very young child,

The closer the bond between the parent and child
the more the child is molded not only by the
conscious but also the unconscious demands of the
parent, reacting in many ways as he is expected
to react instead of in accordance with his
individual type. Such a confusion produces a
sense of unreality and makes the integration of
the personality a difficult task. (p. 130)

Several writers have commented upon the effects the

child's personality type has in the classroom, specifically

in the relationship with the teacher. Wickes discussed

the many ways in which the child is affected by school and

ways that teachers or parents can ascertain (without

20
pigeonholing) clues as to psychological type of the child,
and how to utilize that information to the best benefit of
the child. Wickes (1966) offered the following anecdotal
illustration of why clues as to the psychological type
preferences of the child are important "tools" in the edu-
cational process,

I have known schools where it was possible to
predict the success or failure of a child of
pronounced type as he advanced from grade to
grade and teacher to teacher. In one class he
would be spoken of as a boy with an interesting
mind and showing promise of great ability; in
the next as impractical, dreamy, inaccurate.
Or the reverse type of boy would be first depre-
ciated as too ready, lacking in imagination and
originality, and in the next grade praised as a
fine, accurate, responsive, careful pupil. These
verdicts were primarily based upon the type atti-
tude of the teacher and upon the form of work
which she valued. The same is true of parents.
They feel rapport with the child whom they can
understand. (pp. 134-135)

Fordham (1966) states that it is of help to teachers to
"realize that an introverted child, for instance, is not
unhappy or unadapted if it does not join in activities with
the same zest as extraverted pupils" (p. 45). She gives
the following comparison of an extraverted and an intro-
verted child:

This [the extraverted child] is the type of child
who is popular both with parents and teachers.
He is spoken of as "well adjusted", and is often
considered "brighter" than he really is because
of his earlier development and his capacity to
make a good impression.

The introverted child is shy and hesitant.
He dislikes all new situations, and even
approaches new objects with caution, and some-
times with fear. He prefers to play alone, and
have one, rather than many friends. Because of
the widespread preference for extraversion, such

21

introverted children often cause anxiety to
parents, but they are just as "normal" and intel-
ligent as the other type of child. They are
thoughtful and reflective, and often have a rich
imaginative life. What they need most is time
to develop their less obvious gifts, and to learn
to feel at home in the world. (p. 32)

These external forces (parental and societal, as well
as all life experiences) can be positive or negative in
their influence on type development. In the lifetime of an
individual, the formal education process has the potential
for extensive positive influence. With respect to the
institution of education, Inlow (1970) states that the goals
of formal education and of humanism go hand in hand, "First
and foremost, individual man is of supreme worth. Further-
more, he is an emerging essence, with total personality ful-
fillment constituting both his birthright and his
commitment" (p. 29). Inlow further comments that learners
themselves are an important source of any curriculum, "No
curriculum, irrespective of its antecedents, is functional
unless it relates meaningfully to the differing interests,
abilities, and needs of learners" (pp. 117-118). Myers
(1962) has long favored the smorgaasbord approach, where
students of all types find their psychological differences
both respected and nurtured, and where there is an
atmosphere that fosters fuller personality development.
Such an approach serves a dual purpose; every child is
provided opportunities to develop all functions, as well as
develop his own preferred functions to a higher level of
expertise .

22
Lawrence (1979), in his practical guide to learning
styles, provides examples using psychological type concepts
in developing school curricula, and specifically for planning
instructional strategies for the classroom which "honor the
rights of every type" (p. 27). He clarifies how type con-
cepts suggest solutions for two of education's most persis-
tent and perplexing problems. School personnel across the
country report lack of student motivation a major problem.
Lawrence (1979) suggests breaking motivation down into four
parts corresponding to the four dimensions of psychological
type as follows:

1. The extraversion- introversion preference
shows the broad areas of a student's
natural interest

2. The sensing-intuition preference reveals
basic learning style differences

3. The thinking- feeling dimension shows patterns
of commitments and values of a student

4. The judging-perceiving dimension shows
work habits. (p. 24)

When these four natural motivators are taken into account,

Lawrence suggests that teachers "can better direct student

energies toward learning" (p. 24).

Lawrence describes another problem which is increasingly

becoming known to educators and educational researchers

dealing with psychological type data. He refers to it as

"perhaps the most crucial unrecognized problem of American

education" (p. 27). Lawrence is pointing to the biases of

instruction that may actually operate to the detriment of

students who prefer extraverted sensing (ES types) . It

is estimated (McCaulley, 1978; Myers, 1962) that these

23
extraverted sensing types make up about 70% of the general
population, and hence approximately 707o of the average
public school population. Two examples of these biases may
serve to illustrate the problem. The classroom practice of
presenting abstractions first and applications second
appeals to and fits best the intuitive' s learning style, not
that of the sensing type. Secondly, biases favoring both
introversion and intuition may exist in reading instruction.
Lawrence explains,

Probably the young child first encounters it in
the teaching of reading. Of course, reading is
primarily an introverted activity; it is done
quietly by oneself. And it is fundamentally an
intuitive activity, involving abstractions — the
printed symbols. But reading need not be taught
as if all students were introverts and intuitives.
The skills of reading can be and are being
mastered by students of all types, and they all
can leave school with a positive attitude toward
reading. Tragically many students are alienated
by their first encounters with reading instruc-
tion. Type theory points to the probability that
most of the alienated children are extraverted
sensing types. There are data which indicate that
the dropout rate is much higher among sensing
children. (pp. 26-27)

In addition, Lawrence and others have noted that

standardized test constructors seem to be biased toward

intuitive intelligence (manipulating symbols, abstract

thinking, the drawing of inferences, describing "how to"

rather than doing). As a group, IN types score highest

on intelligence tests, followed by EN , IS , and finally

ES types. Even textbooks seem to favor the intuitive way

of viewing the world. According to Lawrence, "The record of

American education in the twentieth century is a record of

24
neglect of sensing intelligence . . . the kind of

intelligence possessed by the majority of American students"

(1979, p. 27) .

The problems raised by Lawrence are matters of concern
for educators who wish to take an active role in fostering
the total development of their students. Empirical research
dealing directly with these issues is needed. Fuller under-
standing of psychological type concepts and the process of
type development offers great potential for developing
better instructional strategies for individual learners.

Both internal and external forces have a role in the
normal "unfolding" of the developmental process. It is the
favored mode for perceiving and for judging which is
naturally the one most utilized by the child. Habitual use,
through reinforcement, makes for further development and
expertise in the use of the preferred way of perceiving and
the preferred way of judging. This habitual use of one's
preferences produces sets of characteristics, values and
behaviors which a person of a given type shares in common
with other people of the same psychological type.

Humans, in general, seem to feel rapport with others
who resemble them in the preferences for the psychological
attitudes and functions. Van der Hoop's quote of the earth
spirit's answer to Faust is illustrative, "That spirit thou
resemblest, whom thou dost comprehend" (1939, p. 319).

Von Franz and Hillman (1971) , in their treatise on
psychological type, discussed the natural initial one-
sidedness of the process

25

of development. In describing how our psychological type

preferences affect our interpretation of the world around us

Van der Hoop (1939) pointed out that our type differences

have some far-reaching consequences- -

One's immediate experience of one's fellowman
proves itself at once to be much more influenced
by one's own personality than is one's experience
of a natural object, and the way in which this
experience is dealt with is likewise still more
under this influence, since the different modes
of psychological objectif ication are sometimes
determined by typical attitudes .... In the
various schools of psychology, this influence
can frequently be clearly demonstrated. . . . Men
of different types are not all equally interested
in scientific problems. . . . Although I do not
maintain that this recognition of a one-sidedness
in conscious orientation according to type offers
an explanation of all the varieties in outlook,
this differentiation does, nevertheless, make it
possible to get a better understanding of the one-
sidedness in these points of view. (pp. 319-320)

An Interpretation of the Dominant
and the Auxiliary Function

Having dealt briefly with what the basic type
preferences are, Jung's description of the dynamic inter-
play of the preferences within a single individual can be
reviewed. It is a gross oversimplification and probably an
inaccuracy to deal with the effect of the preferences
separately, without reference to one another, because the
preferences have their joint effect in unison. To clarify
this "orchestration" effect we must understand what Jung
called the Dominant and the Auxiliary function.

The Dominant function for any individual is theorized
to be the most conscious and most developed function of the

26
four functions. Therefore, the Dominant function is either
Sensing perception or Intuitive perception, or either Think-
ing judgment or Feeling judgment. Also from the four
functions, Jung postulated the existence of an auxiliary
function. He stated (1921) that the auxiliary is "comple-
mentary," "relatively unconscious," and "in every respect
different from the nature of the primary [Dominant] func-
tion" (pp. 405-406). We recall that the extraverted or
introverted orientation indicates the direction of the
individual's flow of energy; introversion toward the inner
world of thought and ideas, extraversion toward the outer
world of people, objects and events. Based on Jung's state-
ment about the auxiliary function we can assume two things:

(1) If the Dominant function is a judging function
(thinking or feeling, whichever the person prefers), then
the auxiliary function must be a perceptive function
(sensing or intuition, whichever the person prefers) ; and if
a perceptive function is dominant then the auxiliary
function for that person must lie in a judging function; and

(2) If the Dominant function for a person is extraverted,
then the auxiliary function must be introverted; and if the
Dominant function is introverted, then the auxiliary
function must be extraverted. With regard to this Quenk
(1978) stated, "A function cannot be both extraverted and
introverted. Its directionality must be unitary" (p. 8).
Myers (1962) had earlier come to this same deduction when
she and her mother, Katharine C. Briggs, in the development

27

of the Myers-Briggs Type Indicator (MBTI) added the J - P

Index to indicate whether the person uses a perceptive

function (S or N) or a judging function (T or F) in the

extraverted world (that shown most to others). As shown in

Figure 2, the judging or perceptive attitude identifies

(points to) the Dominant function for Extraverts and the

auxiliary function for Introverts. According to Myers

(1962) , persons who prefer the extraverted orientation

(prefer to direct energy to the outer world of people,

places and objects) extravert their Dominant function.

Introverts (prefer to direct energy to their own world of

thoughts and ideas) introvert their Dominant function. The

extraverts introvert their auxiliary function, and the

introverts extravert their auxiliary function. According to

Quenk (1978),

It also follows that an extravert is more
confident of his dominant function, especially
as he receives confirmation from others. An
introvert, because he extraverts his auxiliary
[or second best] function, receives confirma-
tion from the environment only for his auxiliary;
he must rely on his self-awareness to ascertain
his dominant function. . . . Jung has stated
that extraverts fear their inner world just as
introverts fear the outer world. (p. 10)

In line with this argument Plaut (1972) found that intro-
verts have less confidence than extraverts in determining
their dominant function.

In direct relation to this, Myers (1962) pointed out
that all of us must live in both "inner" and "outer" worlds.
Some prefer the outer world; some prefer the inner world.

28

According to Myers (1962) , the excursion into the least

liked place is delegated to the person's auxiliary function.

With regard to the auxiliary function she stated,

If he [the extravert] has no useful development
of an auxiliary process, he will have little or
no inner life, which will make him an extreme
extravert, and better-balanced associates will
find him superficial. ... If the introvert
has no useful development of an auxiliary process,
his outer life will be a very awkward, accidental
and uncomfortable affair. (pp. 60-61)

On the issue of extraversion, introversion, and the dominant

and the auxiliary function Quenk (1978) further stated,

The well-developed function is that function which
has a definite directionality to its energic
charge. Directionality is an all-or-none phenome-
non. Metaphorically, one can only go in one
direction at a time. Thus, as the functions
become specialized their directionality can be
seen as central to their development. (p. 12)

The "evolution" process of psychological type has been

described by Van der Hoop (1939) as follows:

A first point of difference between people of the
same type is found in the stage of development
which they have reached. In every type there is
a simple form, in which the differentiation of
the prevailing function has only just begun,
and its modes of adaptation are still being
tentatively tried out, although a clear prefer-
ence for typical forms of adaptation can already
be observed. At a later stage the dominating
function has found its forms, controlling these
with great assurance. Anything which is not in
accord is, at this stage, suppressed. With a
few people there follows a still further stage,
in which the other functions are permitted more
development, to compensate for any one-sidedness ,
and the pronounced typical picture is again modi-
fied to some extent by the unfolding of a fuller
and richer expression of human nature. (p. 92)

In a lecture in Zurich, Jung (1928) further explicated the

interplay of the functions in the natural unfolding process

29

of type development. He gave several behavior-based

examples for recognizing functions at differing levels of

development ,

Whether a function is differentiated or not can
be recognized from its strength, stability,
consistency, reliability and adaptedness. But
inferiority in a function is often not so easy
to recognize or to describe. An essential
criterion is its lack of self-sufficiency and
consequent dependence on people and circumstances,
its disposing us to moods and crotchetiness , its
unreliable use, its suggestable and labile
character. The inferior function always puts us
at a disadvantage because we cannot direct it
but rather are its victim.

As a final point of clarification, Figures 1 and 2 may
serve as aides to remembering the proposed nature of psy-
chological type as a developmental schema.

FLOW OF ENERGY PERCEPTION JUDGMENT

Extraverted
orientation
(E)

Introverted
orientation
(I)

Sensing
(S)

Intuition
(N)

■ .a Thinking1

!\ (t)

Feeling
(F)

THE FOUR FUNCTIONS

ATTITUDE TOWARD
ENVIRONMENT

Judging
_v attitude
(J)

Perceptive
.^ attitude
(P)

Figure 1. Diagram illustrating the basic psychological
preferences .

30

For the Extravert

For the Extravert:

For the Introvert:

For the Introvert

<r

■)

With the Judging
attitude, the
Dominant Function
is

With the Percep-
tive attitude, the
Dominant Function
is

With the Judging
attitude, the
Dominant Function
is

With the Percep-
tive attitude, the
Dominant Function
is

Figure 2. Myers' schema for determining dominant function.

Evidence Related to the Value of Clarity
and Strength of Psychological Type Preferences

From the literature it is implied that being clear
about one's psychological type preferences is a positive
characteristic and being uncertain or arbitrary in the use
of the preferences is a negative characteristic. As afore-
mentioned, stable type differences are noticeable in very
young children (Myers, 1962; von Franz, 1971; Wickes, 1966)
From Jung's writings the following sequence of development

31
is inferred-- individuals first develop and gain increasing
expertise in their preferred modes of functioning. Only
then can they begin to work on and consciously develop (at
ease and with comfort and confidence) the functions they do
not prefer. These lesser preferred functions are in theory
those under the least conscious control of the individual.
For most people such development is probably a life-long
journey.

From a clinical perspective, Quenk (1978) offered the
following comments regarding two possible problems in the
process of type development. We are reminded that Jung
considered development to be a process of individuation and
differentiation, with great importance placed on extra-
version and introversion,

If both the auxiliary and dominant functions have
the same directionality [extraverted or intro-
verted], one can be confused for the other; i.e.,
they are undifferentiated. . . . Thus, as the
functions [S-N and T-F] become specialized their
directionality can be seen as central to their
development. (pp. 12-13)

Quenk further stated that if the dominant and auxiliary were

to have the same directionality (both introverted or both

extraverted) then the "energic charge" of both would be

lessened as it is shared by two functions. However, if one

of them is extraverted and one is introverted then "the

energic charge available to each would be greater, thereby

assuring the differentiation of the two functions."

The second problem as viewed by Quenk (1978) was what

he called the "fusion of functions"--

32

Fusion is a blending of functions, e.g., thinking
is fused with feeling, or sensation with intui-
tion. If this occurs, it is a sad state of
affairs. The manifestation of the function would
be erratic and primitive. Fusion of functions is
a sign of poor ego development and predisposes a
person to an onslaught from the unconscious which
may result in psychosis. It is the fusion of
functions which is characteristic of latent psy-
chosis, whereas confusion of functions [with
dominant and auxiliary both introverted or both
extraverted] may result in neurosis. Thus,
determining which of the functions is undiffer-
entiated and the extent to which there is fusion
or confusion of functions is one means by which
ego strength can be ascertained. (pp. 12-13)

Myers (1962) does offer an opinion and some evidence

that clarity and/or strength or preference as assessed by

the MBTI may provide a clue in the eventual understanding of

the complex phenomenon of type development. Preference

scores on the MBTI reflect the strength with which a

preference is reported. Myers points out that mean strength

of a preference varies from preference to preference (scale

to scale) and from one kind of group to another (age

groupings are theoretically relevant here) . It is helpful

to refer any individual's preference score to the group to

which they belong,

A preference score well above the mean is likely
to indicate a strong and conscious preference,
which will demand an outlet in the person's work
and make him dissatisfied with any job that does
not give it exercise and scope. Scores in the
neighborhood of the mean are thought to reflect,
on the average, a moderate to strong preference,
with enough development of its attendant qualities
to affect definitely the person's happiness and
effectiveness in his job. With scores well below
the mean there may or may not be an important
development of the preference. And with really
small scores the preference itself may be in
considerable doubt. (p. 69)

33
In a seventh and eighth grade college preparatory
population and in college populations Myers found the pref-
erence for N, and for I and J in combination, to be signifi-
cantly related to grade point average. The point of
relevance here is that Myers also found strength of
preference to be related to mental age (as well as to
chronological age as implied by van der Hoop) . In her
gifted sample Myers found relatively higher preference
scores for both males and females when compared with others
in their age group. Myers found major differences in the
regression of IQ on mean preference scores for EI, TF, and
JP. For each of these three scales, as strength of
preference increased so did mean IQ scores, creating
U-shaped curves. With respect to S-N, Myers found a higher
preference for Sensing to be associated with lower mean IQ
scores in the high school sample. Myers suggested this
finding may be more indicative of the Intuitive type's
greater interest in scholastic or academic activities in
general. She warns against the interpretation that as
preference for Sensing gets stronger grades get lower.

In a related investigation, Weychert (1975) found the
mean IQ for Intuitive males to be 13 points higher than for
Sensing males and 16 points higher for Intuitive females
than for Sensing females. Weychert points out that this
further substantiates Myers' findings regarding the effect
on intelligence of clarity or strength of preference on the

34

S-N scale because Weychert used only subjects with moderate

to high preferences on S-N (some of the variance in

Weychert 's study was thus automatically removed). It should

be further noted that Weychert did not find Sensing types

achieving in a manner inconsistent with their aptitude--as

the Sensing preference was reported stronger and stronger,

grades did not get lower and lower. While "leveling off" of

grades occurred with increasing preference for Sensing

(i.e., grades did not get lower and lower as Sensing was

more strongly reported) , it should be remembered that for

extreme scores on all the other preferences aptitude

increased with extremity of score (Myers, 1962). Weychert

suggested that future studies include all scales of the MBTI

rather than just one (S-N) , and that all students rather

than just those with moderate to high preferences be

included. The failure to do either of the above could mask

interactions .

It is noteworthy that Myers (1962) issued a strong

warning against taking the strength of preference scores for

an individual as a direct measure of the excellence of the

type development.

Scores measure the reported strength of a pref-
erence. The report may for many reasons not
reflect the true strength, either through
accident, or through the well-known hazards of
self -report, or because of unusual appreciation
or supplementary development of the opposite
process or attitude, which would be admirable
rather than detrimental.

Even the true strength of the preference
may not accurately reflect the actual degree of

35

development. Therefore high scores should not
be equated to excellence, nor low scores to
deficit, unless there is positive external
evidence that excellence or deficit exists .
In the latter case the low score, while not
adequate evidence in itself, may provide a
clue which the counselor could find useful in
understanding and seeking to remedy the diffi-
culty, (p. 73) [underlining by the present
writer]

Following up on Myers1 statements regarding quantitative
interpretation of MBIT preference scores, Morse (1975) pro-
vided a table of distributions of random scores on the MBIT,
as well as a table of probabilities that any given type
could result by chance response. His purpose was to provide
"an objective and familiar decision-making technique by
which to evaluate a given score, independent of any other
score by any other person" (pp. 1-2). Morse's calculations
were utilized further in the development of the Discrepancy
Index described in Chapter III, where specific ranges for
low, moderate and high preference scores were set.

Other than size of the preference score as an indi-
cation of strength of preference, there are two other clues
internal to the scoring of the MBIT which warranted
investigation. They are clues as to the clarity with which
a preference is held. One is the split-half reliability
measure. If a person is clear about preference it can be
hypothesized that they will come out relatively the same on
the x-half as on the y-half. The other is the phrase
question/word-pair comparison. Again, if a person is clear
about or sure of the preference it can be hypothesized that

36
they will come out the same type on the phrase questions as
on the word-pair items when scored separately. It is
reasonable to assume that persons with low preference
scores, with x-half /y-half differences, and with phrase
question/word-pair item discrepancies may, indeed, not be
sure about what they trust and prefer. It may be hypothe-
sized that this person has poor type development (especially
if the preference in question is the person's Dominant func-
tion) . If such a person is operating at a deficit due to
poor type development, then we should be able to predict
that in areas of functioning related to that preference, the
person will show evidence of deficit there, too. For
example, a student with the above described discrepancies on
the MBTI for the S-N scale would be expected to show
evidence of academic difficulties. (The S-N scale, since it
deals with perception and how a person comes to "know" about
something, is increasingly being found to be of great impor-
tance in learning situations.) Myers (1962) found lower
internal consistency correlations in her academically
underachieving samples.

Only one study dealing with the problem of discrepancies
on MBTI profiles was found. Metts (1979) looked closely at
discrepancies on MBTI profiles of 113 learning disabled
adolescents. Because of the import of Metts findings for
the present study it is reviewed in depth here.

Metts (1979) addressed the need to help learning
disabled adolescents with problems related to self-knowledge

37
and self-identity in counseling situations. He pointed out
that many researchers and practitioners agree that inter-
vention with the personal development and self-identity
processes of the learning disabled adolescents is as impor-
tant as remediation of academic deficits. Jung's emphasis
on awareness and maturation of one's type preferences is
highly relevant here.

Metts reviewed contradictory research findings with
respect to the importance of low self-esteem and low self-
concept as contributing to the learning disabled adolescent's
problems. Negative results (that low self concept was not
a significant contributing factor) were obtained by Bruner
and Starkey (1974), Busby, Fillmer, and Smittle (1974),
Leviton and Kiraly (1975) , and McDonnell (1975) . Other
studies, however, found significant positive relationships
between self-concept and problems of L.D. students. Examples
cited by Metts were Black (1974), Rosenthal (1973), and
Rosser (1974) . Metts concluded that the differing
measurement devices employed to measure self-concept and
differing definitions of self-concept are largely responsible
for contradictory findings, and thus the noncomparability of
the studies. Metts suggested, "In spite of these problems,
the behavioral observations of Beare (1975) , Gif f in (1971) ,
Gordon (1970), Gow (1974), Siegel (1975), Thompson (1970),
and Wender (1971) suggest that the learning disabled
adolescent continues to experience problems in self-
identity" (pp. 12-13).

38
Metts1 study has four important implications for
education. First, use of the MBIT promises to meet a need
of learning disabled students to gain self-knowledge.
Secondly, an approach is offered where differences are
valued rather than discouraged. Normal aspects of the
student's personality are stressed. Jung's personality
theory has implications for the education of the whole
person, as Metts puts it--"balanced development of individ-
uality" (p. 79). The third area is that of focusing on the
social maturity of the learning disabled adolescent. The
adolescent learns not only about himself and his own person-
ality, but that of his peers- -understanding how he may be
like them and ways in which they may differ. Finally, Metts
concluded that the MBTI can be used in studies to further
explore personality variables in learning disabled students
with the hope of understanding and discovering better ways
of teaching them. Knowledge gained through employing the
MBTI with students has then a two-fold potential. One is
oriented toward personality development, the other toward
academic achievement — better personal understanding as well
as suggesting ways to enhance cognitive learnings.

Metts' basic research question was whether the distri-
bution of the 16 profiles of learning disabled adolescents
on the MBTI was the same as the distribution for normal
adolescents. If they differed significantly one could
question whether or not the learning disability had
influenced type development in some way.

39
Metts' L.D. sample had a similar type distribution
when compared with two normal samples. Only two differences
existed. The most important of these was that the L.D.
students more frequently expressed a preference for thinking
judgments (T) in comparison with the two normal samples. In
general there were more ISTP's and fewer ENFP ' s among the
learning disabled group. Finding only two significant
differences in the type distribution is not uncommon. If
a sample of high school students is divided into those
opting for college and those not, there would probably be
more than two significant differences between the profiles
of these two sub-groups.

Metts also looked very closely at discrepancies in
profiles. Only eight of 113 subjects (7.017=.) came out with
the same type formula (the four letters used to designate
type preferences) when x-y halves and word-pair/phrases on
the MBTI were scored separately. Employing the three
scoring methods (standard scoring, x-y halves, and word-
pair/phrases) 22.127o of Metts1 sample showed discrepancies
on one MBTI scale, 36.28% on two MBTI scales, 25.667, on three
scales, and 8.857. had discrepancies on all four scales.
Metts employed alternative scoring in the hope of verifying
type preference for subjects with low preference scores on
one or more MBTI scales. The high percentage of discrep-
ancies for his sample prevented this.

It is possible that the high percentage of discrepan-
cies in Metts' sample and the questions raised about

40
predominance of thinking may both be understood within the
framework of psychological type development. Metts gave two
possible explanations for the prevalence of thinking judg-
ment. One is that they may truely prefer thinking. The
other is that it may result from educational procedures
stressing clear, logical, objective thinking employed in the
remediation of deficits. Metts further noted that students
may have learned to distrust Feeling due to a lack of
social, academic, and familial success. Since behavioral
descriptions of L.D. students rarely connote traits associ-
ated with thinking judgment, Metts suggested further research
to determine whether the choice of thinking is really the
true preference.

The line of reasoning employed in the present study
suggests that even though the behavioral descriptions of
L.D. students do not frequently mention logical, analytical,
impersonal methods of decision making, the students may
still prefer thinking over feeling; however, they may not in
reality have good command over its use. In other words, the
large number of discrepancies may be related to the level of
type development the student has reached.

Another consideration is the possibility that judgment
(whether it be thinking or feeling oriented) may be the last
process to develop. If this were true it is not unlikely
that in any adolescent population, whether L.D. or not, the
decision making process would be the function over which the
person has the least command in its use, and hence, a greater

41
likelihood of discrepancies in reporting the preference. A
common finding is that reliability of the T-F scale improves
with age (McCaulley, 1978; Myers, 1962). Metts also found
significantly more discrepancies for subjects who reported
a preference for Feeling.

Metts1 study clearly shows the researcher's dilemma in
deciding whether to use profiles with low preference scores.
He supports Myers' procedure for "mutual verification of
results" (1962, p. 78). When low preference scores and
multiple discrepancies exist, discussion of MBTI results
with the person who took it is absolutely necessary from a
counseling standpoint. In this case, the subject can often
shed light on what "fits" and what does not. Metts reiter-
ated Myers caution regarding interpreting very low prefer-
ence scores--the smaller the score the more likely it is to
be wrong.

It was outside the perspective of the present research
to review studies where the direction of preference has been
of central importance (direction of preference being whether
the person prefers E or I, S or N, T or F, and J or P) .
This included studies where basic type differences between
students was the independent variable. There was no intent
to dismiss these studies lightly. Studies focusing on
direction of preference were important in the past and will
continue to gain in importance in the future.

It is interesting to note that in a study not dealing
centrally with the matter of psychological type development

42
Hogan (1970) noted several things which in this researcher's
opinion show probable evidence of the operation of dif-
fering levels of type development in his subjects. Hogan 's
study dealt with the subject of moral development. He
investigated personality dimensions related to holding one
or the other of two contrasting moral principles as the
basis for making decisions in situations of "moral uncer-
tainty." The "ethics of personal conscience" was contrasted
with the "ethics of social responsibility." Hogan found the
adherents of the "personal conscience" ideal (low scorers on
the Survey of Ethical Attitudes, Hogan, 1967) to be -NFP's
on the MBTI, and adherents of the "social responsibility"
ideal (high scorers on the Survey of Ethical Attitudes) to
be -STJ's on the MBTI (N = 57, g<.01 on all three scales of
the MBTI) . He points out that in an independent sample
(N = 37), adherents of "personal conscience" (low scorers)
were described on the Gough Adjective Check List (Gough,
1960; Gough & Heilbrun, 1965) as

rebellious (-.49) complicated (-.42)~"\

handsome (-.47) cynical (-.41) /

smug (-.45) sarcastic (-.41) (p = 05

uninhibited (-.44) progressive (-.41) f -

charming (-.44) vindictive (-.40) J

Those adhering to the "social responsibility" ideal (high
scorers) were described as

thoughtful (.47) sincere (.31)

good-natured (.43) helpful (.30)

conventional (.40) responsible (.29) C p = .05

conscientious (.37) planful (.28)

conservative (.32) honest (.27)

43

Hogan concludes that persons holding the "personal

conscience" ideal,

may be unconventional, liberal, and progressive;
however, they may also be capricious, undepend-
able, and anti-conforming. On the other hand,
persons whose moral judgments are guided by
rational and legal considerations [holding the
"social responsibility" ideal] are often
thoughtful, considerate and honest; yet, they
may also be conventional and overconforming. . . .
[they] tend to gravitate to positions where sup-
port of rules of conduct is viewed as positive
behavior. Low scorers, on the other hand often
seem to delight in changing the "system." Thus,
both high and low scorers on the test possess
unique virtues and drawbacks. (p. 211)

Hogan also comments that based on interviews, persons with

scores below 5 or above 30 (extremely low and high scores)

he considered to be highly sensitive to injustice--"sensi-

tive to the point of paranoia." It is this writer's opinion

that a large portion of the negative characteristics

indirectly and potentially ascribed to -NFP's in Hogan' s

study could be explained if we could differentiate and treat

separately the -NFP's who are mature and fully-developed (in

psychological type terms) and likewise for the -STJ's. It

is granted that the strengths of each of the psychological

types differ and each has their characteristic "blindspots"

and difficulties with which to deal, but this writer

suspects that psychologically mature "examples of their

type" (whichever of the sixteen types they may be) will

exhibit their weaknesses and deficits in far less drastic

ways than someone who is immature and less differentiated

in a developmental sense.

44
An Assessment Problem
We have reached a roadblock in psychological type
research. We know type differences do exist. We are only
beginning to learn how important these differences are.
What we do not know at this point is how to measure the
extent to which a person has developed and has command over
his preferences. Only then can we determine how much vari-
ance in observed differences between different types can be
attributed to differences in the types themselves, and how
much to differing levels of individual development within
each type. Level of development is being assessed in a day-
to-day clinical fashion based on theory. A data base is
needed if we are ever to leave this intuitive stage. Case
histories, anecdotal accounts, and successful experiences
with clients or students where we have operated from a
theory-base, while important, are not enough.

No studies, other than Metts (1979), either
experimental or correlational in nature, were found in the
published literature which dealt centrally with the topic of
assessing an individual's level of psychological type
development. While most investigators have recognized the
developmental aspects of psychological type, research and
application have been limited for the most part to explora-
tion of differences between persons of various psychological
types. The goal in most studies has been to determine how
the direction of preference (i.e., whether the person prefers
E or I, S or N, T or F, and J or P in Myers' system for

45
ascertaining type) relates to other variables, or to test
predictions based on basic type differences (e.g., Sensing
types will prefer a certain thing more than Intuitive
types). These are legitimate goals, as it is necessary to
determine empirically how psychological type manifests
itself in interacting with other variables. Such investi-
gations are, in fact, crucial in assessing construct valid-
ity of the MBTI . It is expected, however, that individuals
will differ not only in the direction of basic preferences
for the use of the psychological functions and attitudes,
but that they will differ also in the command that they have
over their preferred functions as well. In other words, it
is expected that some individuals will have developed their
preferred process or function to a greater extent than
others preferring the use of the same function. The degree
to which one has developed his preferred functions has
implications for how well he functions in day-to-day
situations which call upon powers of perception and
powers of judgment. The perspective of the present
study was at a more individual and micro-level than that of
dealing with the basic type differences in direction of
preference .

It has been proposed that persons giving inconsistent
and discrepant responses on the MBTI scales may be func-
tioning at a lower level of type development than persons
giving clear and consistent responses. In order to measure

46
the effects of clear and consistent versus discrepant and
inconsistent response against academic and nonacademic
variables a Discrepancy Index was constructed. Discrepancy
scores on each of the four MBIT scales were calculated for
each subject in the present study. The Discrepancy Index is
described in the section on Instrumentation in Chapter III.

CHAPTER III
DESIGN OF THE STUDY

This chapter describes subjects utilized in the study
and explains administration procedures. Reliability and
validity data for instruments employed in the study are
given. The rationale and method of developing the Discrep-
ancy Index are also presented. Hypotheses are restated and
statistical methods for data analysis are explained.

Subjects

At the request of school officials a total of 1,171
tenth, eleventh and twelfth-grade students from a metropoli-
tan high school participated in the project. Of this number,
555 students were subjects in the present study. For reasons
explained in the next section, 70 of the 555 students were
excluded from the study, leaving for the purposes of statis-
tical analyses, 485 subjects. Table 1 gives a breakdown of
subjects by grade level. Of this group 271 were females and
214 males. The 456 eleventh-graders whose data were
analyzed represent 727c of the total eleventh-grade population
of the school. The 29 tenth and twelfth-grade subjects were
members of mixed-level English classes where eleventh-graders
predominated.

The total school population was comprised of
approximately 2,000 students. Socioeconomic level of the

47

48

Table 1. Subjects broken down by grade level.

Grade level N

10th 15

11th 456

12th 14

Note: Total N = 485.

students covered an extremely wide range, with family
incomes ranging from less than $4,000 per year to what has
been termed the upper upper class. The school population
included both inner city and rural dwellers.

The only selection factors known to have been operating
in obtaining eleventh-grade subjects were the student's
absence on the days instruments were administered, the stu-
dent's choice not to participate, or the lack of informed
parental consent. Students were guaranteed total anonymity
in relation to the findings of this study.

Because the psychological type distribution of the
sample may be of interest to researchers it is presented in
standard "type table" format (as used at the Center for
Applications of Psychological Type, Gainesville, FL) in
Appendix C. The type distribution of students in the
present study was compared statistically with that of Myers'
sample of 9,320 Pennsylvania high school students. Results
of this comparison are also reported in Appendix C.

49
Procedure

Approximately one week prior to the administration of
the Myers-Briggs Type Indicator (MBTI) , teachers distri-
buted to their students letters explaining the nature of the
research and informed parental consent forms. Teachers col-
lected the consent forms prior to administration time. The
MBTI was administered to students by their classroom English
teachers during their standard 50 minute class period.
Students who did not finish in the 50 minute period were
allowed more time, as their schedules permitted, to complete
their responses. The decision was made by the researcher to
exclude from the study subjects who left blank. 45 or more of
the 166 items (N = 56). Another 14 subjects were excluded
because no data, other than the MBTI, were obtained for
them. For these reasons a total of 70 out of 555 were
dropped from data analysis.

For the purposes of another study on response style
bias operating in subjects" taking the MBTI (Kainz, 1978),
the standard Form F test booklet was given to 244 subjects
from the present study. The remaining 241 students were
given Form FR. Form FR had reversed response formats on 23
of the 166 items. Males and females sat on opposite sides
of the classroom. Distribution of test booklets proceeded
in a fashion so that every other student received a Form FR
test booklet rather than a Form F booklet, thus, male and
female students were provided an equal probability of being
assigned to the Form F group and the Form FR group.

50
Students were unaware that two forms were distributed.
Based on analyses of the results obtained from the alternate
forms, it was found that students did not respond differ-
ently to the two forms (Kainz, 1978). In the present study
the forms were considered equivalent.

In another 50 minute class period on the second day of
administration, combined classes of students were asked
under standardized directions from three test administrators,
to complete a modified form of Landfield's Repertory Grid
for assessing cognitive complexity. A major problem in this
administration session was that a large number of students
did not complete the procedure in the 50 minutes allotted.
Analyses of these data will be performed at a later date and
reported elsewhere.

Due to the impossibility of utilizing more of the
students' daily class time, teachers agreed to assist the
researcher by distributing and collecting from the students
additional materials that students took home following the
in-class testing. The materials in each take-home packet
were general directions for completing the enclosed ques-
tionnaires, the moral judgment questionnaire, the ego
development questionnaire with separate forms for males and
females (responses to which will be analyzed and reported
elsewhere), and a demographic data sheet.

While assessing the concurrent validity between two
instruments or scales measuring comparable or related con-
structs is a legitimate way to assess the validity of any

51
new measure (the Discrepancy Index proposed here) , it is
advisable to search for behavioral or performance-related
evidence where it is available and quantifiable. In the
present study the following additional data were obtained:
academic and overall grade point average; academic and
overall class rank/ standing. The school provided ethnic
codes for the subjects from their census list. Concern for
confidentiality plus time and financial constraints made the
cost of accessing reading test scores for all subjects via
the school system's computer prohibitive. English teachers'
judgments of reading levels were obtained instead. Because
the teacher judgments may or may not be reliable for all
students, they will be analyzed as supplementary data only
and results reported in Appendix A.

Instrumentation Employed in the Study
The Myers-Briggs Type Indicator

Description. The MBTI contains 166 forced-choice items
and was designed to yield preference scores on four separate
indices. The E-I index has two poles, extraversion and
introversion. This index reflects whether the person pre-
fers to focus perception or judgment upon the outer environ-
ment (E) or whether toward the inner world of one's own
thoughts and ideas (I) .

The S-N index has two poles, sensing perception and
intuitive perception. This index reflects whether the
person prefers to deal with the immediate reality as

52
revealed through the senses (S) , or the world of
possibilities revealed through the use of intuition (N) .

The T-F index reflects what kind of judgment the person
relies upon. One pole of the index is thinking (T) which is
the preference for deciding based upon impersonal criteria
and logic. The other pole is feeling (F) which is the pref-
erence for deciding based upon what is, from a subjective
standpoint, valued in the situation. As described by Jung,
thinking and feeling are both rational processes.

The J-P index was designed to reflect whether the
person prefers the use of perception or the use of judgment
as a way of dealing with the outside world (in other words,
when the person is behaving in an extraverted way) .
According to Myers (1962) , the difference here is "between
the judging people who run their lives and the perceptive
people who just live them" (p. 58). Perceptive types tend
to want to perceive or find out more about something rather
than to decide about it. Judging types tend to want mat-
ters decided.

Validity. Because the MBTI is a theory-based instru-
ment, construct validity is of primary concern. With
respect to validity, Kerlinger (1964) states, "Validity
. . . is much more than technique. It bores into the
essence of science itself. It also bores into philosophy.
Construct validity, particularly, since it is concerned
with the nature of 'reality' and the nature of the

53

properties being measured, is heavily philosophical"

(p. 473). Fox (1969) defines construct validity as,

The ability of the instrument to distinguish
between groups known to behave differently on the
variable or construct under study. . . . There
are two sources for the construct or criterion on
which the groups can be said to differ. One
source is overtly relevant behavior. ... A fre-
quent parallel to this behavioral basis for con-
struct validity is the development of the groups
to be tested on the basis of indirectly related
behavior. . . . construct validity is at best
evidence that at a gross level the instrument
measures what it seeks to measure. (pp. 372-373)

In a critical review of the MBTI for Buros 8th Mental
Measurements Yearbook, Coan (1978) gave the following com-
ments (concerning chiefly construct validity) about the
complexities and problems in assessing the Jungian variables,

The Myers-Briggs Type Indicator represents a major
effort to capture the Jungian personality
typology in a psychometric instrument. This is
a formidable task. The Jungian typology rests on
a sophisticated and intricate analysis of the
basic modes of variation in human experience.
. . . To assess fully an individual's type, we
must identify his dominant attitude and his
superior and auxiliary functions. For a number
of reasons, this is a complicated problem. . . .
The Type Indicator is the product of much thought
and research, and its authors endeavored to take
some of these problems into account. They have
not fully solved them, but neither has any other
test constructor. . . . Research has shown the
instrument to be useful for many applications.
. . . On the whole, the test clearly merits
further research and use. (pp. 973-975)

McCaulley (1978), in a voluminous monograph on the

application of the MBTI to Medicine and over twenty other

health-related professions, provides further information on

validity of the instrument. A four-page table (McCaulley,

1978, pp. 28-31) presents the correlations of MBTI

54

continuous scores on the four indices (E-I, S-N, T-F, and

J-P) with scales from other personality instruments used

with medical students. McCaulley (1978) explains,

The table gives the direction of the significant
correlations in type letters [E or I, S or N,
T or F, and J or P] so that the reader can use
the table not only to assess concurrent and
construct validity, but also to gain a better
understanding of qualities associated with
each preference. (p. 27)

Correlations given are for the 16 Personality Factor Test,
the Omnibus Personality Inventory, the Opinion, Attitude and
Interest Survey, the Allport-Vernon-Lindzey Study of Values,
and others.

With respect to predictive validity, the interested
reader should refer to numerous articles cited in the
extensive METI Bibliography (1981) which concern
themselves with prediction. McCaulley (1978) adds, "The
Myers Longitudinal Sample [McCaulley, 1977] is one of the
major research efforts concerned with predictive validity of
the MBTI, since the research question concerned the selec-
tion of specialties 12 or more years after taking the MBTI

as freshmen in medical school" (p. 27).

2
The Manual (Myers, 1962) is another rich source of

data on validity of the MBTI. Chapter IV provides corre-
lational data for the MBTI with (1) the Gray-Wheelwright

2

The original Manual was published in 1962 by Educational

Testing Service. Revised manuals are due for publication

by Consulting Psychologists Press in 1981. Plans exist

for an administrative manual and a statistical manual.

55
Psychological Type Questionnaire (designed with the same
purpose as the MBTI) , (2) the Strong Vocational Interest
Blank, (3) The Allport-Vernon-Lindzey Study of Values,
(4) the Edwards Personal Preference Schedule, (5) the
Personality Research Inventory, and finally, (6) non-test
variables such as faculty ratings, turnover in utility jobs,
and MacKinnon's data on creativity. Chapter V provides
correlational data for the MBTI and scholastic performance
and aptitude (such as SAT scores, Terman's Concept Mastery
Test scores, grade point averages, IQ data, Davis reading
test scores, and others). Hicks (1970) states, "The MBTI
(Myers, 1962) represents perhaps the most satisfactorily
validated, non-cognitive measure that falls into the forced
choice category" (p. 200).

Reliability. Myers (1962) reports split-half
reliability data for several groups (a measure of internal
consistency) . Myers utilized a logical split-half pro-
cedure and applied the Spearman- Brown Prophecy formula to
obtain correlations between the X and Y halves of the MBTI.
Split-half reliabilities range from .75 to .94. Myers notes
that a wide range of age, intellectual ability, and socio-
economic status is included in these data. She points out
that the only coefficients below .75 are for the under-
achieving eighth-grade male sample and the non-college
preparatory twelfth-grade males. The coefficients for these
two groups on the T-F scale are .44 and .60, respectively.
The T-F scale fares lowest of the four scales in general,

56
but it should be noted that for the two college samples
Myers reports, its "disadvantage" has disappeared (with
coefficients of .83 and .86). McCaulley (1978, p. 33) sup-
plies additional split-half and alpha measures.

Carlyn (1977) found MBTI test-retest reliabilities for
the type classifications to be satisfactory. McCaulley
(1978) provides test-retest data for both MBTI continuous
scores and type categories (E-I, S-N, T-F, and J-P) . Test-
retest correlations (over periods of from 8 to 21 months)
range from a low of .60 for the T-F scale to a high of .83
on the S-N scale. For two groups retested after a two
month period, reliabilities ranged from .69 to .83.
McCaulley points out that those people who change type upon
retest usually do so on one or two letters, rarely on three
letters, and almost never on all four letters. Of the seven
samples presented, change on all four letters occurred only
in one sample and then only 170 of the time. For a personal-
ity instrument the MBTI's test-retest reliabilities are
quite respectable.

A final point with respect to reliability of the MBTI
deserves discussion, as it bears centrally on the problem
investigated in the present study. This is the issue of
psychological type development. Myers1 hypothesis as stated
by McCaulley (1978) is that, "Samples assumed to have
reached higher levels of type development will have higher
internal consistency reliabilities (as reflected in split-
half correlations provided by Myers) . The T-F scale is

57

particularly likely to be low in underdeveloped samples"

(p. 32). With respect to the T-F index reliabilities, Myers

(1962) states, "The possibility would seem to exist that the

relative uncertainty on TF may reflect a lesser development

of the judging process, which may prove to be a significant

characteristic of such samples" (p. 20).

Because the MBIT is a self-report instrument, we have

the thorny question of how much variance is due to the

reliability of the instrument, and how much is due to the

reliability of the respondent. McCaulley (1978) addresses

this point when she suggests that studies should be done in

which test-retest reliabilities are computed separately for

individuals originally clear in their preferences, and for

those with initially low (indeterminate) scores. The problem

as stated best by Myers (1962) boils down to the fact that

at present,

The potent but as yet unmeasurable variable of
"type development"--!. e. , the extent to which
the person actually has developed the processes
and the attitudes which he prefers—enters every
equation as an unknown quantity. Also unknown
is the mean level of type development for any
of the various samples that are tested, and how
many individuals in each should be expected to be
answering virtually at random because their type
is insufficiently developed to govern their
responses. (p. 19)

The Discrepancy Index developed for the purposes of

the present study and described later in the section on

Instrumentation is an attempt at incorporating several

clues as to the level of type development of the respondent

58
into a single index. The predictive power of the new index
can then be assessed against outside criteria that are
related to education and to psychological development in
general.
The Moral Judgment Questionnaire

Description. The Moral Judgment Scale (MJS) was
developed for the purpose of ascertaining levels of moral
reasoning corresponding to those of Kohlberg (1958) in an
objective- scoring format. According to Maitland and
Goldman (1974) Kohlberg' s interview procedures, "while com-
prehensive and well based on theoretical foundations,
require great expenditures of time and careful administra-
tion" (p. 699). The administration and scoring format of
the MJS was ideally suited for use in the present study.

The MJS consists of 15 vignettes, each followed by "a
question aimed at evoking one particular issue of moral
judgment and six alternatives representing characteristic
modes of thought about the issue at each of the six stages
of moral development" (Maitland & Goldman, 1974, p. 700).
The order of presentation of the six stage-representative
responses for each vignette was made random in the con-
struction of the instrument.

Kohlberg (1958, 1963, 1964, 1969, 1971) defines the
development of moral judgment as a dynamic process which
proceeds as the child interacts with its social environ-
ment. Kohlberg' s schema identifies six stages or levels of
moral development. The first stage is the "punishment and

59
obedience" orientation. The second stage is termed "naive
instrumental hedonism." In the next stage the primary focus
is that of winning approval and maintaining good relations
with others. The fourth stage involves a "law and order"
orientation, where conformity serves to avoid censure by
authority and the guilt that might accompany such censure.
Level five is the morality of social contract. One obeys
laws because they are the outcome of democratic process.
The highest orientation in Kohlberg's system (level six) is
that of reliance on "individual principles of conscience."
Each successive stage of moral development in Kohlberg's
conceptualization transmutes, restructures and displaces the
preceding stage or level.

Validity. The authors pointed out two important differ-
ences between the MJS and Kohlberg's interview procedures.
First, there is a loss of personal interaction with the MJS.
In order to compensate for this, the authors personalized
the vignettes by asking for judgments on situations where
they themselves are involved (as opposed to judgments on
behavior of a third party). Secondly, it is noted that
subjects in Kohlberg's procedure provide or produce the
response on their own, whereas in the MJS they choose from
responses that have already been provided for them.
Maitland and Goldman (1974) state, however, that "While the
subject's decision to select a justification is a somewhat
different task than that of Kohlberg's interview procedure,
in which the subject supplies a decision and justifies it

60
through probing questions by the interviewer, this
difference is seen as relatively minor" (p. 703). They
point out that the matter of equivalence of selecting a
justification with the spontaneous production of one, is an
open question, and further stated that the direct computa-
tion of a validity coefficient was not possible from their
data. Maitland and Goldman (1974) did find the mean scores
on moral judgment for their subjects to be in accordance
with Kohlberg's expectations for that age group (fifteen to
seventeen-year olds) . The authors compared MJS scores of
their 36 subjects with Moral Maturity Scale scores of 25 of
Kohlberg's subjects of similar age and academic preparation
(Gilligan, Kohlberg, Lerner, & Belenky, 1971) by multi-
plying the MJS scores by a factor of 100/15. Kohlberg's
mean Moral Maturity Scale score for this group was 364. The
converted MJS score for Maitland and Goldman's subjects
(from same age and academic grouping) was 365.

Wilmoth and McFarland (1977) made a comparison of four
measures of moral reasoning. Contrary to Maitland and
Goldman's findings, Wilmoth and McFarland did not conclude
that the Objective Moral Judgment Scale validly assesses the
Kohlberg stages, nor did they report it to have good reli-
abilities. They did find the Maturity of Moral Judgment
Scale developed by Hogan (and reviewed in relation to MBTI
findings in Section II, Chapter II) to be strongly related
to subjects' classifications on the Kohlberg stages, but its
continuous scores did not permit clear stage classification.

61
The goal- of Wilmoth and McFarland's study was to use
common age subjects to assess inter-instrument relation-
ships, because the Objective MJS of Maitland and Goldman had
been administered only to pre-college age students and
Hogan's Maturity of Moral Judgment Scale to under-graduate
age subjects. Wilmoth and McFarland used a sample of 70
adults enrolled in three extension graduate courses for edu-
cators. Over 807o of the subjects were elementary or
secondary teachers, counselors, and/or administrators. Age
ranged from 21 to 51 years with mean age being 30.45 years.
Only two dilemmas from Kohlberg's interview procedure
and two from the Sexual Moral Judgment Scale (Gilligan,
Kohlberg, Lerner, & Belenky, 1971) were utilized for the
comparison. Subjects scores on the Objective MJS were not
found to be significantly related to their scores on the
four moral stories combined (F = 1.82, df = 4/64, p_<.15).
Utilizing Scheffe's post hoc comparison procedure, the
authors found none of the means to be significantly differ-
ent from one another (it is noted, however, that if the
one-way analysis of variance was not significant, no
significant post hoc comparison could be expected) .

Wilmoth and McFarland concluded that their data offer
little support for the Objective MJS as constructed by
Maitland and Goldman. They report substantially lower
internal reliabilities than those reported by Maitland and
Goldman, and they ascribe the absence of relationship

62
between the Objective MJS and the four dilemmas used as the
independent variable to the Objective MJS ' s lower reli-
ability.

In spite of contradictory findings, the decision was
made to employ the Objective MJS in the present study. This
was due primarily to its objective format and secondly to
the acceptable reliability and validity data obtained for
subjects in the same age range as those in the present
study. The hypothesis presented in the present investiga-
tion regarding moral reasoning is concerned foremost with
predicting relative levels of moral judgment from analysis
of discrepant responses and strength of preference for the
Thinking-Feeling scale of the MBTI . The Objective MJS was
seen as adequate for exploratory purposes.

Reliability. Maitland and Goldman (1974) assessed the
test-retest reliability of the MJS by successive administra-
tion to 60 subjects at a ten-day interval. Subjects were
public school students from seventh" to twelfth grade (ages
twelve to nineteen) . The Pearson product-moment correlation
for this group was .83. For a sample of 22 tenth and
eleventh-graders it was .60. The authors concluded that
scores were quite stable over a ten-day period.

Maitland and Goldman (1974) offer evidence that stage
achievement can be reliably assessed by the MJS. They
reasoned that "since stage achievement is somewhat inde-
pendent of specific issue selections" (p. 700), item
content differences might be expected to reduce internal

63
consistency reliability estimates. They obtained a
corrected split-half reliability of .71. They computed a
Kuder-Richardson Formula 20 reliability estimate (r = .67)
in order to rule out particular order effects on the split-
half reliability estimate obtained. In each case a sample
of 125 subjects was employed. Subjects were evenly distri-
buted by age from 12 to 23.

Over short periods (ten days) the MJS appears reliable.
However, its reliability over longer intervals needs assess-
ment .
Demographic and School Performance Data

Demographic. The subject's age was recorded on each
MBTI answer sheet. Information as to grade level and race
or ethnic origin was provided on computer-generated lists
from the school. A questionnaire designed to obtain birth
order information and an inferred estimate of socioeconomic
status was returned by so few students that a range of
socioeconomic level was all that could be provided by school
personnel .

School performance. Academic and overall grade point
average and class rank were obtained for each subject from
a computer-generated list provided by the school. Reading
test scores were available for so few students that English
teachers were asked for estimates of their students' reading
levels. Due to the questionable reliability of these
estimates, they were utilized only in supplementary data
analysis .

64
The Discrepancy Index

A Discrepancy Index (DI) based on individuals'
responses on the MBTI was constructed for use in the present
study. The complex computer- scoring program developed at
the Center for Applications of Psychological Type (CAPT)
provided the detailed scoring utilized in the DI. Four
scores are rendered for each subject on the DI as follows:
(1) a strength of preference score (DI1) ; (2) a word-pair/
phrase item discrepancy score (DI2) ; (3) a split-half
difference score (DI3) ; and (4) an overall discrepancy score
(DI4) .

Strength of preference scores (D1EI, D1SN, D1TF, and
DlJP) were computed by giving a point value from one to four
for each of four possible ranges of scores. Two sources of
information were drawn upon in determining where to set
score ranges for each of the four MBTI scales (E-I, S-N, T-F,
and J-P) . The first source comes from guidelines provided
by Myers which the CAPT center utilizes in its extended
scoring report. Table 2 presents these confidence regions
in continuous score format.

Support for the confidence regions set by Myers was
found in a paper presented by Morse (1975). Morse's goal
was to provide "a measure of the likelihood that a pref-
erence as strong or stronger than a given score could have
occurred by chance" (1975, p. 1). The distribution of
random scores for the four MBTI scales is provided by Morse

65

t^.

O

O

o

CN

<r

cu

iH

rH

rH

u

M

1

iH

O

.H

c-0

Ph|

1

00

o

rH

o

00

M

1

i-H
CN

rH

o

M

o

6

o

<t
l-H

CO
CO

^1

1

ON

o

rH

N

1

cn

CTi

O
CN

rH

>->l

1

cn

|-j|

cu

rH

M

O

LO

CU

r)
O

M

1

Pm|

1

M

1

H

E

cu

U

a

r-H

oo

H

o

a

O

o

CN 1

cu

rH

H

rH

m

u

n

0)

rH

<4H

CU

u

a

CO

o

vO i

en

CO

i-i

a\

00

VO 1

CU

(0

i— i

3

4-1

H|

1

H|

1

H|

•

H|

u

CJ

ON

CN

CTi 1

o

(fl

o\

CT\

r>« 1

m

0)

vO

,0

4-1

O

cu

d

o

o

CN

O 1
-* 1

M

O

>i

rH

<-t

•-i i

CU

B

m

u

S

21

1

0)

SB|

1

55 1

1 1

d

CU

21

u
o

H

o

00

H 1

i-i

CD

O 1

e

o

CN I

CU

i-H

CJ

H

(U

r-H

CD

H 1

<4H

sr

a

M

O

CU

rH

0)

a)

d

n

h

4H

cu

a.

cu

CU

u

MH

r4

a>

00

cn

cu

P.

U-l

d

CO

u

ro |

O 1

cu

O 1

0

cu

o.

tn i

a)

4J

00 1

u

CU

vO i

u

4-1

rH

4J

W|

a)

W|

1

CQ|

1 1

en

ot|

IH

J2

)-i

u

O

00

CTi 1

cu

CN 1

CO

o> ]

>,

•H

ON I

T>

o\ 1

01

P» 1

U

o>

H

o

rH

cu

in

cn

B

CJ

>

cu

cn

cn

en

en

u

cu

r*» i

cu

o l

01

O i

a)

o

00

O 1

oo

CN 1

oo

-d- 1

00

e

d

H 1

c

r-i 1

c

^ 1

d

a)

cd

n3

cfl

u

M

H|

1 !

i-i

H|

1 i

u

H|

1

u

M|

0

cu

-H 1

ai

00 I

0)

rH |

0)

rH

M

O 1

u

O 1

1-1

CN 1

M

■<f

o

rH 1

o

^ 1

o

rH !

O

rH

o

a

CJ

CJ

CO

en

en

en

en

en

en

en

cn

3

3

3

3

cn

0

O

O

O

cu

3

rn i

3

O 1

3

3

i-h

d

ON 1

C

00 1

C

O 1

C

iH

•H

■H

vO 1

•H

u

4-1

d

Ml

1 !

4-1

d

W|

1 i

4-1

c

M|

1 1

4J

d

W|

o

0

a\ i

o

CN 1

o

ON |

o

CTv

o

ON |

o

o\ I

u

^ 1

o

m

66
(Table 3) . Continuous scores falling outside each range are
significantly different from chance at the alpha level
specified. While Morse provides the ranges for alpha levels
from .50 to .001, only the .05 and .01 levels are reported
here.

Table 3. Continuous score ranges differing significantly
from chance.

Alpha levels (2-tail)

MBTI scale

E - I

S - N

T - F (Males)

T - F (Females)

J - P

.05

01

79

- 127

71

- 135

67

- 117

61

- 123

79

- 123

73

- 129

69

- 119

63

- 125

79

- 129

71

- 137

Note: Data from Morse, 1975, p. 8, Table 4.

The agreement between Myers' suggested ranges for con-
fidence placed in reported preferences (Table 2) and Morse's
confidence regions derived through statistical methodology
(Table 3) is quite high. Because Myers' continuous score
range for a clearly reported preference for each scale
reaches or exceeds those reported by Morse as differing from
chance at at least the .05 confidence level, they were uti-
lized in the present study to designate strength of prefer-
ence categories (slight, moderate, clearly reported, and

67
very strongly reported) . A point value of 1 was assigned to
the very strongly reported category, 2 to the clearly
reported category, 3 to the moderately reported category,
and 4 to the slightly reported preference category.

Word-pair/phrase item discrepancy scores (D2EI, D2SN,
D2TF, and D2JP) were computed for each subject. If the
subject did not change type (came out with the same prefer-
ence on the word-pair items as on the phrase question items
when scored separately) the zero difference was assigned a
point value of 1. If the subject had a difference in con-
tinuous score of from 1 to 6 points, a 2 was assigned. A 3
was assigned if the difference in continuous score points
ranged from 7 to 12. A 4 was assigned if the continuous
score point difference was 13 or greater.

Split-half difference scores (D3EI, D3SN, D3TF, and
D3JP) were computed by scoring the x-half and the y-half for
each scale of the MBIT separately. If the absolute differ-
ence in continuous score points ranged from zero to 5 , a
point value of 1 was assigned (whether the subject changed
type or not). A 2 was assigned for differences from 6 to 11,
a 3 for differences from 12 to 17, and a 4 for differences
of 18 or greater.

The fourth part of the Discrepancy Index was total
scores combining discrepancies (D4EI, D4SN, D4TF, and D4JP)
for each subject across each scale of the MBTI . Total scores
were computed by combining the point values previously
calculated for each MBTI scale as follows:

68

D4EI = D1EI + D2EI + D3EI

D4SN = D1SN + D2SN + D3SN

D4TF = D1TP + D2TF + D3TF

D4JP = D1JP + D2JP + D3JP

The fifth and final part of the Discrepancy Index was

an additional set of combined scores (Dl , D2, D3, and D4)

where :

Dl = D1EI + D1SN + D1TF + D1JP

D2 = D2EI + D2SN + D2TF + D2JP

D3 = D3EI + D3SN + D3TF + D3JP
D4 = Dl + D2 + D3

Dl combines point values assigned for strength of reported
preference across all four MBIT scales. D2 combines point
values assigned for word-pair/phrase item discrepancies on
all four MBTI scales. D3 combines point values assigned for
split-half differences on all four MBTI scales. D4 combines
the total points computed for strength of preference, word-
pair/phrases, and split-halves on all four MBTI scales.

While beyond the exploratory scope of the present
study, the Discrepancy Index is open to differential
weighting of its component parts. Components which contri-
bute little to variance may be redesigned by adjusting the
ranges, and re-evaluated, or dropped altogether. Other
components could be differentially weighted as merited.

Statement of Specific Hypotheses
H Students with low Discrepancy Index scores will employ
higher levels of moral reasoning than students with high
Discrepancy Index scores.

To test this hypothesis the following discrepancy
scores will be examined;

69

1.1 reported strength of preference for Thinking/
Feeling scale

1.2 word-pair/phrase discrepancy on Thinking/Feeling
scale

1.3 split-half difference on Thinking/Feeling scale

1.4 total and combined index scores (D4EI to D4JP
and Dl to D4)

2
H Students with low Discrepancy Index scores will have

higher a) overall and academic class rank and b) higher

overall and academic grade point average than students with

high Discrepancy Index scores.

To test this hypothesis the following discrepancy scores

will be examined:

2.1 reported strength of preference for E-I, S-N, T-F,
and J-P scales

2.2 word-pair/phrase discrepancies

2.3 split-half differences

2.4 total and combined index scores

3
H Students with low Discrepancy Index scores on the

Thinking/Feeling scale when it represents their theorized

Dominant function, will employ higher levels of moral

reasoning than students with high Discrepancy Index scores

on the Thinking/Feeling scale when it represents their

theorized Dominant function.

To test this hypothesis the following index scores for

students with either Thinking or Feeling as their theorized

Dominant function, will be examined:

3.1 strength of reported preference for Thinking or
Feeling

3.2 word-pair/phrase discrepancy on T or F

3.3 split-half differences on T or F

3.4 total and combined index scores

4
H Students with low Discrepancy Index scores on the

Sensing/ Intuition scale when it represents their theorized

70
Dominant function will have a) higher overall and academic
class rank or b) higher overall and academic grade point
average than students with high Discrepancy Index scores on
the Sensing/ Intuition scale when it represents their theo-
rized Dominant function.

To test this hypothesis, the following discrepancy
scores for students with either Sensing or Intuition as
their theorized Dominant function, will be examined:

4.1 strength of reported preference for Sensing or
Intuition

4.2 word-pair /phrase discrepancy on S or N

4.3 split-half differences on S or N

4.4 total and combined index scores

Data Analysis

Scoring of MBTI optically- scanned answer sheets was
accomplished via the computer program of the Center for
Applications of Psychological Type with continuous scores,
word-pair/phrase item scores, and split-half scores pro-
vided on standard 80 column computer cards. The Moral
Judgment Scale was hand- scored by the researcher. Demo-
graphic data, school performance data, and MJS scores were
recorded on computer cards. The 20 Discrepancy Index
scores for each subject were computed and coded using
Version 7.1 of the Statististical Package for the Social
Sciences (Nie, Hull, Jenkins, Steinbrenner & Bent, 1975).

Statistical procedures were performed at the Northeast
Regional Data Center of the University of Florida using
Version 7.1 of the Statistical Package for the Social
Sciences (SPSS) . SPSS subprograms utilized were Frequencies,

71
Pearson Correlations, and Breakdowns providing one-way
analysis of variance and linear trend tests. In the pre-
liminary stages of analysis subprograms Scattergram and
Factor Analysis were also used. Post hoc tests for sub-
groups where the overall F tests were significant at at
least the .05 level were performed by hand employing
Scheffe's S test procedure. The Selection Ratio Type Table
program (Allen & Kainz, 1976) was utilized to obtain chi-
square statistics for psychological type distributions and
for comparison purposes between the sample employed in the
present study and that of 9,320 Pennsylvania high school
students (Myers, 1962). Type distribution data are reported
and discussed in Appendix C.

The independent variables consisted of the 20
Discrepancy Index scores computed for each individual. There
were seven dependent variables; overall class rank, academic
class rank, overall grade point average, academic grade
point average, Moral Judgment Scale total score, Moral
Judgment Scale mean score, and teacher estimates of student
reading level. Both grade point average and class rank were
utilized as criterion variables even though they are nearly
perfectly correlated (one derived from the other) , because
class rank reflects the student's academic performance in
relation to that of classmates and to that extent provides
descriptive information not readily ascertained from grade
point average. Teacher estimates of reading level were
analyzed and reported as supplementary data only

72
(Appendix A) . Ordinal level measures were obtained for all
variables in the study and they were treated as though they
were interval level in data analysis. According to
Labovitz (1970) interval statistics can be applied to
ordinal-level variables, especially when the nature of the
research is exploratory.

CHAPTER IV
RESULTS

This chapter provides the results from one-way analysis
of variance and other data analysis procedures, including
post hoc comparisons. Significant findings for the four
hypotheses are presented.

One-way Analysis of Variance

One-way analysis of variance (ANOVA) was performed
for the six criterion variables utilizing version 7.1 of
the Statistical Package for the Social Sciences (SPSS,
Nie, Hull, Jenkins, Steinbrenner & Bent, 1975). Significant
results are reported in Tables 4 through 7. Supplementary
analysis of teacher estimates of reading level are reported
in Appendix A. In spite of the exploratory nature of the
study, only those differences reaching the .05 level of
significance are reported in Tables 4 through 7 and the
Appendices .

The two basic assumptions for legitimate use of ANOVA
were met--the assumptions of normality and homogeneity of
variance. With respect to normality the question of impor-
tance was, do the criterion variables under study have a
normal or near-normal distribution, and examination of
means, medians, and values for skewness and kurtosis
revealed that they do. In addition, measurement in the

73

74
present study reached at least the ordinal level. Labovitz
(1970) argues that interval statistics (parametric) can be
applied to ordinal-level variables, and that error associ-
ated with such treatment is offset by the use of better
developed, more sensitive, and more clearly interpretable
statistics .

The assumption of homogeneity of variance was tested.
The F-test for homogeneity of variance (Kohout, 1974) was
employed for this purpose. In 240 possible comparisons of
sample variances only two reached significance at the .01
level and three at the .001 level (these comparisons were
between groups 1 and 2, and 1 and 3 for D1TF with overall
grade point average, and between groups 1 and 2, 1 and 3,
and 1 and 4 for D1TF with academic grade point average) .
In three of the five cases the obtained value for F just
barely exceeded the F table value. For example, with
degrees of freedom 40,172 the obtained F value was 1.58
where the table value for F was 1.56. In 235 comparisons
sample variances did not significantly differ. In general
then, sample variances did not differ widely enough to war-
rant assuming inequality.

Tests of linearity (linear trend tests) were per-
formed to determine whether the relationship between the
dependent and independent variables was solely a linear
relationship (Nie, Hull, Jenkins, Steinbrenner 6c Bent,
1975) . Linear and non-linear components of variance
explained are given in Tables 4 through 7.

75
Post Hoc Comparisons

Considering the exploratory nature of the study a
liberal test for the post hoc comparisons would have been
desirable. Two such tests were considered. Unfortunately,
Tukey ' s method of multiple pairwise contrasts could not be
used because it requires equal sample sizes (Marascuilo,
1971). Duncan's method was not chosen because according to
McNemar (1969), "The frequently advocated Duncan 'new multi-
ple range test' is currently under suspicion by mathematical
statisticians" (p. 324). Because of its general acceptance
and wide use (albeit extremely conservative) Scheffe's S_
test was selected and utilized in two formats. In a fashion
consistent with the direction of the hypotheses posed, all
possible pairwise comparisons of group means were made. In
addition, non-pairwise contrasts (combining groups 1 and 2
for comparison with groups 3 and 4) were performed. These
post hoc analyses were made only where the overall F for the
one-way ANOVA reached the .05 level of significance. Only
those post hoc comparisons reaching the .05 significance
level are reported in Tables 4 through 7 and Appendix A.
Selection Ratio Type Table Analyses

The frequency distribution of the type preferences
themselves must be closely scrutinized before the analysis
of variance results are interpreted. For each of the four
groups of each discrepancy index the selection ratio type
table program, SRTT (Allen & Kainz , 1976) was utilized to
locate any significant differences in type distribution.

76
As these analyses were not part of the primary focus of the
present study, they are reported and discussed in Appendix
D.

Frequencies

Metts (1979), in working with learning disabled stu-
dents, raised the question as to how many discrepancies on
MBTI scales can be expected to exist in normal samples.
Descriptive statistics for the Discrepancy Indices in the
present study are found in Appendix B. For all 20 Dis-
crepancy Index groups, the mean for each index as well as
the N and percent of sample for each of the four groups
within each index is presented.

Results of Hypotheses Testing

Hypothesis 1.1 was supported (see Table 4). Students
who reported higher preferences for either Thinking or
Feeling had significantly higher Moral Judgment Scale total
and mean scores than students who reported low preferences
for Thinking or Feeling. Hypotheses 1.2 and 1.3 regarding
word-pair/phrase discrepancy and x-y half differences were
not supported.

Hypothesis 1.4 received some support. The relationship
between D4TF and MJS score was not significant at the .05
confidence level; however, means were in the exact order
predicted, with students with high preferences for T or F,

no word-pair/phrase discrepancy, and low x-y half differ-

1 4
ences having highest MJS scores. With respect to H

(dealing with total or combined index scores) there was

col o-l

•Homo

ot| o-I

w|

O 0) -u \<u

•H CO

01

aj

LH -H

cd

U-i

■a e

jj

0)

M -H

c

u

h ca

Q

w

en o.

U

w

G-l

77

m en

I~> t-H
CO CO

O rH
-3" CO

IX

M|

78
a significant relationship between D4JP and MJS score. The
relationship appears non-linear. Group 3 had significantly-
higher MJS scores than did group 2. For J-P then, the
students expressing only moderate preference for J or P,
with slight word-pair/phrase discrepancy, and moderate
split-half difference, had the highest scores on the MJS.

Hypothesis 2.1 received support. Students who reported
the highest preferences for either Extraversion or Intro-
version (on D1EI) had significantly higher overall class
rank and overall grade point average than students expressing
lower preferences for E or I . The same trend existed for
academic class rank and academic grade point average, but
results did not reach the .05 level of significance.

Students who reported the highest preferences for
either Sensing or Intuition (on D1SN) had significantly
higher academic class rank and academic grade point average
than students reporting lower preferences for Sensing or
Intuition. The same trend existed for overall class rank
and overall grade point average, but results were not
significant at the .05 level.

Students expressing the highest preferences for either
Thinking or Feeling (on D1TF) had significantly higher
overall and academic class rank and overall and academic
grade point averages than students expressing only moderate
or slight preferences for T or F.

Another significant non- linear relationship appeared
for the J-P scale as it did in H (see D1JP on Table 5).

79

^

00 rH

m s-*.

cn <-N

CM O

cn in

O H

H in

w| a

1

• o

• o

cn o

CM o

p~ ^-^

CM •

CM •

CM •

jj

<r

-<r

-3"

<r

c

l0

>^>

^

<JS

CO

cn

cn

cn

cn

cn

•H

01

tn

4.)

cn

A

A.

A.

A

UH

■H

CO

CM

CM

CM

CM

•H

2

M

c3

<^J

■^

<Ji

C 1

4-1

ai

ai

01

0)

rH

rH

iH

^H

QJ

0)

a)

co c

•H

C

a

P

c

C

c

c

c

•H 0

CO

0

o

o

o

o

a

(X

a

a

0

o

o

cn c

a.

CJ

c

C

c

c

00

oo

00

00

c

c

c

,-^

,-s

en ^-s m iH

vO r-.

H /-N

.— 1

cn

co m • o

av in

• rH

cn o

r» i—i

<s

in

r^

m

/^

• O vD o

• o

vO O

\45 o

<■ o

•

O

rH

o

</i| a.

I

m • cm •

o> •

o •

<t •

<r •

ON

CM

CO •>»' H ^

m ^

iH v— '

• '

• v-x

m

^~^

•

"s"^

^

w|

4J

c

cn

CO <t

<r

■vf

<t

«*

<r

<r

cfl

cn

•-

cj a)

u

VQJ

a a.

a.

Q,

a

a.

Pu

a

•h cn

tn

C4-J

oo oo

00

00

oo

00

oo

00

CM -H
•H S

CO

u

CM
0)

A A

A

A

A

A

A

A

4-1

JS

0)

a)

01

01

rH H

CN

i-(

i—l

■H

cn

cn

a)

{JO -H

c

a

C

c

c

c

Pi

•H cB

0

cn

0

0

o

0

& P.

a

a.

a.

P.

P.

Pu

o

en a,

u

c

c

c

c

oo oo

00

00

00

00

00

60

c

*

CM

p~.

cn

a\

r^.

P~

P»

CO 1

CM

iH

CM

CM

<r

CM

cn

CM

CM

CM

rH

4—1 1

0)|

o

O

o

o

o

o

O

O

o

o

o

*

CM

M|

00

v£>

00

CO

<r

vO

<f

vO

>J0

\D

<r

1

rH

tH

rH

rH

o

O

O

o

r-i

rH

rH

CM

O

O

o

o

o

o

O

o

CD

O

o

031

4-1 1

at 1

CM

CM

CM

CM

o>

cn

cn

CM

cn

<r

cn

■K

O

o

o

o

cn

CM

cn

CN

o

o

o

CM

O

o

o

o

o

o

o

o

o

o

o

M|

H

m

O

cn

o

<r

CM

-d-

CM

CM

o

o

o

o

CM

CM

m

o

O

o

o

o

o

o

o

O

O

o

a.|

t-t

^^

/_x

^^

^^

^-^

/_N

^

^^

^_^

/_^

^

rH

O0

<r

OO

<r

00

00

<r

<r

00

<r

-3-

CO

CM

r-»

m p-

ON

p-

m

p»

<t p~

m r~-

<r r»

m r~

vD

r»

M0

r^

oo r»

rJ to

14-

CM

<r

p~ <3-

T-H

<r

CM

<r

-H <r

r^ <f

o <r

<r <r

tH

<r

CM

<r

vo -a-

0)

TJ

>

^^'

cn

cn

cm en

cn

cn

CO

CO

p» cn

<r cn

vo cn

-* cn

cn

m

cn

cn

cn cn

0

~

^

s~'

^

^

CM

CO

CM

00

CM

CM

00

00

CM

00

oo

oo

p~

CO

p-

co

CO

r-~

r~

00

r»

r^

z|

<r

<r

<r

Kf

<r

<r

<r

-3-

<r

-<r

<r

rH

rH

rH

iH

rH

rH

.— i

rH

H

r-H

rH

rH

CO

CO

CO

cfl

cfl

cO

w

H

13

-a

U

t3

M

T3

W

H

T3

en

0)

M

<u

CO

Ai

CO

01 ^

CO a

0)

Cfl

OJ

^

0J

cfl

•H

>

C

>

a

fi

CJ

> c

cj a

>

O

>

a

>

CJ

cn

o

CO

o

CO

CO

CO

O CO

CO co

0

Cfl

o

cfl

0

cfl

o>

w

u

u

H

u

j=

-3

<# .

^3

<-3

c3

■-3

^ •

^ .

^3

^)

<J> •

4-1

cn

CO

cn

cfl

cn

CO

cO

Cfl

cn

cfl

CO

0

H

cn

M •

z

cn

Z

ta cn

fn cn

Pn •

to •

P-.

CO

CH

P-4 •

a,

W

CO

W D,

w

CO

<Xt

0*

H co

H co

h a

h a

^

co

»n

Cu

*-i 0.

>.

rH

H

rH •

rH

H

rH

rH rH

.-i rH

iH •

1-t •

rH

rH

rH

rH .

K

o

a

O 00

Q

a

P

00

P o

O O

Q 00

Q OO

Q

u

P

00

P 00

80

Students who reported only a moderate preference for
either the judging or perceptive attitude had significantly
higher overall class rank and overall grade point average
than students reporting higher and lower preferences for
J or P.

Hypothesis 2.2 was only partially supported (see
D2JP, Table 6). No significant relationships emerged for
word-pair/phrase discrepancies with academic variables for
the E-I, S-N, or T-F scales of the MBTI . A significant
relationship existed for the J-P scale. In the post hoc
comparisons, group 4 (students with the largest word-pair/
phrase discrepancy for J or P) came out highest on all four
academic variables. In 3 of 4 post hoc comparisons, group 1
(students with no word-pair/phrase discrepancy on J or P)
were second highest on academic variables. The relation-
ship thus appears curvilinear with students highest and
lowest in word-pair/phrase discrepancies on J-P faring best
on academic variables .

Hypothesis 2.3 was not supported. No significant
relationships emerged between x-y half differences and
academic variables .

Hypothesis 2.4 dealt with total discrepancy scores
(D4EI to D4JP) which combines strength of preference, word-
pair/phrase discrepancies, and x-y half differences for E-I,
S-N, T-F, and J-P, and combined discrepancy scores (Dl to
D4) where Dl combines strength of reported preference for
all four MBTI scales, D2 combines word-pair/phrase

\D *-N

VO >-%

^^

vo m

cn in

on m

u>\a

1

• o

• O

as o

rH •

m *•»

in ^

rH •

u

<r

o-

<r

C

^a

m3

^

CO

cn

cn

cn

cn

cj

■H

CU 4J

cn cn

A

A

A

U-l

•i-l co

CN

CN

CN

■H

3 u

«3

t8

w3

C 1

U 4-1

CU

0)

CU

ai

CU

CU

aj

rH

rH

rH

so c

•H C

c

c

C

e

C

a

c

•H O

CO O

0

o

o

o

o

0

o

Cm

Cm

C-

cn C

(X o

a

c

c

c

c

c

C

00

00

60

r^

r-> *-n

I-l •-s

O /->

<r /-n

rH /"V

m /— >

cn /^

• t-i

o> in

• in

<T rH

cn m

o m

oo m

<r

m

cn h

m m

^ m

/^N

CN O

• O

o o

00 o

\0 o

• o

• o

a\

c

m o

• O

• o

m| a

1

CI •

CN •

rH •

<r •

<r •

o •

as •

CN

cn .

o\ .

m •

as ^-s

^-^

w|

u

a

en

CN

rH

CN

CN

CN

CN

m

CN

en

cn

<t

CO

en -

O (U

4J \CU

a

a.

0.

a.

a

&

Cm

c-

a

a

Cm

•h cn

en u-l

oo

00

00

oo

oo

00

oo

00

00

00

00

14-1 -H

■H 3

co y-i
u cu

A

A

A

r\

A

A

A

A

A

A

/\

C M

4-1 J3

<r

<r

<f

<r

<r

rH

rH

0)

rH

rH

eu

rH

■H

CU

00 -H

C o

c

C

a

■H CO

O C/3

p.

Q.

a.

a

a.

Cm

C»

o

a

Cm

o

Cu

Cm

0

cn a.

a ~— -

oo

00

00

00

oo

00

00

c

00

00

c

00

oo

c

*

en

m

CN

o

a*

~3"

as

m

^

CO

a\

CN

col

cn

CN

cn

CN

CN

rH

CN

rH

CN

r-j

4-1

O

o

O

o

o

o

o

o

O

O

<u|

•K

CN

1-1

m

LO

CN

^D

<f

CO

CN

<r

CN

1

CN

rH

CN

rH

rH

rH

o

o

O

CN

o

o

o

o

o

O

o

o

o

O

col

4-1

<u|

^1-

, — 1

in

CN

m

CN

vD

00

*

rH

rH

rH

rH

rH

o

rH

rH

CN

CN

CN

O

o

o

o

o

o

o

o

o

O

i— 1

m

rH

en

r~

<r

o

o

O

o

O

m

o

CN

rH

m

o

o

O

o

O

o

o

o

o

o

a. I

*

H

^_^

rH

00

00

<r

<t

00

-3-

-*

00

00

<r

f r ,

cn r-~

rH r--

H r-~

o r~

r- r-

ON

r^

o

r~-

vO r~

cn r-»

oo r»

rJ &m|

M — 1 1

OS <t

oo <r

m <r

r-- -<r

o -cr

in

<r

^D

<t

rH ^3-

m <!•

as <t

at

TJ|

>

w

m en

<r cn

m cn

<r cn

<r m

CN

cn

<r

'TS

en cn

cn m

cn cn

0

^

^

^

W

""

CN

CN

00

00

CN

00

CO

CN

CN

00

00

00

r^~

r-

00

r~«

r^-

00

00

r^

z|

<r

<r

-<r

<r

<f

<T

<r

<r

st

s*

H

rH

rH

rH

rH

H

rH

rH

CO

CO

CO

Cfl

rH

rH

cn

M

-a

1-4

T3

T3

u

T3

M

rH

rH

•H

0) ^i

cfl ^

CU

CO

cfl ^

eu

CO

eu ^!

cfl X

CO

cn

> c

cj a

>

CJ

o a

>

CJ

> f3

h a

S-i

0)

0 CO

cfl cfl

o

cfl

cfl cfl

0

CO

O cfl

eu cfl

CU

JZ

M

u

U

u

> M

>

4-1

m3

<^J

l£ .

ca .

•a

<^J

<&

<Ji

O

0 •

o

cn

cn

cfl

cfl

CO

CO

cfl

CO

cn

cfl

a

cn 0h cn

cu cn

Oh •

0m •

<r z id

rZ

Z

fn cn

m3 en

<-3 .

>>

• >-) CO

n co

|-3 a

n Cm

• CO CO

w

P-

m

a

H cfl

cfl

a

33

CN N H

CN rH

CN •

CN •

CN vT H

<r

<t

<T rH

H rH

rH •

PB Q O

o u

Q 00

Q 00

33 Q O

Q

00

c

00

a o

Q cj

Q 00

82
discrepancies for all four MBTI scales, D3 combines
split-half differences on all four MBTI scales, and D4 com-
bines total points computed for strength of preference,
word-pair/phrase discrepancies, and split-half differences
on all four MBTI scales. Significant relationships, in the
direction predicted, did exist for the total Discrepancy-
Index for Sensing or Intuition (D4SN) and academic class
rank, overall grade point average, and academic grade point
average. Students with strongly reported preference, low
word-pair/phrase discrepancy, and low x-y half differences
for S or N were higher on these academic variables than
students reporting lower preferences, higher word-pair/
phrase discrepancy, and higher x-y half differences for S or
N. A single relationship existed for the total Discrepancy
Index for T or F (D4TF) where students with high preference,
low word-pair/phrase discrepancy, and low x-y half differ-
ence for Thinking or Feeling had significantly higher over-
all class rank than students with low preference, high
word-pair/phrase discrepancy, and high x-y half difference
for T or F. A significant relationship did not emerge
between the total Discrepancy Index for J or P (D4JP) and
the academic variables.

A significant relationship existed for the combined
Discrepancy Index for strength of preference on all four
MBTI scales (Dl) and academic variables. Students with
highest reported preference for E or I, S or N, T or F,
and J or P had significantly higher overall class rank and

83
overall grade point average than students with only moderate
or slightly reported preferences. No significant relation-
ships emerged for D2 , D3, or D4 and academic variables.

3
H was not supported.

Hypothesis 4.1 received partial support. For students
with either Sensing or Intuition as their theorized dominant
function, those with strongly reported preference for S or N
(group 1 on D1SN) had higher academic class rank and higher
academic grade point averages than those reporting lower
preference for S or N. Neither word-pair/phrase discrepancy
(H ' ) nor x-y half difference (H ) on S or N for those
students with S or N as their theorized dominant function,
was significantly related to academic variables.

Hypothesis 4.4 was supported. The total index for S or
N (D4SN, where S or N was the theorized dominant function)
did significantly predict to the academic variables in the
direction expected. Students with S or N dominant who had
strongly reported preference, low word-pair/phrase discrep-
ancy, and low x-y half differences on S or N had signifi-
cantly higher overall class rank, academic class rank,
overall grade point average, and academic grade point
average than students with lower preference, higher
word-pair/phrase discrepancy, and higher x-y half differences
on S or N.

84

W| D-|

4-1

c

cfl CO

CJ CD 4-1

•h cn en

14-4 -H CO

C 1 & 4-1

CU

1 CU

01

cu

oj

0)

so b -h c

C

1 c

c

c

C3

a

•H 0 CO O

o

1 o

o

0

c

0

« C ft o

c

1 c

c

c

c

c

,-v

.-1

• in

• m

• T-l

• rH

<* m

i-H i-l

vO M

~tf O

co o

CO O

m o

m o

r~> O

m o

CN O

CM O

w| 0- |

o •

O •

CN •

CN •

CO .

<r •

m •

vD •

s — '

>-N

OT|

4-1

C CO

CM

CO

CN

CO

CN

CO

CN

CO

co co -

o cu 4-i \aj

&

a

Q.

P.

a

p.

a

a.

•h cn cn u-i

00

oo

oo

oo

00

oo

00

00

■4-4 -H CO 14-1

•H 2 S-i CU

A

A

A

A

A

A

A

A

cu

CU

1— 1

r-4

i— 1

tH

H

i-l

rH

i-H

on -H c o

c

c

•H CO o w

0

o

a-

a.

D.

a.

O.

a.

0.

a

cn a. o ^

c

c

oo

oo

00

oo

00

oo

oo

oo

-K

CN

CN

CN

00

00

CO

t-H

CO

-a-

<r

<r

o

in

CO

4—1 1

CU |

o

o

o

o

o

O

*

CM

H

r^

Ml

O

00

00

r~-

ON

00

1

<T

cn

cn

m

<J-

^C

CM

col
4-1 1

O

o

O

o

o

o

•

01 1

CN

<■

a\

H

CJ>

CO

O

o

o

.-H

o

^H

CM
M

O

o

o

O

o

O

CM

CO

O

CN

CN

1-1

O

o

O

0.1

O

o

o

O

o

O

tH

^.

^

^

^^

^^

^

H

c^

M3

a>

<J\

vO

vO

CO ^-,

vO co

cn co

r-\ cn

co

CO

m co

M b| u-il

<r cn

<r cm

O CN

00 CN

00 CN

CT\ CM

cu -a

> >-'

co cn

cn cn

-3- cn

m co

<f CO

vO CO

o

^

^^

s —

%*"'

cn

o

<m

CO

o

o

21

<t i

<r

<t

<r

<r

-3-

CN

CM

CN

Csl

CN

CN

iH

.-1

z

H

cfl

H

CO

i-i

T3

t3

u

T3

U

"O

CO O *-s

CO ^Jl

cfl

CU ^

cfl M

CU

cfl

•H 4-1

O C3|

o

> c

CJ C3

>

o

cn cn e

cfl cfll

cfl

0 cfl

cfl cfl

o

cfl

a) co

S-l 1

u

U

-C 0) C

"3 1

^ .

^

^a

^ •

ca •

4-1 M -H

col

CO

CO

cn

ro

CO

0 0) E

H Z Wl

z •

<r z co

Z cn

z •

Z •

ft£ O

cn cfl 1

cn a.

cn cfl

C/l Cfl

cn &,

CO D.

^ 3T)

-3- t-1 iH 1

iH .

<f -J H

^r ^-i

<r •

■<t •

33 '—'

33 QUI

Q 60

33 Q CJ

Q cj

O 00

Q 00

CHAPTER V
DISCUSSION AND CONCLUSIONS

In the first part of this chapter the results presented
in Chapter IV (Tables 4 through 7) are discussed and sum-
marized in relation to the hypotheses tested. Next, impli-
cations for education are discussed. Finally, suggestions
for further research are given.

Discussion of Results

Hypothesis 1, that students with low Discrepancy Index
scores will employ higher levels of moral reasoning than
students with high Discrepancy Index scores, received
support. The Discrepancy Index dealing with strength of
reported preference for either Thinking or Feeling (D1TF)
did significantly discriminate between students expressing
high and low levels of moral reasoning on the MJS . As
predicted, students who expressed the strongest preferences
for T or F employed higher levels of moral reasoning than
students expressing only moderate or slight preferences for
T or F. In terms of Jungian psychological type theory this
suggests then that students who have developed their judging
function, regardless of whether they prefer to make thinking
judgments or feeling judgments, tend to make decisions which
can be categorized as employing higher levels of reasoning
than students who have a lesser-developed judging function.

85

86
The Discrepancy Indices dealing with word-pair/phrase
discrepancy (D2TF) and split-half differences (D3TF) did not
significantly predict the level of moral reasoning employed.
For D2TF, students in group 1 (those who indicated the same
type preference on the word-pair items as on the phrase
questions) did have the highest total and mean MJS scores;
however, the results were not statistically significant. A
problem with the word-pair/phrase index is that, as it was
constructed in the present study, the number of students
falling in each of the four groups was highly dispropor-
tionate. For example, for D2TF and MJS score, 84 of the 115
students completing the MJS did not have a type discrepancy
on word-pair/phrases. Only 4 students had discrepancies of
from 1 to 6 points, 8 students had discrepancies of from 7
to 12 points, and 19 students had discrepancies of 13 points
or greater. This index needs refinement and re-evaluation
before it can be concluded that word-pair/phrase discrepancy
per se does or does not discriminate between students at
differing levels of psychological type development.

For D3TF, dealing with split-half differences, dis-
proportionate numbers of students across the four groups did
not appear to be a problem. As constructed here, no
significant relationships emerged for split-half difference.

A significant relationship emerged for D4JP (total
index for reported strength of preference, word-pair/phrase
discrepancy, and split-half difference for the J-P scale of
the MBTI) and MJS score. The relationship appears

87
non-linear, with students expressing only moderate
preference for J or P, with slight word-pair/phrase discrep-
ancy, and moderate split-half difference having highest
scores on the MJS. It appears that students expressing very
high or very low preference for J or P may not fare as well
as those expressing more moderate preference for the judging
or perceptive attitude. Personality factors such as ten-
dency to overcontrol or failure to control, to over-organize
or under-organize one's affairs may be involved as well as
factors such as dogmatism. A person with an extreme judging
attitude may not be open to change or grant consideration to
incoming perceptions. Myers has commented that they may
"slam the door" on perception too soon. In opposite fashion,
the person with an extreme perceptive attitude may fail to
make important decisions or meet deadlines by wanting to
keep on perceiving and taking in new information indefinitely,
Myers likened this to "refusing to shut the door" and hesi-
tating to act on what is already known. The data here do
indicate that a balanced or moderate degree of development
or usage of either the judging or perceptive attitude may
prove more adaptive.

Hypothesis 2, that students with low Discrepancy Index
scores will have 1) higher overall and academic class rank
and 2) higher overall and academic grade point average than
students with high Discrepancy Index scores, was supported.
Results were in the direction predicted for reported
strength of preference for E-I, S-N, and T-F. In terms of

Jungian theory this suggests that students who have a well
developed orientation (either Extraversion or Introversion) ,
a well developed perceptive function (either Sensing or
Intuition) , and a well developed judging function (either
Thinking or Feeling) are more likely to be higher achievers
in the traditional school environment than students with
lesser developed functions and orientation.

A significant relationship emerged for strength of
preference for J or P (D1JP) and academic variables. Stu-
dents with moderate preference for J or P had higher overall
class rank and grade point average than students with
highest and lowest preference for J or P.

As with the previous hypothesis, word-pair/phrase dis-
crepancy predicted very little. The only significant rela-
tionship between word-pair/phrase discrepancy and academic
variables was for the J-P scale of the MBTI (D2JP) , with the
relationship non- linear as before. The students with the
highest and lowest word-pair/phrase discrepancies on J-P
fared best academically. Further research regarding
personality factors associated with extreme scores on J-P
may shed more light on this finding.

No significant relationships between split-half
differences and the academic variables emerged. As
constructed here, the split-half index did not predict
academic standings.

Significant relationships in the direction predicted
existed for the total Discrepancy Indices for the

89
Sensing/ Intuition scale of the MBIT and for the Thinking/
Feeling scale (D4SN and D4TF) . As predicted, students with
strongly reported preference, low word-pair/phrase dis-
crepancy, and low split-half difference fared better on the
academic variables than students reporting lower preference,
higher word-pair/phrase discrepancy, and higher split-half
difference for Sensing or Intuition and for Thinking or
Feeling. That these significant relationships emerged is
important in terms of Jungian psychological type theory.
They are believed related to the developmental framework
upon which type theory rests. This development involves the
gradual growth and continued strengthening and evolving of
the four psychological functions (Sensing, Intuition,
Thinking, and Feeling). In terms of the theory, it has been
hypothesized that persons who have a well-developed percep-
tive function with which to perceive (either Sensing or
Intuition) would fare better in real life situations and
particularly in academic settings than someone with a lesser
developed perceptive function. Likewise, a person with a
mature, well- developed judging function (either Thinking or
Feeling) would be expected to make more effective decisions,
a factor that might contribute to academic success. The
achievement data obtained here clearly lend credence to this
aspect of the theory.

Significant relationships existed for the combined
Discrepancy Index for strength of preference on all four
MBTI scales (Dl) and the academic variables in the direction

90
expected. Students expressing the highest reported
preferences for E or I, S or N, T or F, and J or P had
significantly higher overall class rank and grade point
average than students expressing only moderate or slight
preferences. Since Dl is a combined index for D1EI, D1SN,
D1TF, and D1JP it adds little new information. (The degree
of relationship between the variables would probably be
higher if D1JP were removed, since unlike D1EI, D1SN, and
D1TF, it appears to be a non-linear relationship.) This
significant finding for Dl does, however, add validity to
the significant relationships already discussed regarding
the theoretical value of holding clear, consistent prefer-
ences. If the significant relationships for Dl had not
emerged, the foregoing interpretations would certainly
appear less valid.

Hypothesis 3, that students with low Discrepancy Index
scores on the Thinking/Feeling scale when it represents
their theorized Dominant function, will employ higher levels
of moral reasoning than students with high Discrepancy Index
scores on the Thinking/Feeling scale when it represents
their theorized Dominant function, was not supported. Of
the 115 students who completed the MJS, 57 of them had
Thinking or Feeling as their theorized Dominant function.
For these students none of the Discrepancy Indices signifi-
cantly predicted MJS total or mean scores. Two trends in
the direction predicted were present. For strength of
reported preference (D1TF) where T or F was the theorized

91
Dominant function, all four group means were in the exact
order expected with group 1 (strongest reported preference
for T or F) scoring highest and group 4 (slightly reported
preference for T or F) scoring lowest. For D4TF with T or
F dominant, groups 1 and 2 were in first and second place
for highest MJS scores. The differences, however, did not
reach the .05 level of significance.

Two points deserve discussion here. Since the group
means were in the direction predicted, the hypothesis
warrants further investigation with a larger sample and
with more sensitive criterion variables. An ideal test of
prediction would be to take individual judgments or
decisions in a particular setting that had been judged by
some clearly specified external criteria to be effective or
ineffective to varying degrees and then see if the Discrep-
ancy Index yields accurate predictions. Another test would
be to actually employ Kohlberg's system for assessing levels
of moral reasoning. Both of these would be time-consuming
but warranted.

Secondly, the aspect of Jung's theory dealing with the
Dominant and the Auxiliary function is more complex and more
controversial among Jungian analysts and other clinical
professionals than other aspects of the theory. There is
some difference of opinion among experts as to how to best
determine the Dominant and Auxiliary functions for a given
individual. The system employed by Myers was utilized in
the present study.

92
Hypothesis 4, that students with low Discrepancy Index
scores on the Sensing/Intuition scale when it represents
their theorized Dominant function will have higher overall
and academic class rank and higher overall and academic
grade point average than students with high Discrepancy
Index scores on the Sensing/ Intuition scale when it repre-
sents their theorized Dominant function, received partial
support. Significant relationships in the direction pre-
dicted emerged between strength of reported preference for S
or N (D1SN) and academic class rank and grade point average
(where academic class rank was known for 243 students hold-
ing S of N as their theorized Dominant function and academic
grade point average was known for 240 students with S or N
as their theorized Dominant function) .

The total Discrepancy Index for S or N (D4SN, for
subjects with S or N as their theorized Dominant function)
significantly predicted to all four academic variables.
Students with S or N as their Dominant function who had
strongly reported preference, low word-pair/phrase
discrepancy, and low split-half difference for S or N had
higher overall and academic class rank and higher overall
and academic grade point average than students expressing
lower preferences, higher word-pair/phrase discrepancy, and
higher split-half difference on S or N.

These findings for hypothesis 4 tend to support those
already obtained for hypothesis 2 with respect to the
apparent value of holding strong, clear preferences for

93
Sensing or Intuition (the perceptive functions) in relation
to performance on academic variables. They take the analy-
sis one step further, however, to look at differences
between students for whom Sensing or Intuition should
theoretically be their Dominant or most developed function.
The Discrepancy Index did discriminate at that level as
well .

Summary
Strong support for the hypotheses regarding the value
of holding strong type preferences was found (D1EI, D1SN,
and D1TF) . The total Discrepancy Index scores combining
strength of preference, word-pair/phrase discrepancy, and
split-half difference (D4EI to D4TF) and the combined index
for strength of preference on all four MBTI scales (Dl)
also had predictive validity. These relationships appear to
reflect actual "within type" differences. In other words,
it was shown that students with high preference for their
perceptive function, whether it be for Sensing or for
Intuition, fared significantly better on academic variables
than students expressing low preferences. The same rela-
tionship existed for level of moral reasoning and strength
of preference for their judging function. Students with
strong preferences for either Thinking or Feeling attained
higher scores on the MJS than students with low preferences
for Thinking or Feeling. Several non- linear relationships
emerged for the Discrepancy Indices when applied to the
J-P scale of the MBTI.

94
Implications for Education

The purpose of the present study was to explore the
validity of an index developed to assess "within type"
differences. The DI was created in the hope that it could
differentiate between students high in type development and
those less well developed. It was believed, since so many
students are taking the MBTI for numerous reasons, that
information regarding strength of preference, word-pair/
phrase discrepancy, and split-half difference could be used
in a diagnostic and possibly preventive way to identify
potentially "high risk" learners. Based on the present
study, it appears that strength of preference (D1EI, D1SN,
and D1TF) , the total indices (D4EI, D4SN, and D4TF) , and the
combined index (Dl) may give clues as to level of type
development .

The information provided by the MBTI, or simply knowl-
edge of type concepts, can be utilized in many ways in the
classroom to the benefit of students, teachers and adminis-
trators without making reference at all to the indices
investigated here or without reference to the concept of
type development. However, since strong preference appears
to be positively correlated with academic achievement and
level of moral reasoning, this is information that guidance
counselors, school psychologists, evaluators, and teachers
can use in helping students develop their powers of per-
ception and judgment, making real the hope expressed so well
by Myers--to head off problems before they progress too far.

9^
As stated in an earlier chapter, the MBTI is being
increasingly utilized in all types of educational settings.
Guttinger (Guttinger, 1974; Guttinger & Hines 1977) has used
it in an individualized reading program that has been adopted
in numerous school settings across Florida, and on a trial
basis in Brazil. Rowe (1978) utilized psychological type
concepts in her work with science teachers as a means for
creating awareness that there are aspects of science that
appeal to all types of people. Knowledge of psychological
type can also suggest to teachers the best individual way to
teach science to a particular student, or how to at least
interest that student in science.

The important point to bear in mind here is that in the
current uses of psychological type in educational settings
we have been dealing with type differences primarily between
students. The area where we have needed more knowledge is
the difference between students holding the same preferences
at differing levels of development. This is the question of
type development. As Jung proposed it, psychological type
is not a static state, but rather a developmental process.
Anecdotal evidence regarding the developmental nature of
psychological type has been presented, but data obtained
from empirical study were not found in the literature. As
the MBTI and Jung's concepts become even more widely used,
the need to explain "within type" differences has become
increasingly important. Many who employ Jung's develop-
mental concepts in applied settings have been finding clues

96
in the scores yielded by detailed scoring of the MBTI
as to the level of type development of the respondent.
Attempting to ascertain the validity of these clues was the
topic of the present study. Dealing with type concepts on
two levels (basic preferences as well as degree of
development) appears justified and worthy of further study.

Lawrence (1979) and others have pointed out advantages
of employing psychological type concepts in the teaching-
learning process. In the area of learning styles, psycho-
logical type concepts are being taken into account. Even
more studies are needed to increase our understanding of how
different type concepts would provide experiences and activi-
ties that would encompass the smorgaasbord approach where
there was "something for everybody" and where none of the
type preferences were underrated or made to seem less
important. A second dimension of such a teaching approach
would be in providing instruction which would help students
develop their preferred way of perceiving and judging more
fully.

Metts (1979) addressed the problem of interpreting
discrepant and inconsistent responding on the MBTI in a
learning disabled population. Only eight of 113 subjects
came out with the same type formula when x-y halves and
word-pair /phrases were scored separately. He further
pointed out the following advantages of using type concepts

97
with learning disabled students: dealing with the basic
type preferences addresses a need of learning disabled stu-
dents to gain self-knowledge; secondly, the approach is one
where differences are valued rather than discouraged; such
an approach may enhance social maturity when the L.D.
student comes to understand ways in which he is like his
peers and understands and accepts the ways he is different;
and finally, in further exploring personality variables in
L.D. students we may discover better ways of teaching them
(a goal which applies as well to learners who are experi-
encing no apparent learning disability) .

Another place in special education where being able to
ascertain the level of type development of the student would
be beneficial is in the writing of the individualized edu-
cational plans required by Public Law 94-142. Again there
are two dimensions regarding psychological type--one, what
are the basic preferences. When we have an idea of the
student's basic preferences we have clues as to what might
be interesting, motivating, and rewarding to that student.
Secondly and in relation to type development, we must
consider whether the student has had experiences and
opportunities to develop those preferences. If not, we can
begin to provide them.

Those who evaluate children must provide information
regarding the child's strengths and weaknesses. For some
children we end up with a long list of weaknesses and very
few strengths. Knowledge of basic preferences allows us to

98
add to the list of strengths. For example, if the child
prefers extraversion then he is likely to enjoy group
activities and likely to be motivated and rewarded by social
praise (attention and approval of others) . If the child
prefers introversion, then he is likely to appreciate the
opportunity to look deeply into things that interest him.

Just as the psychological type preferences themselves
are basic strengths (gifts, as Myers preferred to call
them) , not having developed them to their full potential is
a weakness. As an example, for the child who prefers to
make feeling judgments, being unclear or unaware of his own
values is a real handicap because the "scales" he uses to
weigh the pros and cons of certain actions in human terms,
are empty.

Refining the measurement of the level of development
of a person's psychological type preferences is important.
When it has been accomplished through reliable and valid
means we will hold one more key to better understanding and
better educating our students.

Suggestions for Further Research

Further research is needed before the Discrepancy Index
as constructed and evaluated here, can be relied upon as
providing anything more than suggestive evidence that level
of type development is being validly assessed by it.
Findings for reported strength of preference for E-I, S-N,
and T-F (D1EI, D1SN, and D1TF) and the total and combined
indices (D4EI, D4SN, D4TF, and Dl) are supportive of data

99
earlier presented by Myers indicating that persons with
strong and clear preferences fare better academically than
persons with low preferences. Evidence was also found that
clear and consistent preference for the judging function
(whether Thinking or Feeling) was related to employing
higher levels of moral reasoning.

Discrepancies on the word-pair/phrase items and
differences on the split-halves need closer attention and
further evaluation. Word-pair/phrase discrepancy, as calcu-
lated in the present study, predicted very little. The
split-half difference, as calculated here, did not signifi-
cantly predict either Moral Judgment Scale scores or aca-
demic standings. Other methods of index construction need
to be investigated. It is possible that simply adjusting
the range of difference for each of the four categories or
groups for these two types of discrepant responding would
result in better discrimination. With the word-pair/phrase
discrepancy index, in particular, this is advised. A
disproportionate number of cases fell in group 1 because of
the large proportion of subjects who did not change type
(i.e., they consistently reported themselves in the same
type category on the word-pair items as on the phrase
items) .

The discrepancy indices, when applied to the J-P scale
of the MBTI, showed several interesting relationships which
appear non-linear. For strength of preference for J or P
(D1JP) and academic variables, students with moderate

100
preferences fared better than students at the extremes
(those reporting highest and lowest preferences) . With
respect to Moral Judgment Scale scores and the total dis-
crepancy index for J-P (D4JP, combining strength of pref-
erence, word-pair/phrase discrepancy, and split-half
differences for J or P) , the relationship was again non-
linear, with students at the extremes scoring lowest on the
MJS and students in the middle groups scoring highest. This
trend was reversed in the relationship found between
word-pair/phrase discrepancy on J or P (D2JP, see Table 6)
and the academic variables. The students with the highest
word-pair/phrase discrepancy for J or P (group 4) were
consistently higher on the four academic variables than the
other three groups. However, for three of the four academic
variables the students with no word-pair/phrase discrepancy
(group 1) were in second place. Students in the middle
(with moderate discrepancies) did not fare as well on the
academic variables as those at the extremes. Further study
is needed to determine how extreme scores on the J-P scale
of the MBTI are correlated with variables known to reflect
varying levels of psychological and personal adjustment.

The findings of this study suggest the need for
research in several areas. The utilization of more
sensitive academic variables than grade point average and
more extensive assessment of moral reasoning and decision-
making variables is suggested.

101
In conclusion, the Discrepancy Index does appear to
assess "within type" differences. These differences warrant
further investigation because of their implications for edu-
cation and for personal adjustment, in general.

APPENDICES

APPENDIX A

SUPPLEMENTARY ANALYSES OF TEACHER ESTIMATES

OF READING LEVEL

The supplementary hypothesis that students with a well
developed (i.e., strong, clear and consistent) perceptive
function would have higher reading levels than students with
a lesser developed perceptive function was supported. Stu-
dents with high preference for either Sensing or Intuition
(D1SN) had significantly higher estimated reading levels
than students with low preference for S or N. In addition,
strength of preference for the judging function (either
Thinking or Feeling) was also positively related to higher
estimate of reading level (see D1TF on Table Al) . For the
total index (D4SN) students with high preference, low word-
pair/phrase discrepancy, and low x-y difference had signifi-
cantly higher estimated reading level than students with
lower preference, higher word-pair/phrase discrepancy, and
higher x-y difference. The combined indices (Dl and D4)
were also positively related to estimates of reading level.
Strength of preference (D1SN) and the total index (D4SN) for
students with Sensing or Intuition as their theorized domi-
nant function showed the same significant relationship.

103

104

<■ *-%

1 <r *->

i t->

cn in

1 H i-H

i o Ln

CO O

1 CT\ O

i cn o

co| o-l

rH w

u

<r

1 <r

<r

c

<■«

1 <-3

1 <&

co en

cn

1 co

cn

U 0» 4-1

•h en en

a

A

A

UH -H CO

CN

1 CM

CN

•h an

c3

1 <-3

i f-a

C 1 6 4J

iH

1 i-l

at

CD

<U

rH

01

oo c «h a

1 C

1 C

1 B

e

•H 0 CO O

a

P.

1 O

1 O

1 0

P.

0

M C P. 'J

00

oo

C

c

c

00

c

/-V

^

^

,_v

,— . /--N

rH rH

cr\ m

cn m

on m

oo m m m

oo m

m o

CN o

o o

o o

cn o cn o

cn o

co| p.|

rH ^

rH ^-'

!-| ^

rH "-^

CN •-' CN ^

rH "^

co|

4J

C en

co

<r

CO

cn

cn ~s-

CO

CO C0 -

u eg 4J mu

a

a

p.

p.

P. P.

P.

•h en en y-i

00

oo

00

oo

oo oo

00

in ti an

•H 3 H 0)

A

A

A

A

A A

A

fl M 4J JZ

eu

rH

rH

T— 1

rH

01

rH iH

rH

00 -H C CJ

c

a

•H CO 0 CO

0

a.

P.

p.

a.

0

P. P.

P.

81 D. O^

c

oo

oo

oo

oo

c

oo oo

00

•X

CN

rH

ON

CN

ON

r~»

<•

CO

col

<•

m

-3-

-d-

cn

o^

oo

4-1

o

o

O

o

O

o

o

cu|

•K

CN

H

CN

rH

cn

rH

00

CN

CN

1

CN

CN

CN

CN

o

m

m

CN

o

o

o

o

o

o

o

cOl

4-1 1

at |

a>

00

a-.

00

o>

CN

rH

*

rH

CO

h

CN

CN

<r

co

CN

O

o

o

o

o

o

O

-*

CO

o

cn

rH

<r

rH

CN

a|

o

o

O

o

o

o

o

i— i

s~.

^

^

^

/_s

^^

^^

H

vP

\D

V40

vO

^

CN

CN

CO /^

~* rH

O rH

00 H

rH rH

<T rH

00 rH

00 rH

U Cn 1 u-il

O CN

m cn

rH CN

r-~ cn

r-- cn

00 rH

CO rH

cu -a

> '-'

cn cn

^r cn

m cn

cn cn

cn cn

CO CO

CO CO

0

^^

s

v*'

^~^

**^

y~'

w

O

o

o

o

o

vP

vp

CN

CN

CN 1

CN

CN

rH

.H

Z|

CN

CN

CN

CN

CN

.H

rH

en

01

1)

•H

M

u

en

11 4-J

0) 4-1

a)

-X3 C

J= c

£

3 2 CO

3 Z CO

4-1

C 1

a

o

Z

P*

Z 1

Z w -H

Z M -H

P.

e/3 1

H

CO 1

CO 0 E 1

co o e

>>

H 1

H

-J- 1

<r

rH 0 1

<r o

K

Q 1

a

Q

Q

a

Q CO T3

Q CO T3

APPENDIX B
DESCRIPTIVE STATISTICS
FOR DISCREPANCY INDICES

The following five tables provide the mean for each
index, the N, and the percent of sample for each of the
four categories or groups within each index. Metts (1979)
raised the question as to how many discrepancies can be
expected to exist in normal samples. If other researchers
will report these data, a basis for comparison (with differ-
ent types of well-described samples) can be established.

105

Table Bl. Descriptive statistics for DI's reflecting
strength of preference.

106

Discrepancy
Index

X

Breakdown for four groups
within Discrepancy Index

N %

D1EI

2.73

group 1

(strongly reported
preference)

48 9.9

group 2

(clearly reported
preference)

156 32.2

group 3

(moderately reported
preference)

161 33.2

group 4

(slightly reported
preference)

120 24.7

D1SN

2.87

group 1

25 5.2

group 2

152 31.3

group 3

167 34.4

group 4

141 29.1

D1TF

2.84

group 1

41 8.5

group 2

134 27.6

group 3

173 35.7

group 4

137 28.2

D1JP

2.73

group 1

52 10.7

group 2

149 30.7

group 3

160 33.0

group 4

124 25.6

Table B2 . Descriptive statistics for DI ' s reflecting
word-pair/phrase discrepancies.

107

Discrepancy-

X

Breakdown for

four groups

Index

within Discrepancy Index

N %

D2EI

1.82

group 1

(no discrepancy

S did not change

type)

326 67.2

group 2

18 3.7

(cont. score disc.

of 1 to 6 pts.)

group 3

41 8.5

(cont. score disc.

of 7 to 12 pts.)

group 4

100 20.6

(cont. score disc.

of 13 pts. or

greater)

D2SN

1.85

group 1

314 64.7

group 2

23 4.7

group 3

53 10.9

group 4

95 19.6

D2TF

1.79

group 1

328 67.6

group 2

21 4.3

group 3

45 9.3

group 4

91 18.8

D2JP

1.95

group 1

304 62.7

group 2

24 4.9

group 3

35 7.2

group 4

122 25.2

Table B3. Descriptive statistics for DI ' s reflecting
split-half differences.

108

Discrepancy
Index

X

Breakdown for four groups
within Discrepancy Index

N %

D3EI

2.01

group 1 194 40.0
(x-y difference
from 0 to 5 pts . )

group 2 147 30.3
(x-y difference
from 6 to 11 pts.)

group 3 91 18.8
(x-y difference
from 12 to 17 pts.)

group 4 53 10.9
(x-y difference

of 18 pts. or

greater)

D3SN

2.08

group 1 160 33.0
group 2 179 36.9
group 3 92 19.0
group 4 54 11.1

D3TF

1.97

group 1 183 37.7
group 2 171 35.3
group 3 94 19.4
group 4 37 7.6

D3JP

2.02

group 1 183 37.7
group 2 157 32.4
group 3 98 20.2
group 4 47 9.7

Table B4. Descriptive statistics for DI's reflecting
total discrepancies for each MBTI scale.

109

Discrepancy
Index

X

Breakdown for four groups
within Discrepancy Index

N

%

D4EI

(DIEI + D2EI
+ D3EI)

2.37

group 1

(combines point
totals of 3 6c 4)

104

21.4

group 2

(combines point
totals of 5 & 6)

155

32.0

group 3

(combines point
totals of 7, 8,
& 9)

169

34.8

group 4
(combines point

totals of 10, 11,

& 12)

57

11.8

D4SN

(DISN + D2SN
+ D3SN)

2.48

group 1
group 2
group 3

70
174
177

14.4
35.9
36.5

group 4

64

13.2

D4TF

2.40

group 1

82

16.9

(D1TF + D2TF
+ D3TF)

group 2
group 3

177
174

36.5
35.9

group 4

52

10.7

D4JP

(DIJP + D2JP
+ D3JP)

2.43

group 1
group 2
group 3

93
156
172

19.2
32.2
35.5

group 4

64

13.2

110

m r-»

a

00

<r

vO

CO

r~»

rH O

00

CN

oo r--

oo

00

O ro

<r

!-H

iH

m

<r

CO

CN O

^-1

vD

cn in

CO

rH

**\

H <r

rn

^H

CN

<r

CN

cn <r

cn

r-i cn

CO

•-t

X

01

T3

C

H

rH CN

in

r-^

<r

^H

CO

CN

r^ <r

<-

O

r^. cn

CO

r-~

ss|

LH i-l

<D

in

o

CN

H

-3-

O c^

m

cn

^o r^

CO

m

o

C

CN

H

H

CN

H

H H

H

r-H

rH

CO

Q.

0)

*X

i£>

vD

rH VD

CN

00

l-l

/->> ^H

H

rH

rH

^H

.—i

CN CN

CO

<r

CJ

CO ^H

oil

00

H

vO

-H

en

X

ot|

X

X

X X

X

X

•H

■C £

co

00

X

X

oo

X

X

oo

OO 00

00

00

O

oo so

r^

3

i

00

00

3

00 /— -

OO

3

3 3

3

3

3 3

O

3

3

0

3 00

3

0

o o

O

O

c

O 0

TJ

M

<r

O

0

l-i

o

0

1-1

U U

SH

rl

•H

in )-l

C

X!

l-i

1-1

X

1-4 T3

l-i

X

X X

X

X

*

X x

cfl

4-1

4-4 ^

X

X

4-1

X C

X

4-1

4-1 4-1

4-1

4-1

4-1

4-1 4-1

o en

4-4

4-1

4-1 CO

4-1

•H

CN

<r

ai

CN

CN

CN CN

r~~

cn

3

-a- a\

H

H

en en

H CO

m

o>

H

<r r^

cr>

<-i

.H CN

CN

cn

en

U-. UH

uj

M-i

CO 1-4

U-l

UH

uh

U-l U-l

u-i

U-l

U-l U-l

UH

UH

a.

o o

0

O

■U X

o

O

0

0 O

O

o

0 O

O

0

3

o a

O

tn en

tn

CO

4-1 ■>«.

en

en

en

en en

cn

cn

en en

en

en

M

H H

H

rH

1-1

.-i

rH

H

.-1 H

j-i

T-l

rH r-l

rH

rH

oo

CO CO

CO

CO

u a.

cd

CO

CO

n) CO

CO

CO

CO cfl

CO

Cfl

4-1 4J

4-1

4-1

C 1

4-1

4-1

4-1

■U 4-)

4J

4-1

4J 4-1

4J

4-1

u

O 0

o

0

■H T3

0

O

o

o o

0

0

O 0

o

O

3

4-1 4-1

4-J

4-1

O 3

4-1

4-1

4-1

4-1 4-1

4-1

4-1

4-1 4-1

4-1

4-1

o

P.

U-l

4-1 4-1

4-1

4-1

c

4-1

4-1

4-1

4-1 4-1

4-1

4-1

4-1 4-1

4-1

4-1

a c

C

a

>, o

C

c

c

c c

C

C

c c

c

c

U

•H iH

•H

•H

T— 1

■H

•H

•H

•H 1-4

■H

i-l

•H T-t

•H

•H

0

o o

O

O

c •

O

0

O

0 o

O

o

O O

O

O

U-l

a. a

a

cu

O U-l
U-l

P4

Cu

O.

ex a.

a

D.

a. cu

Cu

Cu

C

en en

Bl

tn

tn i-i

01

tn

en

en en

tn

cn

cn cn

en

en

3

01 0)

01

ai

aj -a

QJ

0)

01

01 CU

0J

cu

cu cu

cu

a)

o

c c

c

a

-a

C

a

c

(3 C

c

a

C C

C

C

-a

■H -H

•H

•H

3 a)

•H

■H

•H

•H -H

■H

•H

•H i-l

•H

•H

j*s

-a x

X

X

h a.

X

X

XI

X X

X

X

X X

X

X

cfl

B B

e

E

o >>

e

s

B

e e

e

s

e e

e

e

01

o o

o

o

a 4J

O

o

0

o o

0

o

O 0

o

o

u

u cj

u

O

i-t

CJ

CJ

o

CJ CJ

CJ

CJ

CJ CJ

a

CJ

«

c

^H CN

CO

<r

X

CN

co

<r

.-1 CN

cn

<r

<-* CN

co

-*

a a

a.

a

a- 4-i

a.

a

a

D. a

Cl

D.

D- CX

a

Cu

3 3

3

3

3 i-l

3

3

3

3 3

3

3

3 3

3

3

O O

0

O

O 3

0

o

o

O 0

0

O

O O

o

O

U U

u

u

1-1

H

i-i

(H

u u

1-1

1-4

l-i l-i

u

Sh

oo oo

00

00

oo

00

to

00

oo oo

OO

00

oo oo

oo

00

r^

o

CN

00

<r

CN

CN

<-

1 X!

CN

CN

CN

CN

^

^

^

PL|

a.

fl4

>-)

>~>

•-3

H

CN

ro

Q

Q

a

+

+

+

U,

to

Pi.

H

H

H

X

— i

CN

n

QJ

Q

Q

Q

-a

e

+

+

+

z^

M

Z

S3

2

cn
Q

^

LO

t/1

e/3

CJ

rH

CN

cn

+

CI

a

Q

Q

cfl

CN

PL,

ai
u

+

+

+

Q

H

H

M

+

u

M

W

W

en

r^

CN

cn

r-l

•H

rH Q

CN Q

<n q

<f Q

Q

a X-'

a *-*

a ^

Q ^^

APPENDIX C
TYPE DISTRIBUTION IN STUDY SAMPLE

The type distribution for subjects in the present study
is presented in Table CI. This type distribution was com-
pared with that of Myers' sample of 9,320 Pennsylvania high
school students (Table C2) . The Selection Ratio Type Table
program, a 2x2x16 Chi- square contingency table format (Allen
& Kainz, 1976), was utilized for this purpose. Results of
this comparison are reported in Table C3.

Several general points must be covered before inter-
preting the data in Table C3. Based on a large number of
samples, McCaulley (1978, pp. 32-38) estimates that in the
general population extraverts outnumber introverts about
3:1; sensing types outnumber intuitive types about 3:1; in
female samples, feeling types outnumber thinking types
about 2:1 to 3:1; in males thinking types probably out-
number feeling types about 3:2. McCaulley further points
out that for high school student samples the majority of

students fall in the extraverted sensing grouping (ES )

with relatively fewer students in the introverted intuitive

grouping (IN ). Further, because S-N and J-P are

positively correlated, we can expect to find the number of
judging types increasing as the number of sensing types

111

112
increase, and conversely the number of perceptive types
increasing as the number of intuitives increases.

It can be seen from Table C3 that the number of extra-
verts and introverts in the present study sample are nearly
equal. This is an unusual finding, and Table C3 shows this
high proportion of introverts in a high school sample
significant at the .001 confidence level.

The number of sensing and intuitive types in the
present sample are in roughly the same proportion (3:1) as
McCaulley expects and as found in Myers' sample. Since SN
and JP are positively correlated as noted above, it is
highly unusual then to find the large and statistically
significant (p = .001) proportion of perceptive types, with
perceptive types outnumbering judging types about 3:1.

In the study sample a highly significant (p = .001)
proportion of feeling types to thinking types exists (about
3:1). Since the proportion of females to males is close to
equal (271 females, 214 males) the proportion of feeling
types to thinking types could be expected to be more nearly
equal as in Myers' sample with 4,387 females and 4,933 males
(see Table C2) .

The study sample is atypical from most high school
student samples in another important way. It was noted
earlier that a large proportion of extraverted sensing (ES)
types, possibly as high as 50 to 707o could be expected with
a smaller proportion of introverted intuitive (IN) types.
Instead, only 297o of the sample are ES types

113
(X2 = 46.44, p = .001), 35% are the IS types (X2 = 34.99,
p = .001), and 15% are the IN types (X2 = 5.08, p - .05).
The EN types were found in the proportion expected when com-
pared with Myers' sample.

114

Myers-Briggs Type Indicator

Table CI. Psychological type distribution
for 485 high school students.

Type Table

SENSING TYPES INTUITIVE TYPES

with THINKING with FEELING with FEELING with THINKING

ISTJ

N=32
°o = 6.60

ISTP

- 37
= 7.63

ESTP

N - 20
'•= 4.12

ESTJ

N- 30

% = 6.19

ISFJ

N = 43
"a -8.87

INFJ

N=14
% =2.89

ISFP

N = 58
°° = 11.96

ESFP

N= 64
°o = 13.20

INFP

N = 45
°° ^9.28

ESFJ

N= 27
% = 5.57

£NFP

N= 57

°== 11.75

FNFJ

N= 23

%= 4.74

/N7J

N=4
°o =0.82

/N7P

N = ll

"• =2.27

ENTP

N=ll
"o = 2.27

ENTJ

N=9
•/„ = 1.86

241

244

311
174

154
331

182
303

74
100
170
141

49.69
50.31

64.12
35.88

31.75
68.25

37.53
62.47

IJ

93

19.18

IP

151

31.13

EP

152

31.34

EJ

89

18.35

ST

119

24.54

SF

192

39.59

NF

139

28.66

NT

35

7.22

SJ

132

27.22

SP

179

36.91

NP

124

25.57

NJ

50

10.31

TJ

75

15.46

TP

79

16.29

FP

224

46.19

FJ

107

22.06

15.26
20.62
35.05

29.07

NOTES:

115

Myers-Bnggs Type Indicator

Table C2. Myers' students in Pennsylvania
high schools (N = 9,320).

Type Table

SENSING TYPES

INTUITIVE TYPES

withTHINKING with FEELING

with FEELING withTHINKINC

N

%

ISTJ

ISFJ

INFJ

INTJ

E

6044

64.85

N = 645

N=636

N=167

N=244

c

1

3276

35.15

°°=6.92

*. =6.82

%=1.79

°o=2.62

o
0

s

6350

68.13

Z

a

N

2970

31.87

z

T

4432

47.55

33

o

<

m

F

J

4888
5091

52.45

ISTP

ISFP

INFP

INTP

54.62

N-388

N = 503

N = 363

N=330

■n

P

4229

45.38

'• = 4.16

°« =5.40

°°=3.89

°o=3.54

n

IJ

1692

18.15

T3

IP

1584

17.00

<

EP
EJ

ST

2645
3399

28.38
36.47

ESTP

ESFP

ENFP

ENTP

SF

3036
3314

32.58
35.56

N = 608
°° - 6.52

N = 873
% = 9.37

N = 708
°o =7.60

N=456
°o =4.89

TJ

33

n

NF
NT

1574
1396

16.89

14.98

TJ

SJ

3978

42.68

<

SP

2372

25.45

X

NP

1857

19.92

33
>
<

m

NJ
TJ

1113
2650

11.94

ESTJ

ESFJ

ENFJ

ENTJ

28.43

N-- 1395
"» = 14.97

N= 1302
°° = 13.97

N=336
0/o = 3.61

N = 366
°o=3.93

on

C

o
0

TP
FP
FJ

1782
2447

2441

19.12
26.26
26.19

z

IN

1104

11.85

o

EN

IS

ES

1866
2172
4178

20.02
23.30
44.83

NOTES:

116

Myers-Briggs Type Indicator

Table C3. Comparison of 485 high school
students with Myers1
Pennsylvania high school
student sample (N = 9,320).

SENSING TYPES INTUITIVE TYPES

with THINKING with FEELING with FEELING with THINKING

Type Table

Legend: % = percent
of total choosing
this group who fall
into this type.
I = ratio of percent
of type in group to
7o in sample.

ISTJ

ISFJ

INFJ

INTJ

E

1

241
244

49.69
50.31

0.77*
1.43*

N = 32

N-43

N=14

N=4

c

°°=6.60

'• =8.87

°°=2.89

°° =0.82

O

s

311

64.12

0.94

I =0.95

I =1.30

I =1.61

I =0.32^

z
O

N

174

35.88

1.13

z

T

154

31.75

0.67*

TO

o
<

F

J

331
182

68.25
37.53

1.30*

ISTP

ISFP

INFP

INTP

0.69*

N = 37

N=58

N=45

N=ll

P

303

62.47

1.38*

°o =7.63
I =1.83*

'. =11.96
I =2.22*

°==9.28

I =2.38*

°o=2.27
I =0.64

TO

n

IJ
IP

93
151

19.18
31.13

1.06
1.83*

<

EP
EJ

ST

152

89

119
192

31.34
18.35

24.54
39.59

1.10
0.50*

0.75*
1.11

ESTP

ESFP

ENFP

ENTP

SF

N = 20
'• "4. 12

N = 64
°o =13.20

N=57
%=11.75

N=ll
°° =2.27

■70

n

NF
NT

139

35

28.66
7.22

1.70*
0.48*

I =0.63"

I =1.41//

I =1.55*

I =0.46//

■o

SJ

132

27.22

0.64*

<

SP

179

36.91

1.45*

NP

124

25.57

1.28#

X)

>
<

m

NJ

TJ

50
75

10.31
15.46

0.86

ESTJ

ESFJ

ENFJ

ENTJ

0.54*

N= 30
"• = 6.19
I =0.41*

N = 27
°° = 5.57
I =0.40*

N = 23
°°=4.74
I =1.32

N=9

°o=1.86
I =0.47"

on

C

o
o

TP
FP
FJ

79
224

107

16.29
46.19
22.06

0.85

1.76*

0.84"

z
O

IN
EN
IS
ES

74
100
170
141

15.26
20.62
35.05
29.07

1.29"
1.03
1.50*
0.65*

NOTES: concerning symbols following the selection ratios:
" implies significance at the .05 level, i.e., chi sq.> 3.8
// implies significance at the .01 level, i.e., chi sq.p> 6.6
* implies significance at the .001 level, i.e., chi sq.> 10.8
- (underscore) indicates Fisher's Exact Probability used instead of chi-square.

APPENDIX D
RESULTS AND DISCUSSION OF SELECTION
RATIO TYPE TABLE (SRTT) ANALYSES

Additional analyses of the four categories or groups of
each discrepancy index were needed to properly interpret the
analysis of variance results presented in Chapter IV. If
significant differences in the basic type preferences of
subjects existed between the four groups for each discrep-
ancy index, the analysis of variance results would be diffi-
cult to interpret.

For example, if for D1SN, category 1 and category 4
significantly discriminated students who did best academi-
cally from those who performed less well, then for D1SN,
category 1 (students who reported a strong preference for
Sensing or Intuition) and category 4 (students who reported
only a slight preference for Sensing or Intuition) must be
looked at closely to make sure category 1 is not predominantly
Intuitive types and category 4 is not predominantly Sensing
types. If this were true the picture would become very
blurred. Two different interpretations would be supported.
The finding might result from the fact that Intuitive types
tend to out-perform Sensing types on academic variables, or
it could be that students with a clear, consistent, and
well-developed preference, either for Sensing or for
Intuition, do better academically than those students who

117

118
have not developed a strong preference to use in perceiving.
In reality, the finding might be explained in part by both
factors and by other factors as well.

The SRTT also helps answer the previously posed ques-
tion as to whether grades get lower as Sensing is more
strongly reported. Three sample SRTT's are provided in
Tables Dl through D3, and their implications discussed.

Table Dl is the selection ratio type table (SRTT)
between group 1 and group 3 for D1EI (strength of preference
for Extraversion or Introversion) , the students in group 1
reporting very strong preferences for E or I and group 3
reporting only moderate preference. This SRTT table is of
interest because group 1 was significantly higher in overall
class rank than group 3. Since in past studies Introverts,
in general, have been found to make higher grades than
Extraverts, it was necessary to make sure that group 1 did
not contain significantly more Introverts than group 3. It
did not. Hence, this lends further credence to the finding
that it is strength of preference for either E or I which
accounts for the difference, not preference for I over E,
per se.

Table D2 is the SRTT between groups 2 and 4 for D1TF
(strength of preference for thinking or feeling) , with group
2 reporting clear preferences for T or F, group 4 very
slight preference. This table is of interest because group
2 had significantly higher Moral Judgment Scale scores than
group 4. In group 2 we find 28 thinking types compared with

119
106 feeling types. Group 4 had 68 thinking types and 69
feeling types. Thus in the low MJS group (group 4) prefer-
ence for Thinking and for Feeling are equally distributed.
However, in the high MJS group (group 2) we find a signifi-
cantly higher proportion of Feeling types than Thinking
types. It appears that the high number of ENFP ' s in group 2
accounts for this difference. There are, nevertheless, 28
students in the high MJS group with very strong preference
for Thinking. This does not change the interpretation that
high preference for either T or F goes with high level of
moral reasoning, but the SRTT does raise the question as to
why the feeling types (particularly ENFP) are so highly
represented in the high MJS group.

A look at the level 5 and 6 responses for the MJS
vignettes may shed some light on this. They appear to be
the type of response which would appeal to feeling judgment.
They call for a subjective weighing of values important to
persons involved in the situation, rather than reasoning
based on clear-cut principles of right or wrong, or legal
contrivance. While not within the scope of the present
investigation, this could be determined in future research
on type preference and the MJS. It is possible that the
very task of completing a moral judgment questionnaire
appeals more to feeling than thinking types.

Table D3 shows the SRTT for groups 1 and 3 on D4SN
(combining strength of preference, word-pair/phrase dis-
crepancy, and x-y half difference for Sensing or Intuition).

120
As predicted, group 1 (high preference with low
word-pair/phrase discrepancy and x-y difference) had sig-
nificantly higher overall and academic grade point average
than group 3 (moderate preference with high word-pair/phrase
discrepancy and x-y difference). We do not find a signifi-
cant difference in the proportion of Sensing to Intuitive
types in these two groups. Thus, this lends further support
to the interpretation that strong, clear and consistent
preference for one's perceptive function (whether it be
Sensing or Intuition) goes with academic achievement. If
there had been a higher than usual proportion of Intuitives
in group 1 this would have made the above interpretation
less credible, because preference for Intuition has been
found to be positively correlated with academic success.
To some extent this offers evidence for the question posed
by Myers (1962) and explored by Weychert (1975) . This was
the question as to whether grades get lower as preference
for Sensing increases. Since we find an equal proportion of
types reporting high preference for Sensing as we do types
reporting high preference for Intuition in the group with
highest academic and overall grade point average, we find no
significant trend for grades being lower when preference for
Sensing is strong.

121

Myers-Briggs Type Indicator

Table Dl. SRTT analysis for D1EI
groups 1 and 3.

SENSING TYPES INTUITIVE TYPES

with THINKING with FEELING with FEELING with THINKING

ISTJ

N = 3
°»=6.25
I =0.72

ISTP

N = 3
%=6.25

I =0.72

ESTP

N-l

'• -2.0
1 = 0.

67

ESTJ

N-0

% -0.0
I =0.0"

ISFJ

%=6.25
I =0.67

N

INFJ

N=l

•■.=2.08

I =1.12

ISFP

N=5

°o =10.42
I =0.76

ESFP

N=5

°°=10.42
I =0.99

INFP

N=4
°o=8.33
I =0.84

ESFJ

N=l

"'.=2.08
I =0.30

ENFP

N=ll

°o =22.92
I =2.46"

ENFJ

N=8

D/o=16.67
I =5.37#

INTJ

N=0
°„ =0.0
I =0.0

INTP

N=l

°o=2.08
I =1.68

ENTP

N=l

°°=2.08
I =0.84

ENTJ

N=l

%=2.08
I =1.68

Type Table

Legend: % = percent
of total choosing
this group who fall
into this type.
I = self selection
index: ratio of per-
cent of type in group
to 7o in sample.

N % I

28
20

21
27

10
38

17
31

7
13
18
10

58.33
41.67

43.75
56.25

20.83
79.17

35.42
64.58

14.58
27.08
37.50
20.83

1.27
0.77

0.62*
1.89*

0.59
1.23

0.86

1.09

0.71
0.81
1.47
1.02

ST

7

14.58

0.49"

SF

14

29.17

0.72

NF

24

50.00

2.06*

NT

3

6.25

1.12

SJ

7

14.58

0.43//

SP

14

29.17

0.81

m
X

NP
NJ

17
10

35.42
20.83

1.54
3.05#

>

TJ

4

8.33

0.42

en

TP

6

12.50

0.81

FP

25

52.08

1.20

FJ

13

27.08

1.28

IN

6

12.50

0.91

EN

21

43.75

2.71*

IS

14

29.17

0.72

bb

7

14.58

0.49"

NOTES: concerning symbols following the selection ratios:
" implies significance at the .05 level, i.e., chi sq. > 3.8
// implies significance at the .01 level, i.e., chi sq.>6.6
* implies significance at the .001 level, i.e., chi sq.>10.8
- (underscore) indicates Fisher's Exact Probability used instead of chi-square.

122

AAyers-Briggs Type Indicator

Table D2 . SRTT analysis for D1TF
groups 2 and 4.

SENSING TYPES INTUITIVE TYPES

withTHINKING with FEELING with FEELING with THINKING

Type Table

Legend: °L = percent
of total choosing
this group who fall
into this type.
I = self selection
index: ratio of per-
cent of type in group
to 7o in sample.

N % I

ISTJ

ISFJ

INFJ

INTJ

E

1

69

65

51.49
48.51

1.22
0.84

N = 8

N = o,

N=5

N=0

c

°°=5.97

"° "6.72

°°=3.73

°° =0.0

o

o

S

77

57.46

0.80"

I =0.48

I =0.92

I =1.28

I =0.0

z
O

N

57

42.54

1.53"

2

T

28

20.90

0.42*

73

o

<

m

F
J

106

43

79.10
32.09

1.57*

ISTP

ISFP

INFP

INTP

0.76

N=7

N = 18

N = 16

N=2

on

P

91

67.91

1.18

°<- = 5.22

"• =13.43

% =11.94

°.=1.49

TO

IJ

22

16.42

0.68

I =0.51

I =1.08

I =1.82

I =0.34

m
-a

IP

43

32.09

0.96

<

EP

48

35.82

1.49"

m

EJ
ST

21

24
53

15.67

17.91
39.55

0.86

0.48*
1.13

ESTP

ESFP

ENFP

ENTP

SF

N=5
'.= 3.73

N = 19
'■• = 14.18

N=24
% = 17.91

N=0
\ =0.0

-o
n

NF
NT

53
4

39.55
2.99

2.58*
0.24#

I =0.57

I =1.39

I = 4.91*

I =0.0

■o

SJ

28

20.90

0.64"

<

SP

49

36.57

0.93

x

NP

42

31.34

1.72"

3D
>
<

m

NJ
TJ

15
14

11.19
10.45

1.18

ESTJ

ESFJ

ENFJ

ENTJ

0.42#

N-- 4
%= 2.99
I = 0.37

N = 7
°o = 5.22
I = 1.02

N= 8
% = 5.97
I =2.73

N=2
°o = 1.49
I = 0.51

C
O

o

TP
FP
FJ

14
77
29

10.45
57.46
21.64

0.42#
1.75*
1.24

z
o

IN
EN
IS
ES

23
34
42
35

17.16
25.37
31.34
26.12

1.12
2.04#
0.74
0.87

NOTES: concerning symbols following the selection ratios:

" implies significance at the .05 level, i.e., chi sq.>3.8

// implies significance at the .01 level, i.e., chi sq. >6.6

* implies significance at the .001 level, i.e., chi sq.> 10.8

- (underscore) indicates Fisher's Exact Probability used instead of chi-square.

123

Myers-Briggs Type Indicator

Table D3 . SRTT analysis for D4SN
groups 1 and 3.

SENSING TYPES

INTUITIVE TYPES

with THINKING with FEELING

with FEELING

withTHINKING

ISTJ

ISFJ

INFJ

INTJ

N=10

N=6

N=l

N=l

°o=14.29

••. -8.57

%=1.43

% =1.43

I =3.16#

I =1.17

I =0.28

I =1.26

iSTP

ISFP

INFP

INTP

N-5

N=9

N=5

N= 2

°o = 7.14

°o =12.86

% =7 . 14

°0 = 2.86

I =1.05

I =1.03

I =0.79

I = 5.06

ESTP

ESFP

ENFP

ENTP

N-2

N = 3

N = 6

N= 2

°° = 2.86

°o = 4.29

°o=8.57

°„= 2.86

I =0.51

I =0.23#

I =1.01

I = 1.26

£577

ESFJ

ENFJ

ENTJ

N= 10

N=5

N = 2

N = 1

%~ 14.29

°= = 7.14

% = 2.86

%= 1.43

I =2.30"

I = 1.05

I =0.72

I = 1.26

Type Table

Legend: % = percent
of total choosing
this group who fall
into this type.
I = self selection
Index: ratio of per-
cent of type in group
to 70 in sample.

N % I

E

31

44.29

0.83

1

39

55.71

1.19

S

50

71.43

1.04

N

20

28.57

0.90

T

33

47.14

1.67#

F

37

52.86

0.74#

J

36

51.43

1.42"

P

34

48.57

0.76"

IJ

18

25.71

1.42

IP

21

30.00

1.04

EP

13

18.57

0.53"

EJ

18

25.71

1.42

ST
SF
NF
NT

SJ
SP
NP
NJ

TJ
TP
FP
FJ

27

23

14

6

31

19

15

5

22
11
23
14

38

57

1.67"

32

86

0.73

20

00

0.75

8

57

1.69

44

29

1.78//

27

14

0.62"

21

43

1.05

7

14

0.63

31

43

2.42*

15

71

1.03

32

86

0.68"

20

00

0.86

IN

9

12.86

0.81

EN

11

15.71

0.99

IS

30

42.86

1.38

ES

20

28.57

0.77

NOTES: concerning symbols following the selection ratios:
" implies significance at the .05 level, i.e., chi sq. > 3.8
// implies significance at the .01 level, i.e., chi sq. > 6.6
* implies significance at the .001 level, i.e., chi sq.> 10.8
- (underscore) indicates Fisher's Exact Probability used instead of chi-square.

BIBLIOGRAPHY

Allen, J. S., & Kainz , R. I. Selection ratio type table

program. Gainesville, Fla. : Center for Applications
of Psychological Type, Inc., 1976.

Anast, P. Personality determinants of mass media prefer-
ences. Journalism Quarterly, 1966, 4_3, 729-732.

Beare, D. Self -concept and the adolescent L/LD student.
TPGA Journal, 1975, 4, 22-29.

Black, W. F. Self-concept as related to achievement and age
in learning disabled children. Child Development, 1974,
45, 1137-1140.

Bruner, J., & Starkey, J. Interpersonal relationships and

the self -concept . DeKalb, 111 . : Northern Illinois
University, 1974 . (ERIC Document Reproduction Service
No. ED 089 515)

Busby, W. A., Fillmer, H. T., & Smittle, P. Interrelation-
ship between self -concept , visual perception and
reading disabilities. Journal of Experimental Educa-
tion, 1974, 42, 1-6.

Campbell, J. (Ed.) The portable Jung (R.F.C. Hull, Trans.).
New York: Penguin Books, 1971.

Carlyn, M. An assessment of the Myers-Briggs Type Indicator.
Journal of Personality Assessment, 1977, 41_(5) , 461-473.

Center for Applications of Psychological Type, Inc.

Bibliography: Myers-Briggs Type Indicator. Gainesville
Fla. , 1981.

Coan, R. W. Review of the Myers-Briggs Type Indicator. In
0. K. Buros (Ed.), The Eighth Mental Measurements
Yearbook. Highland Park, N.J ~ The Gryphon Press, 1978

Fordham, F. An introduction to Jung's psychology (3rd ed.).
Baltimore^ Md. : Penguin Books, 1966.

Fox, D. J. The research process in education. New York:
Holt, Rinehart, & Winston, 1969.

124

125

Giffin, M. How does he feel? In E. Schloss (ed.), The

educator's enigma: The adolescent with learning dis-
abilities^ San Rafael, Calif. : Academic Therapy
Publications, 1971.

Gilligan, C, Kohlberg, L. , Lerner, J., & Belenky, M.

Moral reasoning about sexual dilemmas : The development
of an interview and scoring system. In Technical
report of the Commission on Obscenity and Pornography
(Vol. 1) . Washington, D.C. : U.S. Government Printing
Office, 1971.

Golanty-Koel, R. The relationship of psychological types and
mass media preferences to the values of non-academic
high school students (Doctoral dissertation, University
of California, 1977) . Dissertation Abstracts Inter-
national, 1978, 38, 4683~rAT (University Microfilms
No. 73-29, 395)

Gordon, S. Reversing a negative self-image. In L. Anderson
(Ed.), Helping the adolescent with the hidden handicap.
Los Angeles: Academic Therapy Publications, 1970.

Gough, H. G. The Adjective Check List as a personality

assessment research technique. Psychological Reports,
1960, 6, 107-122.

Gough, H. G. , & Heilbrun, A. B., Jr. Manual for the

Adjective Check List. Palo Alto, Calif.: Consulting
Psychologists Press, 1965.

Gow, D. W. Dyslexic adolescent boys: Classroom remediation
is not enough. Bulletin of the Orton Society, 1974,
24, 154-163.

Guttinger, H. I. Patterns of perceiving: Factors which

affect individualized reading instruction at the high
school level. Paper presented at the meeting of the
American Educational Research Association, Chicago,
April 1974.

Guttinger, H. I., & Hines , V. A. Field testing and

diffusion of an experiment in developmental, individ-
ualized reading at the middle and high school level's"
(Research Monograph No. 20) . Gainesville, Fla. :
P. K. Yonge Laboratory School, University of Florida,
1977.

Hicks, L. E. Some properties of ipsative, normative and
forced-choice measures. Psychological Bulletin,
1970, 74, 167-184.

126

Hogan, R. T. Moral development: An assessment approach.
Unpublished doctoral dissertation, University of
California, Berkeley, 1967.

Hogan, R. T. A dimension of moral judgment. Journal of
Consulting and Clinical Psychology, 1970, 35(2),
205-212. —

Inlow, G. M. Education: Mirror and agent of change.

New York: Holt, Rinehart and Winston, Inc. , 1970.

Jung, C. G. [Psychological types] (G. G. Baynes, trans. &
R. F. C.'Hull, Ed7) . Princeton, N.J. : Princeton
University Press, 1971. (Originally published,
1921.)

Jung, C. G. A psychological theory and types . Lecture

delivered at Congress of Swiss Psychiatrists, Zurich,
1928.

Kainz, R. I. An investigation of response style in the
Myers-Briggs Type Indicator, a Jungian typology
inventory. Unpublished masters thesis, University of
Florida, 1978.

Ker linger, F. N. Foundations of behavioral research

(2nd ed.). New York: Holt, Rinehart & Winston, 1973.

Kohlberg, L. The development of modes of moral thinking and
choice in the years ten to sixteen. Unpublished
doctoral dissertation, University of Chicago, 1958.

Kohlberg, L. Moral development and identification. In

H. W. Stevenson (Ed.), Child psychology. (62nd Year-
book of the Study of Education) (Pt. I). Chicago:
University of Chicago Press, 1963.

Kohlberg, L. Development of moral character and moral

ideology. In M. L. Hoffman & L. W. Hoffman (Eds.),
Review of child development research (Vol. 1). New
York: Russell Sage Foundation, 1964.

Kohlberg, L. Stage and sequence: The cognitive-developmental
approach to socialization. In D. A. Goslin (Ed.),
Handbook of socialization theory and research.
Chicago: Rand McNally, 1969.

Kohlberg, L. From is to ought: How to commit the

naturalistic fallacy and get away with it in the study
of moral development. In T. Mischel (Ed.), Cognitive
development and epistemology . New York: Academic
Press, 1971.

127

Kohout, F. J. Statistics for social scientists: A

coordinated learning system! New York: John Wiley &
Sons, 1974.

Labovitz, S. The assignment of numbers to rank order

categories, American Sociological Review, 1970, 35,
515-524. ~

Lawrence, G. D. People types and tiger stripes: A practi-
cal guide to learning styles. Gainesville, Fla. :
Center for Applications of Psychological Type, Inc.,
1979.

Leviton, H. , & Kiraly, J. Achievement and self-concept in
young L.D. children. Academic Therapy, 1975, 10,
453-455. —

Maitland, K. A., & Goldman, J. R. Moral judgment as a
function of peer group interaction. Journal of
Personality and Social Psychology, 1974, 30(5), 699-

Marascuilo, L. A. Statistical methods for behavioral
science research. New York: McGraw-Hill, 1971.

McCaulley, M. H. The Myers-Briggs Type Indicator and the
teaching- learning process. Paper presented at the
meeting of the American Educational Research Associ-
ation, Chicago, April 1974. (ERIC Reproduction Service
No. ED 093 965)

McCaulley, M. H. Personality variables: Modal profiles

that characterize various fields of science. In M. B.
Rowe (chair) , Birth of new ways to raise a scientifi-
cally literate society: Research that may help.
Symposium presented at the meeting of the American
Association for the Advancement of Science, Boston,
February 1976. (a)

McCaulley, M. H. Psychological types in engineering:

Implications for teaching. Engineering Education,
1976, 66(7), 729-736. (b)

McCaulley, M. H. The Myers longitudinal medical study

(HRA contract 231-76-0051 Monograph 2). Gainesville,
Fla. : Center for Applications of Psychological Type,
1977.

128

McCaulley, M. H. Application of the Myers-Briggs Type
Indicator to medicine and other health professions
(Monograph 1, Contract no. 231-76-0051, Health
Resources Administration, DHEW) . Gainesville, Fla.:
Center for Applications of Psychological Type, 1978.

McDonnell, M. The comparative effects of teacher reinforce-
ment of self-esteem and academic achievement on
affective variables and achievement in learning-
disabled children (Doctoral dissertation, University of
Southern California, 1974) . Dissertation Abstracts
International , 1975, 35, 4287A-! (University Micro-
films No. 75-1073)

McNemar, Q. Psychological statistics (4th ed.). New York:
John Wiley & Sons, 1969.

Metts, R. E. The Myers-Briggs Type Indicator: Its use with
learning disabled adolescents. Doctoral dissertation,
Teachers College, Columbia University, 1979.

Morse, P. K. Distribution of random response scores on the
Myers-Briggs Type Indicator" Paper presented at the
First National Conference on the Myers-Briggs Type
Indicator, Gainesville, Fla., October 1975.

Myers, I. B. Manual: The Myers-Briggs Type Indicator.
Palo Alto" Calif. : Consulting Psychologists Press,
1975. (Originally published, 1962.)

Myers, I. B. Gifts differing. Palo Alto, Calif.:
Consulting Psychologists Press, 1980.

Nie, N. H., Hull, C. H., Jenkins, J. G., Steinbrenner , K. ,
& Bent, D. H. SPSS: Statistical package for the
social sciences (2nd ed.). New York: McGraw-Hill ,
T9~75\

Nuernberger , A., & Lawrence, G. Teacher self -understanding .
Competency area 1. Module A: Personality influences
in teaching and learning . Tallahassee , Fla. : Florida
Department of Education, Florida Center for
Professional Development Materials, 1974.

Plaut, A. Analytical psychologists and psychological types:
Comment on replies to a survey. Journal of Analytical
Psychology, 17, 2(7), 1972.

129

Quenk, A. T. Psychological types: The auxiliary function
and the analytic process. Diploma thesis submitted in
partial fulfillment of the requirements of the Inter-
regional Society of Jungian Analysts. January 1978.

Rosenthal, J. H. Self-esteem in dyslexic children.
Academic Therapy, 1973, 9, 27-39.

Rosser, G. J. A comparative analysis of real-ideal self-
concept of non-disabled and language and/or learning
disabled children (Doctoral dissertation, Baylor
University, 1973) . Dissertation Abstracts Inter-
national, 1974, 35, 270A. (University Microfilms
No. 74-7294)

Rowe, M. B. Who chooses science? A profile. The Science
Teacher, 1978, 45(4), 25-28.

Siegel, E. The exceptional child grows up. New York:
Dut ton , 1975.

Thompson, A. Moving toward adulthood. In L. Anderson (Ed.),
Helping the adolescent with the hidden handicap. Los
Angeles : Academic Therapy Publications, 1970.

Van der Hoop, J. H. Conscious orientation: A study of

personality types in relation to neurosis and psychosis
(Hut ton, L. , Trans.) . New York: Harcourt Brace, 1939.

von Franz, M-L., & Hillman, J. Lectures on Jung's typology.
New York: Spring Publications, 1971.

Wender, P. Minimal brain dysfunction in children. New
York: John Wiley, 1971.

Weychert, M. C. The effects of similarity-dissimilarity of
student-teacher personality type on student attitude
and achievement (Doctoral dissertation, Temple Uni-
versity, 1975) . Dissertation Abstracts International,
1975, 36(06), 3351A-3352A. (University Microfilms
No. 75-28, 258)

Wickes, F. G. The inner world of childhood: A study in
analytical psychology (Rev"! ed. ) . New York :
Apple ton- Century, 1966 .

Williams, Jr., D. L. An analysis of the interrelationships
among elementary school teachers' personality types,
beliefs, observed classroom practices, and reports of
how broadcast instructional television should be used
(Doctoral dissertation, University of Florida, 1973) .
Dissertation Abstracts International, 1973, 34, 3226A.
(University Microfilms No. 73-29, 230)

130

Wilmoth, G. H.,& McFarland, S. G. A comparison of four

measures of moral reasoning. Journal of Personality-
Assessment, 1977, 41(4), 396-40T!

BIOGRAPHICAL SKETCH

Janie Darlene Sweet is the only child of Everett and
Selma Kellams, born in Evansville, Indiana, on February 4,
1950. She attended public schools, through tenth grade,
in Charlotte, North Carolina. Moving to Los Angeles
County, California, in 1966, she graduated from Birmingham
High School in the San Fernando Valley in June, 1968. She
entered the University of Florida in Fall of 1968 and
received her B.A. in psychology in 1972, M.Ed, in psycho-
logical foundations of education in 1975, and Ph.D. in
foundations of education in 1981.

As a graduate student at the University of Florida
she served on the staff of the Center for Allied Health
Instructional Personnel, under the directorship of Dr.
Margaret K. Morgan, as a graduate assistant and graduate
research assistant until June, 1975. She then joined the
staff of Dr. Mary H. McCaulley at the Center for Applications
of Psychological Type, serving as general manager of the
center from the time of its establishment to December, 1978.

She moved to Chester, South Carolina, in 1979 and
completed state requirements for certification as a level II
school psychologist in June, 1980, through additional course
work at Winthrop College in Rock Hill, South Carolina, and
internship to Chester County Schools. She is employed full

131

132
time as a school psychologist for Union County Schools,
Union, South Carolina, and will begin the required internship
for level III certification in school psychology. Following
graduation she will teach part-time at Winthrop College.

She has two sons, Scott, age 10 and George, Jr., age 2.
Her husband, George Montgomery Sweet, also a UF graduate,
is a math instructor at Winthrop College and is working
toward completion of his doctoral program in mathematics.
Her major interests are family activities, small-scale
farming, fishing, and reading.

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

Donald F. Avila, Chairman
Professor of Foundations
of Education

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

Wilson H. Guertin
Professor of Foundations
of Education

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality,
as a dissertation for the degree of Doctor of Philosophy.

Richard J. Anderson
Professor Emeritus,
Psychology

This dissertation was submitted to the Graduate Faculty of
the Department of Foundations of Education in the College
of Education and to the Graduate Council, and was accepted
as partial fulfillment of the requirements for the degree
of Doctor of Philosophy.

August 1981

^fjj^t^^Cuu^

Chairman , Foundations of Education

Dean, Graduate School

UNIVERSITY OF FLORIDA

3 1262 08553 1290