An Investigation of Operant Elements in Desensitization: A

Connparison of Differential Reinforcement of Other

Behaviors and Desensitization in the Reduction

of Phobic Responses in Rats

By
ARTHUR M. WELLS, JR.

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
1970

To
My Wife

ACKNOV/LSDGSHEi^TS

The author wishes to acknov/ledge his indehtedness and to ex-
press his ;:^ratitude to those v.'ho gave so willingly of their guidance,
support, and patience. First of all appreciation is due to the
Supervisory Coamittee Chairi'ian, Dr. Harry A. Grater, Jr., and to all
mem'bers of the Supervisory Cocunittee, Dr. Benjamin Barger, Dr. Hugh
C. Davis, Jr., Dr. Mary lou McSver, and Dr. Henry S. Pennypacl:er.
Appreciation is also due Dr. Frederick A. Xing for his kind assistance
with experimental equipien.t and to Dr. Edv/ard F. Malagodi, Jr., for
his lean of equipment and lahoratory facilities.

Thanks are due also to Mrs. Irma Smith for her careful
preparation of the m.aiiuscript.

Finally, special gratitude goes to the author's wife-., Anastasia,
to v;hora this study is dedicated. She provider an invaluable source of
meaning to this work and to the achieveaent it reoresents.

ill

TABLE OF COHTEI^TS

Page

ACKNOWLSDaMSOTS iii

LIST 0? TABLES v

LIST OF FIGaRiS vi

ABSTHACT vii

INTRODUCTIOIJ 1

METHOD 13

RESLTl.TS 25

DISCUSSION /}1

REFEIlSiCESo 51

APPEl^iDlCL^ 54

APPENDIX A S5

APPENDIX B 56

/J'PENDIX C 57

BIOGRAPHICAL SKETT-H 58

iv

LIST OF TA5LES

Ta.-hle P^2

1. ONSET AKD DURATION 0? TONE MI) IX)OD PRiSSNTATION IN

TRSATMEt^T 19

2. ANALYSIS 07 7ARIANCE 0? AVOIDANCE RESPONSE OF THRE3 GROUPS

ON THE EIETH DAY 0? ACQUISITION TRAINING 2?

3. DAILY TOTAJ.. NUMBER OF PHOBIC RSPONSSS MADE BY THE

OPERANT ANII-'IALS OYER FIVE TREATMENT DAYS 29

4. MEA-MS ^m) STAl'IDARD DEVIATIONS OF >IU]1BSR OF PHOBIC RESPONSES

MADE DURING THE FIVE DAYS OF TEN TRIALS PER DAY

EXTINCTION PHASE 32

5. ANAJ.YSIS OF VARIANCE OF NUl-IBER OF PHOBIC RESPONSES I'ulDE

Bf THE THREE GROUPS OVER FIVE DAYS OF EXTINCTION .... 33

6. POST-HOC COilPARISONS OF THE THREE GROUPS IN ilFMBFJl OF

AVOIDANCE R5SP0N3JS DURING EXTINCTION. 3-'^

7. MEANS AND ST/JIDARD DEVIATIOIS OF CS - OR LATE^TCIIS OF

THREE GROUPS OVER FIVE EXTINCTION DAYS 38

8. ANALYSIS OF VAPJ/JCE OF CS - CR LATENCIES OF THREE GROUTS

OVER FIVE EXTINCTION DAYS 39

9. POST-HOC COMPARISONS OF THE THREE GROUPS IN LAI'ENCY OF

AVOIDANCE RESPONSE DURING MTINCTION 40

LIST OF FIGURiS

rigor e Page

1. Mean number of aToidance responses of three groups of

nine subjects over five days of conditioning 26

2. Mean nujnber of avoidance responses of tl-n^ee groups of

nine subjects over five extinction days 31

3. Average time betv;een CS onset and phobic response of

three groups of nine subjects over five

extinction days 37

Yi

Aostracl of Dissertation Presented to the Graduate Council in

Partial i\ilfillcent of the Require:: entr, Tor the

Degree of Doctor of Philosophy

at the University of Florida

M INYESTIGATION OF OPERAITT ELE1-3NTS IN DESffilSITIZATIO}!:

A COMPARISON OF DIFFiiL^.aiTIAL RSniFORCSISiT OF OTHER

B3HAVI0RS A?iD DSSENSITIZATION IN THE REDUCTION

OF PHOBIC PJISPOIISES IN MTS

Arthur K. Wells, Jr.
March, 19?0

ChairEian: Dr. Harr/ A. Grater, Jr.
Major Dspartnent: Psycholo^

An animal analogue studi' was designed to conpave the effectG of
desensitization and operar:t conditioning In reducing phobic tehayiors.
The intent of this comparison was to investigate opDiV-nt ilcirents in
desensitization rather than to evaluate the relative sfficacv of the
two treatments.

On the tasis of previous research an e:yplana.tion of desensiti-
zation in terms of operant rather than classical conditioning principles
was thought to be feasible, leading to the follov;ing hiypotheses:

1. An operant treatment procedure similar to an operant pro-
cedure possible in desensitization and presented within the framework of
desensitization facilitates the extinction of phobic behavior:" more so
than a desensitization treatment in which operant elements exe cinimizedj
thereby demonstrating the dependence of desensitization on operant con-
ditioning.

?.. Both the operant and desensitization procedures facilitate
the extinction of phobic beliaviors to a greater extent than does a

Tii

control procedure, therety demonstrating treatment
effects.

To test these hypotheses, rats v'ere conditioned to hurdle juip
in a shuttle hox to avoid electric shock \:hich followed a 500 llz.
tone (C3). After five days of avoidance conditioning the rats v;ere
assigned to tv;o treataent and one control groups, matched according
to their strength of avoidance responding.

Shock then v;as discontinued and treatment began. Both treat-
ment groups received eq.ual exposure to the CS tones (phobic stimuli)
which were presented in hierarchial arrangement. Initially the treat-
ment CS was an 8,000 Hz., 66 db tone presented for two seconds, follo-.;-
ing t'jo seconds of a seven second food presentation. On subsequent
treatment days the CS becane lower, louder and longer. By the fifth
(final) treatment day it was equal to th3 original CS tone used in
avoidance conditioning. Food presentation increasingly followed CS
onset in the operant procedure and was made contingent upon the animal's
fflaking various inhibitory responses involved in confronting the CS for
increasing periods without making an avoidance (phobic) response. Koweirer,
in desensitization the tone onset followed and was concomitant with food
presentation.

Following treatment all three groups received five dsjs of
extinction, during which the number and latencies of the phobic responses
•rfere recorded. Statistical analyses revealed the following:

1. The hj-pothesis of superiority of the operant procedure as
compared with dCiGnsitisation in reducing phobic behaviors v;as rejected.

viii

In fact, the deEie.nsitization procedure demonstrated a stronger treat-
ment effect—though not significantly so.

2. Both treatments resulted in less resistance to the
extinction of photic behaviors than did the control procedure.

The results of this work are consistent with the view that the
treatment of phohic ^jehavior may "be accomplished either hy rewarding
phohic object approach behaviors at the motor level or by the desensiti-
zation of anxiety at a more internal level. These results further
support a unified view of behavior and experience which v;ould regai'd
changes occurring at one level of overt or covert beliavior as having
implications for cha.ngc:s at other such levels.

A futui'e determination of the most effective of these two
treatments could be accomplished within the experimental framework
here introduced by removing certaj,n restrictions which were placed on
the operant procedure. These restrictions were regarded as necessary
in this present study in order to ma2:e the operant treatment more
comparable to the operant procedure involved In desensitization.

Desensitisation with humans could perhaps be made more effective
by employing and maximizing conditioned positive reinforcers, making
rewards contingent upon the subject's report of successful progression
thi'ough the hisrarcliy of threatening stLtiulus situations.

ix

INTRODUCTION

Desensitizt.tion techniques as introduced '07 V/olpe (I958) have
been successfully employed ty many therapists for the relief of pri-
marily psychoneurotic symptoms (V/olpe, 195S; Lazarus, 19^1; Lang and
Lazovik, I963; lazai-us, I963, and Paul, I966). This study represents
an attci'ipt to further our understanding of the processes by v;hich
dessnsitization success is achieved in reducing phobic behaviors.
Fears and phobic behaviors are comrion distressful problems and the
treatment of these problems is worthy of investigation.

The term £hoMc behavior is used here in agreement with
tfolpe's (1958) definition as behavior wliich is persistently mal-
adaptive in view of changed environruGutal conditions of reinforc^ricnt.
According to classical conditioning theory (Pavlov, I927) such phobic
behavior is esta'^lished when a previously neatral stimulus situation
is followed closely by punishment. The neutral stimulus v;ill be
referred to as the conditioned stimulus (C3) while the punishing situ-
ation will be regarded as an unconditioned stimulus (UCS). After re-
peated presentations of the C3, follo.-;ed closely by the UCS, the
subject (S_) reacts to the CS in a manner similar to his reaction to
the UCS. This overt avoidance behavior, termed the conditioned response
(CR), is generally accepted as a measure of fear on the part of the S_
learning this phobic behavior. The persistence of phobic behavior in
response to the CS, lorg after removal of the UCS, wry be regarded as
unadaptive behavior. Treatment of phobic behavior will be defined here
as any process which facilitates, or otherwise results in, the reduction

of the overt phobic behavior as a response (CR) to the CS when the
UCS ceases to occur. In this study shock (UCS) will he used to con-
dition auinals to jump (CR) over a hurdle frcni one end of a shuttle
box to the other in response to a tone (CS). This jumping behavior
following onset of the tone persists long after the shock is discon-
tinued. The jumping is regarded as a phobic response rather than an
avoidance response under these conditions. A pai^ticular interest here
is with the explanation of treatment of phobic responses by desensiti-
zatlon.

£oijjit_eix:Oj;iditiqnijj|^oi;_^es_en^^

Couiiterconditioning is the tera sonetiaes used in describing
experimental laboratory v.-ork with animals which is analogous to
desensitization in clinical treatment v/ith ha-nans. However, both teri.!S
describe essentially equivalent p'l'ocesses (Uolpe, 1952; i?5'3). Wclps
bases his desensitization technique on early animal studies using
shock as the UCS and using eating responses on the pfvrt of the S_ as the
reciprocal inhibitor to the previously conditioned fear (Wolpe, 1952 j
1958). This approach was similar to ths.t employed by VJatson eM Rayner
(1920) and Jones (1924); a previously conditioned phobia in a child was
treated with food given to the child at progressively diminishing
distances to the phobic object.

In a similar vein, flooding (Weinberger, I965) is the laboratory
animal study equivalent of implcsive therapy in treating humans (Sta-apfl
and Levis, I967). This latter treatment involves repeated and continued
presentation of the CS until the CR diminishes, in a massed extinction
procedure.

The treatment procedure of desensitizatiorij as describsd by
V.'olpe (1958), will be employed here with anicals. This treatnant in-
voItos having the phoMc aniaal experience the phobic stimulus while
eating food. The CS or phobic stimulus is presented along a dimension
of stimulus generalization ranging from weakest to strongest in terms
of CR evocative potential. V/hile the animal is eating, the C3 is
presented in increasingly stronger forms until the original CS form Is
presented. In working with humans the first step in desensitization is
constructicn of a hierarchial arrangement of stimulus situations which
evoke the phobic behavior, arranged from weakest to strongest in terms
of their evocative potential. After rali-Tation training, the subject
is asked to imr^gine these stimuras situations while remaining relaxed.
Wolpe caution^, that the new responce to be learned (relaxation) must be
stronger than the unadaptive response to be eliminated (anxiety). Thus,
by presenting weak portions of the C3 first, and by not progressing to
a stronger CS item until all anxiety in response to that CS item has
been reciprocally inhibited by the reciprocal inhibitor (relaxation),
desensitization will be maintained without hindrance of sensitization
(conditioning instead of removing anxiety in response to the CS). This
caution on Wolpe's part is important, though not necessarily for the
reasons which V/olpe would offer. Gale, Sturmfels, and Gale {I966) found
that though this use of progressive approximations in presenting these
threatening stimuli (C3) facilitates extinction, this method is most
effective v;hen used in conjunction with desensitization. Goldstein's
(1967) carefully controlled study suggests th;-.t desensitization is more

effective than extinction alone onlx v.'lien progressive approxiraation in
presentirxg the CS is employed; without hierarchially arranged presentation
of the CS, there is no significant difference between extinction and
counterconditioning. Thus, the procedure of progressive approxiraations
in presenting the CS is of therapeutic importance, particularly in
conjunction with the desensitization treatment method.

9l£l^ ical Conditioning- Explanation of
Desen'sitisatioii

I'/olpe (1958) offers an explanation of desensitization based
primarily on classical conditioning principles. Wolpe's explanation of
desensitization treatment of phobic behavior primarily involves the
"masking" or reciprocal inhibition of anxiety by a reciprocal in-
hibitor, i.e., a response state which is antagonistic to anxiety and
which prevents the occurrence of the anxiety. In desensitization, cues
which originally evclsed anxiety and phobic behavior come to be associ-
ated v;ith and come to evoke relaxation, which being antagonistic to
anxiety comes to eliminate the anxiety aid phobic behavior.

Wolpe (1961), using his principles cf reciprocal inhibition,
clearly suggests that in desensitization it is the inhibiting properties
of the incompatible responses (relaxation) that reduces araiety and thus
eliminates phobic behavior: "if a response inhibitory to anjciety can
be made to occur in the presence of anxiety-evoking stimuli so that it
is accompanied by a complete or partial suppression of the anxiety
response, the bond between these stimuli and the anxiety
response will be weal:ened" (VJolpe, I96I, p. I89). Note that '..'olpe

believeo that ;.'hat is treated is the anxiety and not its outv:ard raani-
festations in the foriu of oLserrable phobic behavior. Also note th?.t
Wolpe believes it is the associations between the C3 and internal
response states, such as anxiety and relaxation, which are changed in
desensitization treatment.

As previously noted, counterconditioning of laboratory animals
is the equivalent of the desensitization process. In his research
with animals, VJolps (1952) used eating responses as a reciprocal in-
hibitor and argued that the neural and physiological responses associ-
ated with eating and those associated with fear cannot occur simul-
taneously because the first are antagonistic to the latter. Thus, in
desen3iti7:ationj eating responses reciprocally inhibit fear and thereby
decrease and eliminate those beMviors based en such fear. This study
investigates desensitization with aninals cmployiiig eating responses
as the reciprocal inhibitor.

Cautela {I966) offers an explanation of the desensitization
process in ter^ns of Pavlovian theory. V/hile relaxed, the subject's
cortex is thought to be in a predominantly inhibitory state when the
excitatory stinulus (CS) is presented. The excitatory stimulus thus
loses its anxiety arousing potential after desensitization. This ex-
planation is similar to V/olpe*s in that cue-anxiety bonds are weakened
as a result of strengthening of cue-relaxation bonds.

Reciprocal Inhibitors as^Rej-vards

Relaxation responses are nost commonly used in desensitization
as a reciprocal inhibitor of anxiety. EovevGr, a wide variety of other

such inhibitors have heen eiTiployed including assertive responses
(Wolpc &M Lazarus, 1966), dinigs which induce relaxation (IJolpe and
La?anis, I966) , a comforting relationship (Bentler, I962), emotive or
enjoyable iroasery (Lazarus and Ahranjovitz, I962), playing responses
with children (White, 1959), and eating responses ('/Jolpe, 1952; 1958).
An elcnent coiar.on to all these reciprocal inhihitors is their potential
reinforcement value in operant conditioning.

Solyon and Miller (196?) h-ave used withdrav/al of punishment
(termination of shock) instead of relaxation as a "reciprocal inluhitor"
in treating pLohias. The fact th^t reciprocal inhihitors can also he
positive reinforcers suggest? the possilility of operant processes
in desensitization treatment of phohic tehavior.

Operant Processes in Desensitization

Desencitization involves having the su.hjcct imagine increasingly
threatening scenes vhile he is in a state of relaxation. The seq.uence
occurs as follows: relaxation—therapist's su^estion of the imagination
of an anxiety provoking image (the C3)-»suhject's report of having
successfully iaiagined the scene without disturbance — therapist's
suggestion that the subject relax (Wolpe and Lazarus, I966, p. 81). The
subject reports v/hether or not he is disturbed over imagining the scene
by signaling with raising his finger. B:,'- slowly progressing through
the hierarchial list of threatening scenes, from least to most threaten~
ing, anxiety experiences on the part of the subject can be eliminated
or kept to minisial levels.

Although 7/olpo offers an explar^ation of this process in terms
of classical conditioning principles as previously described, an

alternate erplanation in terms of operant learning principles appears
to be tenable and more parsimonious. Relaxation instructions or any
other "reciprocal inhibitor" would serve as a conditioned reinforcer
in sucli an operant explanation. The behavior rev/arded would be the
subject's self-report of having imagined a threatening scene. As
previously noted this self-report is given via finger signal in the
desensiti nation procedure. As the subject reports that he has suc-
cessfully imagined threatening scenes without experiencing disturbance
he is then operantly conditioned to progressively continue in this
process until he has imagined the most anxiety provoking scene in the
hierarchy. If the subject reports being disturbed in the descnsitization
procedure then he is presented with previously given discriminative
stimuli in the hierarchy and must progress back through the hierarchy
again. This procedure could be regarded as operant conditlouir^ of
finger elevation (representing the person's report of successful
imagination of the threatening discriminative stimulus) in the presence
of a fading procedure v;herein the discriminative stimuli progress in
presentation along a stimulus dimension from least to most threatening
in terms of phobic response evocative potential.

The alternate explanation of desensitizatiou in operant terms
involves modifying a response by bringing behavior under stimulus
control rather than by hypothesizing the reduction or elimination of
anxiety, upon which the phobic behavior is based. The sequence of
events in descnsitization plus the reward potential of reciprocal
inhibitors make such an alternate exolanation feasible if it can be

8

shovm that reduction of anxiety is not a necessary element in eliminating
phoMc tehavior. VJolpe, as previously descri'oed, attributes treatment
success of photic tshavior as due to reduction or elimination of
anxiety. The successful modification of phohic "behavior using an
operant procedure vould show that reduction of amiiety is not a
necessary eleir.ent in the reduction of phohic behavior, though it might
still be regarded as a sufficient element.

Studies SuriT-'Ortive of an Operant

Exjjlanation"

Studies which have attempted to analyze and study the components
of desensitization 1-iave shovm that whereas relaxation has no treatment
value alone (Cooke, I968; Johnson and Sechrest, 19^8), px-essntatio'i of
the C3 in hierarchial form without usiri^^ relaxation does Lave treatment
value (Cooke, I968; Ra,chman, I968; and Rachman and Hodgson, 19^7) •
Thus, it appears that relaxation to counteract ajixiety is not a
necessary treatment condition; ar^iety does not need to be directly
eliminated in the successful treatment of phobic behavior.

Wolpin and Raines (I966) found that having subjects simply
imagine the original full CS (most threatening scene in the hierarchy)
for more than two minutes resulted in extinction of the phobic behavior.
Implosive therapy, which involves the technique of flooding whereby
the subject is repeatedly presented with the CS in massed and prolonged
extinction trials, has been successfully used in eliminating phobic
behavior (Hogan and Kirciner, 1963; Stampfl and Levis, I967). These
procedures also mal:e no attempt to minimize or reciprocally inhibit
anxiety. V/olpc biio. Lazarus (I966, p. 1^-0) believe that v;hen such

treatment is sonetimes successful , it is only teca.usc of "transsiarginal
(protectivG) inhibition . . . the diminution of response that is ol>-
servecl vhen stimulus intensity exceeds a certain limit." Hov/ever, this
does not explain the lasting effect of such treatment to all degrees
of stimulus intensity in CS confrontation.

Reduction of phobic 'behavior is facilitated or phobic behavior
is successfully treated v/hen the subject is prevented from raalcing the
unadaptive response v;ith repeated presentations of the CS without
any presentations of the UCS (Carlson and Black, 1959 j Weinberger, I965).
In these studies the subject was prevented from makir^g the unadaptive
response (phobic behavior) ly utilizing a barrier in a shuttle box,
where jumping from one end of the box to the other upon presentation
of the CS V7as defined as the phobic behavior. An operant method of
reinforcing the subject to stay in the situation vhen the CS is presented
without the UCS should be equally successful in facilitating elimina-
tion of such phobic behavior, demonstrating the possibility of operant
processes in desensitization.

Lang and Lazovik (19^3) reported that following desensitization
of snake phobia, using relajcation as a reciprocal inhibitor, seme sub-
jects were able to demonstrate treatment success by touching a snake,
in spite of the fact that after treatment they reported no decrease
in their pre-treatment anxiety concerning touching a snake. This is
exactly what vrould be expected in terms of an operant learning view of
desensitization technique. Other subjects who reported a decrease in
anxiety levels following treatment may have been responding to the
demand characteristics of the experimental situation (Orne, I96Z) or

10

the e>:pectations of th3 erpsrinenter (Rosenthal, I966), Another
possihility for dscrenent in anxiety is that airciety vfas experienced
and then extingaished in the absence of the UC3 during desenpitization
treatment.

Results of studies cited in this section all demonstrate that
anxiety does not hs.re to he reciprocally inhibited in the treatment
of phohic hehavior. This fact coupled with an alternate explanation cf
events in the desensitization procedure in terms of operant conditioning
leads to the hypothesis that at least part of desensitization success
is due to operajit procedures.

Formulation of. ,.the^ Hypotheseg

Treatment of phohic heha?ior may involve facilitation of ex-
tinction processes. These extinction processes do not al\:o.ys operate.
^Ihevi the subject cakes an unadaptive response (CR) in response to the
CS, this cay preclude the subject from experiencing the absence of the
UCS. Fii.i.'theraore, the frequently short latency betv.'een onset of the
CS and the occurrence of the CR may serve to prevent an anxiety reaction,
thus conserving conditioned anxiety from undergoing extinction (Delude
and Carlson, 196k ; Solonon and Wynne, 195^). Desensitization, operant
techniques, flooding/carrier methods, and other procedures used in
treating phobic behavior are considered here to be successful to the
extent th^t they allow, encourage or force the subject to experience
both the C3 and the lack of the occurrence of the UCS, so that extinction
caji occur.

11

Wolpe's eiqilanation of descnsitization in terms of a primarily
classical conditioning paradigm, in v;hich anxiety rast be eliminated to
eliminate phobic behavior, is questioned here. This e'-qjlanation
requires inferences as to unobservable interral processes vMch have
not been experimentally established (Breger and I'cGraugh, I965).

This present study concerns an attempt to show that phobic
beha,vior can be treated effectively by operant methods, concerned only
with bringing the subject's observable behavior under stiEulus control,
without attempting to reduce or eliminate internal anxiety. This bears
on the question of the necessity of reciprocally inhibiting anxiety
in treating phobic beliavior. The first hypothesis is that an opera.nt
treatment procedures presented within the frajnev/orl: of the technique
of desensitization, v;ill result in significojitly less resistance to
extinction compared to the disuse or controlled rest condition.

Previously an alternate explanation vas offered of the desensiti-
zation process in terns of oper?^t principles rather than in terms of
classical conditioning theory. Such an alterriite explanation v;ould
find support if an operant procedure proved to be a more effective
treataent procedure than a desensitization procedure in vhich operant
processes were minimally operative, i.e., restricted to a continuous
and simultaneous presentation of revfard with the presentation of the
CS (so tiiat coni'rontation v;ith the CS is possibly rev/arded). The
second h^-pothesis is that an operant treatment procedure, presented
within the fraineworl: of desensitization, v/lll result in significantly
less resistance to extinction compared to adesensitization procedure, in
which operant procedures are minimally involved.

12

It is assumed that these treatment procedures can be applied
to the previously conditioned phobic behavior of animals in the follov;-
ing v:ay. First, using a 500 Hz. tone as the CS, a conditioned avoidance
response to shoclc (UCS) can be established. This response can then be
treated by using the progressive approximations method in presenting the
CS without the UCS, starting vjith an 8,000 Hz. tone which does not evoke
the avoidance response, then progressing tc the original CS (500 Hz.
tone). Such progressive presentations of the CS, in hierarclrial arrange-
nent frora veekest to strongest item, can then be combined with either
a desensitizatioii treatment or an operant treatment. In the desensiti-
zation treatraent the hierarchially arranged conditioned stinrali can
be gradually introduced in increasing approxination to the original CS
while the animal is eating. The various responses involved in this
eating behavior should serve to reciprocally inhibit the anxiety aroused
by presentation of the CS tone. In the operant treatnent the food is
presented in a way such that obtaining it and malting an avoidance
response are mutually exclusive. In this operant procedure, obtaining
food requires that the animal inhibit his making an avoidance response,
making instead various other responses which are competitive with the
avoidance response, for increasing periods of time.

Both of these procedures can be called effective treatment
procedures only if they result in significantly less resistance to ex-
tinction than the disuse or controlled rest conditions. Thus, the
third b^-pothesls is that the desensitization treatment will also
result in significantly 1g-=5s resistance to extinction compared to the
disuse or controlled rest conditio:)..

METHOD

Subjects

Thirty experimentally naive, initially 80 days old Long-E^ans
male rats were used as subjects. These 30 rats were selected from a
gro::i.p of k-Z on the basis of their overall performance during the
Acquisition Phase. Selection v:as accomplished by taking out the four
fastest a2id eiglit slov.'est rats in terns of average reaction tins
bet'.;cen CS onset and CH or hurdle jumping. In the middle of this
study three rats died, leaving data on Z? animals (see discussion
of matching S_s in H^catinint section belo.^').

Animal Maintenance

Each ani-Tal was separately housed in a hanging, letal C2,^q In
a controlled tenperature environment. ¥ater was given ad ]ib in all
phases of the experiment. Food vas presented ad lib for five days
prior to experimentation and then the animals were reduced to 30 per
cent of their normal ;;eight. Normal weight v;as determined oj taking
the median of the daily weights over th.e last three days of this five
day pre-experinenlal period. After normal bcdy veiglit v;as dcterainGd
the ajiimals went on their diet for tv/o djajs prior to beginning
acquisition training. They were maintained at 80 per cent of normal
body weight during the Acquisition and Treatment Pha.se3 of the experi-
ment, but were again placed on an ad lib food schedule during the
Extinction Phase of the exoeriment. Tlie rationale for placing the

13

14

animals on an ad lib focd schedule during the extinction trials is
presented in the Discussion section. All animals were fed each night
at the end of the experimental work^ Purim Laboratory Chow was used.

Apparatus

The wooden shuttle box was painted black and had interior
dimensions of 2''i inches long by 6 inches wide by 18 inches high. The
floor consisted of k6 copper alloy bars spaced one-half inch apart.
The grid floor was divided in half by a four and one-half inch high
wooden hurdle which v;as in place at all tines. Only one end of the
floor was wired for shock. This saaie "hot" end also contained a food
tray which could be pushed into the side of the box pnd which v:as
centered betv.'eoii the hurdle and the end of the box.

During the Treatment Phase, 20 mg lloycs food pellets were
presented by sliding this tray into the bo/.. The empty tray was kept
out of the box, the end of the tray covering the box tray opening,
during the Acquisition and Extinction Phases. A plexiglass sheet
covered a rectangular opening on one side of the box and permitted
full observation of S^Sj behavior. The other side of the box contained
numerous one-quarter inch holes to permit sound to enter from an
adjacent speairer.

The shock for the grid was delivered by a Lehigh 7alley
Electronics Shock Generator and Scrambler unit which reversed the
polarity randomly aniong the 23 grid bars on the hot side of the box.
Level of shock was set at one and two-tenths milliapperes (D.C.) as
measured by the Simpson meter on the shock unit.

Iji AR^x spealier was used adjacent to ons side of the shuttle
box with the woofer nsar the "hot" end of the tox. The tones were
supplied froa a Keulett-Packard, model 200AB audio Oscillator. Tones
vere aeasured ly a General Radio Coicpany Sound-Level Meter, type 1551-B,
inside of the shuttle hox with background noise at a 35 db level. A
Hunter interval timer was used to measure event duration to the nearest
hundredth of a second. Manually operated switches controlled the
shod:, tone and timer. V^hen appropriate the switches fro^i the tone
generator could be yoked to the timer to measure the duration of the
tone. The miming ticier was also used to measure food presentation
duration.

Procedure

Acquisition Phase

rurin^ this Phase each of the 42 rats received 20 trials of
acquisition training for each of five consecutive days. On the first
trial of each day the animal was placed in the hot end of the shuttle
box and allowed to roam for a period of two minutes. The 500 Hz., 90 db
tone (the original CS) was then turned on if the animal was in the hot
end of the box. If the animal was not in the hot end after the two
minute period then he was gently placed there. After the CS tone h3.d
been turned on for a period of ten seconds shock was delivered. The CS
tone (and shock if turned on) was immediately terminated and time
recorded as scon as the an.icial jijmped, withdrawing all contact from the
hot grid floor. In effect then the animal quit receiving both the

16

sliock and the tone as soon as he jumped. He did not receive shock
prior to ten secon'^.s from the tone onset.

The latency "bctvfeen the onset of the CS tone and the jumping
or avoidance rci-ponse was recorded to the nearest hundredth of a
second. Recording of latency was accomplished hy having the CS tone
duration switch yoked to the timer so that the duration of the CS
tone (v/hich terr'.inated v/ith ^s jumping clear of the hot grid floor)
was timed. A latency of ten seconds or greater indicated that the
animal did not jump lofore receiving shock.

On trials follov/irig the first trial of each day during the
acquisition training, the animals \/ere allowed 20 seconds in the
"cold" end of the shuttle toz following hurdle jyjnpir>>g and CS termina-
tion vhich ended a trial. They did not receive the CSj marking the
beginning of a nev; trial, until at least ten seconds aftc-r "being placed
hack in the hot end of the "box. Animals jumping vitMn ten seconds of
"being placed in the hot end of the "box to "begin a new acquisition
trial did not receive either the 03 or the UC3 (shock) but were gently
placed "back into the hot end to "begin the trial again. Following the
20 daily trials each animal v;as weighed, returned to his ce^e and fed
an amount of Purina La"boratory Chov; "based on the deviation of his weight
from 80 per cent of his normal weight.

By the fifth day of training, all animals reached the acquisition
criterion by hurdle jumping at least 7^ per cent of the time before the
dui'ation of the CS reached ten seconds (thus avoiding the shock).

1?

Hatching

Folloying the fivs days of acquisition training all i|2 animals
vere ra_nked according to their overall average reaction time (betv/een
the CS onset and jumping behavior). The four animals with the fastest
reaction tiaes as well as the eight slowest reacting animals were dis-
carded from further experimentation. Three groups of ten animals each
vere then drawn for the tv.'o treatment and control groups. The Control
group vas comprised of the following ranked aninals: 1, 6, 8, 12, 14,
16, 20, 22, 27 and 30. The Operant treatment group was sade up of
animals v;ith these ranked numbers: 2, A"-, 9$ 10, 15, 1?, 21, 23, 25
and 29. The Deseusitization treatnsnt group vas coir:?csed of the follov;-
Ing rank numbered aninals: 3, 5, 7, 11, 13, 18, 19, 2^1, 26 and 28.
The average reaction time over the 100 trials of acquisition training,
computed for ea-^h group of Icn rats aft^r the anioials were formed into
the three groups, v,'as as follGv;s: Control group, k.33 seconds; Operant
group, 4.94' seconds; and the Desensitization group, 4.94 seconds. The
groups of animals v.-ere ihus thought to be well matched in terms of the
overall average time it took then to jump the hurdle upon onset of the
CS during acquisition training.

On the third day of treati:.ent three of the experimental animals,
one from each of the three groups, were founl dead. Although no autopsy
vas performed the presence of blood traces near the external urinary
region of these animals suggested a urinary tract infection. The remain-
ing 2? S^s survived the experiment, apparently in good health.

The elimination of these three animals did not appear to upset
the previous matching. The ranl-ed positions of these S_s, in terms of
overall rraction time in the acquisition trials, were 17, 18 and 20.

10

The average reaction times of the three groups changed as fol],ov;s wltl;
the eliDination of the three animals: Control group, dropped fron 4.93
to 4.90 seconds; Operant group, reiaained the same at 4.9^ seconds; and
the Desensiti7ation group, dropped from 4.9^ to 4.93 seconds. Since
matching appeared to he maintained the study progressed as scheduled v;ith
the remaining 2? S_s,

Treatment Phiise

Treataent tsgan the day after conpletion of the acquisition
trials. Both the Operant and the Desensitization treatment gi-oups
received equal exposure to the CS stimuli components during treatment.
Both groups also received equal exposure to the food pellets used in
treatment. The tonal hierarchy presented during treatment in accord-
ance with ths method of progressive approximations consisted of the
following five tones! an 8,000 Hz., 66 dh tone presented for two
seconds; a 4,000 H2., ?2 dh tone presented for three seconds; a 2,000
Hz., 78 db tone presented for four seconds; a 1,000 Hz., 84 dh tone
presented for five and four seconds; and a 50O Kz., 90 dh tone (original
CS) presented for four, six and seven seconds. These tones were pre-
sented over the five dsy Treatment Phase period in a total of 84 treat-
ment trials. Food was also presented. The timing of the food presenta-
tion in relation to the tone presentation defined and distinguished the
Operant and Desensitization treatment groups. Table 1 provides a
detailed description of the tonal hierarchy and the presentation of
these CS tones In relation to food pellet presentation in the two treat-
ment groups.

19

1-f o

CD

-C) f:i

o o

o .-^

f:;

O

rj

c-

+

q3

f-i

•O' C>

E-^

O f-1

o o

p;

Ps &-.

o

'i

t:

CJ

■♦-•'

Ti d

o o

10

O r-i

C

P=H<J

o

in

<D

o r!

C! O

O rH

C-l <;•

o

'd t:i

o o

O I— i

P":; -<

+-

+

K

o

Ti <D

e

o r:

*->

o o

cd

(is E-.

<D

t-4

^

4-'

t:;

<D

c3

■zi «

,%

o o

o

O r- 1

p

^ <

o

C>

Qj a

« o

O r-H

E-. <

t-j-c!

t> c r!

c it! o

a> ■r^

;U i-i +^

W CD c3

ei t> f 1

^^B

r-t

ri

•r^

fH

EH

>?

c5

« I

O w

C0"0 CO

O ^ O

O ,0 O ,o
O w

O «n

T-H vr\

20

-Cf 03

•ci <D

O V)

O C3

OO

o o

V^o^vO

Vr^a^C^.

21

As can "be seen frora Tatlc 1, both treatment groups received the
same stimuli conditions during the 16 trials of the first treatiaent day
and the first four trials of the second treatment day. On the first
treatment day each animal v;as initially put into the hot end of the
shuttle hox (no longer receiving shock, houever) and immediately the
filled food tray ua? pushed inside the hox. The animal v;as allov/ed
either to eat or to roam freely, hut as soou as he had eaten frora the
tray for 30 seconds the food tray was v/ithdrawn. After a ten second
delay the tray was reintroduced and two seconds after the animal had
again star-ted eating the 8,000 Hz., 66 dh tone was turned on for tvjo
seconds. After completion of this tone the food tray remained In the
box for an additional three second? and then was withdrawn. There was a
ten second interval between trials and S was always in place in the hot
end of the box at the beginning of a trial. These first 20 trials con-
bined both operant and desensitization treatment elements but on subse-
quent and succeeding trials food presentation beca'ie increasingly de-
layed in the operant treatment in a shaping procedure designed to rein-
force S_*s confronting the OS stimulus component, making various other
and unmeasured competitive responses, without making the phobic response.
The Desensitization group never received the OS tone without prior and
concoaitant food presentation.

Beginning with the twenty-first trial, which, was the fifth
trial of the second treatment day, timing of the onset of food presenta-
tion began with the introduction of the food tray and not, as previously,
with initiation of eating behavior on the part of _S. It is impossible
to reliably obtain a constant delay of food presentation relative to

22

previous tone onset when food presentation is defined in terms of S_'s
initiation of eating tehaviors, but food presentation defined in terms
of food tray introduction can easily be delayed at a fixed period after
a tone onset in a consistent and accurate fashion. There vas no
assurance that S_ would begin eating as soon as the food v;as presented
in the initial trials because though deprived he had no previous ex-
perience with the food's availability. This made it necessary to
define the start of food presentation in terms of S*s initiation of
eating behaviors in these first 20 trials. Stimulus conditions in
these first trials made such a definition quite feasible in regards to
procedural reliability. After the first 20 trials, S_s had all becorae
acquainted with the food tray's location and availability such that
introduction of the food tray v/as practically equivalent to S_'s
initiation of eating behavior.

After the first treatment day all treatment trials were con-
ducted in the following manner. First an animal was introduced into
the hot end of the shuttle box and after a period of at least ten
seconds, and while the animal was in the hot end of the box (being
gently put back there if he had roamed over the hurdle during the ten
second period), the treatment stimuli were introduced. In the case of
the operant procedure the treatment stimuli introduction consisted of
the onset of the C3 tone component followed later by food presentation
which was contingent upon the animal's confronting the CS withoiit
Railing an avoidance response. In the desensitization method the onset
of the CS tone component followed the food presentation.

23

As treatment progressed each S_ was exposed to tones which
increasingly approached the original CS tone in terms of frequency or
pitch, volu5}8 level and duration. Thus, after "beginning treatment
vith a relatively soft, high-pitched and short tone, v/hicli had very
weak CR evocative potential, treatment ended vfith the loud lov;-
pitched and long lasting tone used as the CS in the acquisition
training. No shock was administered to any anit^^l at any tiiae follo'j-
ivg the completion of the acquisition training; thus hurdle jujuping on
the part of £ in response to the tone fulfilled the experimental
definition of a phobic response. Food was presented for a period cf
seven seconds in all trials except the last eight trials where it was
presented for eight seconds. The stimulus conditions for the two treat-
ment groups are described more precisely in Table 1.

During the Trcatasnt Phase the control animals were transferred
from their ciiges to another cage prior to their daily weighing. This
was done to equate all animals for handling. The control animals were
not othenr;ise manipulated and were kept in the saae isolated environaient
receiving the S£'jue stimulus exposure as the treatment animals, except
of course for the treatment stimuli.

Extinction Phrise

The extinction trials began on the day after the fifth (final)
treatment day, and consisted of ten trials per day for five consecutive
days. These trials were similar to the acquisition trials in most
respects. Of course neither UCS (shock) nor food was presented as was
the case respectively in the acquisition and treatment trials. Initially

Zii

on each day of extinction the animal v;as placed in the hot end of the
"box, then alleged to roam for two minutes- During: extinction the
empty food tray was always pulled in the "out" position so that the
end of the tray covered the tray opening in the side of the shuttle-
box. At the end of the two minute exploratory period the animal was
placed again in the hot end of the "box (if he was not already there)
and the CS tone (500 Hz., 90 db) v/as presented and concomitantly was
titled. A3 soon as the aninal made a phobic response (by withdrawing
all physical contact from the grid floor on the hot side of the box)
the CS tone was terminated and simultaneously the timer stopped. If
the animal did not jump at the end of 20 seconds, chosen as the
infinite latency period, then the CS tone v;as terminated and his
latency was recorded as 20 seconds (indicating that no phobic response
was made). This procedure provided measures for the two dependent
variables, e.g., latency or delay between onset of the CS and the
occurrence of a CR, and whether or not the phobic response or OR was
made. After the first trial of each extinction day S_ v;as allowed to
freely roam for a period of 30 seconds before the CS presentation for
the next trial. If an animal roade a CR during this pre-trial period
then ten seconds later he v;as placed back into the hot side to begin
the 30 second exploratory inter-trial period again.

RESULTS

Data obtained from each of the three phases of this study
v;ill be preseoted in sequential correspondence to the experimental
order.

Acquisition Phase

The latency betv;een CS onset £:id CR occurrence was recorded
over the 100 trials (20 trials per day) of acquisition training-. A
recorded latency of ten seconds or greater indicates that shock was
administered on that trial. These data then provide two measures of
CR acquisition strength in the speed with which a CR v/as evoked and
in the frequency of CR occurrence prior to UGS onset (or vlthin ten
seconds).

Figure 1 demonstrates these data graphically showing the mean
number of avoidance responses (CR's) made by each of the three groups
for each day of acquisition training.

Inspection of Figure 1 indicates similarity of conrlitioning
rates among the three groups. All of the sninals achieved the criterior
of avoidance training' by making a CR prior to UCS onset at least on
15 of the final 20 acquisition trials,

A one-way analysis of variance on the number of avoidance
responses for the final da^'- of acquisition training supported the
contention that the t'nree groups of Ss had been successfully matched

7A

26

Conti-ol c^ — =»—— o

Operant t*— ••—•■©

Desensitization ©..........es

19-

17-
16-.
15„

y 13-

12.

11.

lOu.

7-
c-

12 3^ 5

Acquisition Phase: Jive Days, Tv'enty Trials Per Day

Fig. i.—Mean numt-sr of avoidance responses of thi-ee groups ol
nine su"bjactr= over five days of cor.ditioaiiig.

27

on the strength of acquisition variable. This aiialysis failed to
indicate significant differences ojnong the Vhree gro-ops in nimtier of
avoidance responses for the last 20 acquisition trials, as can "be
seen in Table 2 belcv;.

TABLE 2

ANALYSIS 01 VARIANCE OF AVOIDANCE REPOIISE OF THREE GROUPS ON
THE FIFTH DAY OF ACQUISITION TRAINING

Sources of Variation df Mean Square F

Between 2 A5 .^^^

Ei-ror 24 1.03

Total 26

df (2, 24) .05 = 3.'.'0 .01 =: 5.61

Matching was acconiplished not on the basis of the number of
avoidance responses made prior to the UCS, but rather on the latency
of these avoidance responses. The average CS - CR latency by group
over all five days of acquisition training was as follows: Operant
group, k,9hi Desensitization group, 4.93; and the Control group, 4.90
seconds. These mean latencies for the fifth and final day of acquisition
training were as follovjs: Control group, 3*63; Openint group, 3«27; and
Desensitization group, 3.22 seconds. Note that although the Control
group demonstrates the fastest overall mean reaction time, indicating
strong acquisition of the avoidance response, the Control group's
reaction time for the fifth and final day of acquisition training (just

28

prioi- to treatneat attempts to increase this reaction time) is the
slov;est of the three groups. As in the case of the numl^er of avoidance
responses made during acquisition, the speed with which these responses
were made is essentially equivalent araong the three experimental groups.

Trgatment Phase

The only recording made during the treatment trials was the
nuTiher of avoidance or phohic responses made by S_s during C3 tone
component presentation. In the Desensitization group, which received
food prior to and continuing throughout exposure to the CS tones, only
two of the nine animals made phohic responses. These two anin.als only
made one such response each for a total of tv/o, and these two responses
were toth made only on the first treatment day. In the Operant group,
which received food usually seconds after the onset of the CS tones,
eight of the nine S_s made a total of 21 (of a possiWc I'^o total
trials) phoMc responses. Taole 3 descrities the distribution of these
21 phobic responses over the five treatment days. Exaaiination of the
data in Ta.ble 3 reveals that the operant fading procedure employed was
generally successful iu maintaining the desired beh?,vior, and that a
small percentage of phobic responses occurred.

Both treatments appear to have been accomplished successfully.
The nearly total prevention of CR occurrence in the desensitization
procedure attests to the effectiveness of the hierarchy used. Probably
the only way to have eliminated CR occurrence in the operant treatment
would hr'.va been to make food presentation less delayed relative to 03
onset, tat this would have blurred the distinction between the two

29

TABLE 3

DAILY TOTAL. KOl-IBER OF PHOBIC RESPOIISES }y..DE BY THE OPSPANT Al^Ii^XS
CTER FIVE TPE.IT}-IE1IT DAYS

Treatment

Day Nmnter and Nuntsr of
Per F^ch /jiinal

Treatment

Trials

Animal

Day 1
16 Trials

Day 2
Id Trials

Day 3
16 Trials

Day 4
16 Trials

Day 5
20 Trials

1

0

0

0

0

1

2

1

1

2

1

0

3

0

0

0

0

2

4

2

0

1

0

3

5

0

0

0

1

2

6

1

0

1

0

0

7

0

0

0

0

0

8

0

0

0

0

1

9

0

0

0

0

1

30

treatment groi.'ps; the operant treatment would have tecoiiie more of a
desensitization treataient*

Extinction Phase
Treatment effects v/ere measured during the extinction trials
both hy the munLer of photic responses made and hy the latency in
responding. These data can be found in Appendices A, B and C.
Considering the nraber of avoidant responses made by the three
groups in the extinction curves in Figure 2, differences a'song the
groups are readily noted. These differences are further reflected
in the means and standard deviations of the avoidance responses as
contained in Table h. Both treatment groups made less phobic responses
during post-treatcent extinction trials than did the ncntreated con-
trols. This difference was most evident on the first day of extinction,
imnisdiately follov/ing the last treatment day. A two-\;ay, mixed model
(Hays, 19^6) analysis of variance is summarized in Table 5« These
results require the rejection of the hypothesis that the three groaps
were sampled from the same population. To further describe differences
attributable to treatment, a Newm.an-Keuls test is presented in Table 6.
These results indicate that both the Desensitization and Operant treat-
ment groups are significantly different from the Control groiip, but
that the Operajit and Desensitization groups are not significantly
different from each other. Thus, the significant treatment effect is
primarily due to differences between the treated and untreated groups.
Changes over "days" of the extinction trials are to be expected; the
significant effect of this variable is of little concern to this

31

Control ©■ — --^

Operant e^ ©

Desensitization © -©

icu

7

6-

5-

4

3

2-

0«

^^^.
X)..

■^^..^..

*'C"

nine

Fig.
su'bject;

12345

Extinction Phase: Jive Days of Ten Trials Each

2.-~Mean numter of avoidance responses of three groups of
; over five extinction days.

32

TABLE 1}

MFANS AND STAl'IDAIU} DEVIATIONS OF lIUi-IBEH OE PHOBIC RESPONSES 14&DE
DURING THE FIVE DAYS OF TEl^ TRIALS PER DAY EXTINCTION PHASE

E:cti notion Day

Group

1

2

3

4

5

Descnsitization

X

3.2

^.7

6.8

5.4

5^6

h

3.18

3.64

2.47

1.72

2.70

Operant

I

ii.Z

6.0

7-3

C.7

6.7

%

3.35

2.58

2.99

2.79

2.94

Control

X

8.9

8.6

8.7

9.3

9.4

h

.89

1.61

1.59

1.03

1,14

33

TABLE 5

ANALYSIS OF VARIANCE OF IIIIIBKR OF PHOBIC MSP0FS2S MADi] Bf THE
THRE3 GROUPS OVER FIVE DAYS OF EXTINCTION

Source of Varia

,tion

df

Mean Square

F

Treatment

2

177.80

36.U'«-^^

Days

4

19.85

2.53^^

Interaction

8

4.92

.73^

Error

120 (128)^

6.90 (6.78)'^

Totals

134

^df (2, 8) .05 = ^.h6 .01 = 8.65

^df (4, 128) .05 = 2.42

^df (8, 128) .05 = 2.01

•^Pooled MS error, used to test Days and Interaction effects

34

TABLE 6

MO^

POST-HOC COMPAPJSOKS 0? THE TH.RE'E GROUIS IN MTOiBER OP AVOIDANCE
RESPOHSK; DUHING EXTINCTIOxI

Desensiti-
zation

Treatment Totals

Differences

Differences ~
|/inis"erroi''

231

Treatment Methods

^'Nev.-man- Kelt} s Test
\.95 (?•» 120) ■-= 2.80
%99 (2, 120) = 3.70

Operant

47

2.72

278

126

7.15^

-..•.c

Control

404

35

investigation. It should "be noted that coniplete extinction did not occrar,
and that the treatment effects were measured during "but not tliroughout
the extinction process. The "days" factor in the analysis of variance of
nrnnher of phobic responses uas regarded statistically as a random factor,
sajrpled from all possible extinction days.

Turning now to tlie second dependent variable, pliobic response
latency, differences aiuong the three groups can be seen in the ex-
tinction carves of figure 3. These differences are most evident, as
in the case of the phobic response occurrence variable, immediately
after treatment on the first day of extinction. Also, as in the case
of the first dependent varie.ble, both treatment groups appear to be
clearly superior to the Control group in decrement of phobic behavior
and the densensitization treatment appears perhaps more effective than
does the cpsrcxnt procedure. The moans and standard deviations of these
latency data are presented in Table ?. Analysis of the variance of
these results, summarized in Table 8, again as in the case of the
first dependent variable, demonstrates a strong treatment effect. A
two-';;ay, cixed model analysis was en'ployed with an error term for test-
ing interaction and row effects (the random effects variable) derived
by pooling the interaction and error mean squai'es, again as was done in
the first dependent variable analysis of variance.

To more precisely delineate the effect of treatments on latency
of the phobic response during extinction another Nev:man-Keuls test was
calculated (Table 9). These results again parallel those obtained with
the CR occurrence data previously described in that either the Desensiti-
zation or Operant groups were significantly slouer in responding to the

36

CS than v;as the Control group, but that the two treatment groups raisr,
loii]^ significrjitly different.

The effect of desensitization and operant procedures in the
reduction of tv;o behavioral indices of phobic behavior appears to be
highly facilitative as compared with no treatment procedure at all,
but the hypothesis that the two successful treatment procedures
differed in effectiveness escaped rejection and the differences found
in this study aust av;ait futui-e investigation for clarification.

Jf

K
s

o

en

i

16-1
15

13 ~
12

11 J

.10

9
8
7
0

Control

Operant
Desensitization

N

V

— _„^ — , J„ ^ — --.»--J .____^_

1 2.3 ^ 5

Extinction Phase: Five Days of Ten Trials Each

Pig. 3.— Average time betv/een CS onset and phobic response oi
three groups of nine subjects over five extinction days.

38

TABLE 7

^ffiAlI3 MD STANDARD DL7IATI01JS OF CS - CR LATEilCI^S OF THREE
GROUPS OVFii FIVE EXTINCTION DAYS

Ext:

Lnction fiav

Group

1

2

3

4

5

Desensitization

X

16.09

14.83

11.9?

13.63

12.44

^D

4.68

5.35

3.69

3.50

6.03

Operant

I

15.24

13.73

11.10

12.31

11,16

h

3.91

3.45

4.08

3.48

4.01

Control

1

7-55

9.72

8.87

8.85

7.09

h

2.18

4.33

2.?4

3.58

4.58

39

TABLE 8

ANAI.YSI3 0? VARIANCE OF CS - CR LATENCIES OF THREE GPOIJPS
OVER FIVE EITIIiCTION DAYS

Source of Y?.riation

df

Mean Square

Treatment

Days

Intci-action

Error

Total

2
if

8
120 (128)'
134

356.08

39.08
12.40
18.12 (1?.76)'

^-df (2, 8) .05 - ^M .01 = 8.65

28. 72*^

2.20

.70"

°df (4, 128) .05 ^ 2.42
'^df (8, 128) .05 = 2.01
'^Pocled M3 error, used to test Days and Interaction effects.

TABL3 9

POST-HOC COKi^ARISOI-iS OP THE THRKE GROUPS IN LATffllCI OF
AVOIBMGE RE3PONSS DURING EXTINCTION

40

l)esensi'fi~
zation

Treatment Totals

620.63

Treatment Methods
Opcra.nt

572.85

Control

381.43

Differences

Differences -:-
/nTlS~erfor

47.78

1.67^

191.42

i,7^'

Q.95 (2, 120) = 2.80
V99 (2, 120) - 2.70

Discussion

Analysis of the acquisition data indicates that successful
matching of groups vas achieved in this investigation so that the three
groups of animals had approximately equal concHtioncd avoidance
response tendency prior to the treatment phase. Thus the three groups
were prohahly nearly equal in the strength of their phobic response
tendency, and were prohahly also nearly equal in their ability to
"unlearn" the previously conditioned avoidance response, when treat-
ment began.

Data obtained during the Treatment Phase indicate that both
treatments cjmaged the ra.te of the avoidance responses of the aiiijials
during treatment trials. Both treatoent groups received a nearly
equal exposure to the CS throughout this experiment, so that the treat-
ment effect involved in CS ejpnsu.re vas equivalent araong treaticent
groups. Treatment effects due to extinction vrere considered minor since
the phobic behaviors of the Control group on the fifth (final) day of
extinction trials were nearly equal to those on the first day of ex-
tinction, and since the phobic behaviors of the Control group on the
final day of extinction v;ere much greater than those of either the
Operant group or the Desensitization group on the first day of er.tinction.
Both treatnent groups also received an equal opportunity to eat during
treatment. All 3s were maintained on similar food deprivation schedulss
prior to and during* the first two phases of this study, and all the
groups were fed ad lib duririg the firial phase.

&1

42

The reason for changing to aJi ad lib food schedule in the final
experimental phase was to achieve v.'hat vfas thought to he a more pure
mea^iure of the effects of treatment "by reducing deprivation effects on
response maintenance. Under food deprivation conditions the operant
treatment v/ould have tieasured out as more effective than v;as the case.
The fact that the measures of phohic hchavior on the final t\f0 days
of extinction, when any food expectation should have extinguished to a
large degree, are very similar to these saine measures over all five
days of extinction suggests that food expectation played a rtinor role
in the effects of treatment (as measured with well-fed animals).

Effects of operant procedures are typically measured after
removal of previous reinforceaent, hut not necessarily v;hiie needs
satisfied by the reward are reduced to a state of satiation. In attempt-
ing to measure more lasting effects of treatment, rather than temporary
and perhaps artificial effects, the operant treatment was ha^npered in
this investigation.

The operant procedure employed was a discriminative differential
reinforcement of other behaviors. The desired event in the operant
procedure, not jumping, involved a vai'iety of irihibitory responses
(largely superstitious in nature). It was these varied ur^measured
irihibitory responses v/hich were actually reinforced. These reinforced
responses, which were somewhat incompatible and antagonistic to the
phobic response, were brought under the control of discriminative
stimuli presented along stimulus dimensions (frequency, loudness, and
duration of tout) through a fading procedure. The tone ca-ae to ser/e
as a discrimiP-ative stimulus for food presentation in the operant

^3

procedure. D-aring the extinction trials, when the animf.ls were satiated
for food, the tone was irrelevant as a discrirainative stimulus, making
the operant animal much more likely to malce the phobic response than
would have heen the case had the food-deprivation schedule "been continued.
However, by providing; all ariiaals with food during the Extinction Pha^e,
the operojit treatment procedure Eore closely approximated the hj^pothe-
sized operant elenents in the desensitization treatment. The desensitized
animal dononstrates his treatment effect whether food deprived or food
satiated and any operant procedure involved in the desensitization
treatment must also stand up under food -satiation conditions.

A one-way shuttle box was used zdch that one side consistently
received shock dui'in^; acquisition, food during treatment, and the CS
during all experlBental phases. If a two-way shuttle boz ha.d been
used, in which both sides alternately received shock, then both ends
of the box vTOUld have taken on an aversive quality for the aniaals.
This potential source of confusion for the animals was avoided, though
thereby probably making treatment more difficult, by consistently offer-
ing the animals a sanctuary in the no-shock end of the box. Thus the
quick phobic response was alv;ays clearly associated with avoidance of
shock, and was always an available course of action to the animal.

The operant treatment involved the reinforcement of responses
involved in an animal's confronting the CS without m.aking a phobic
response. These reinforced responses were shaped in that their occurrence
over increasing periods of time was required in order to obtain food.
Operant treatments usually atteiapt to increase the rates of more active

¥i-

and positive responses. Hovever, in the experimental procedure employed
here, the reinforceniont of a single, active response ciiglit have intro-
duced desensitization elements in tlmt the active responses would have
involved inhihition of anxiety. VJolpe (1958) descrihes how an active
motor response can reduce anxiety, even though this motor response
does nothing toward the avoidance of the threatening stimulus. The
choice of rev;arding a p-assive C3 confrontation response avoided these
prohlecij; and thus more clearly offered a test cf motor level operant
conditioning versus Piore central (perhaps) autonomic level ajixiety
inhitition. If an active, positive, clearly defined competing motor
response had been positively reinforced then the operant treatment would
prooatly have measured out as more effective than was the case. But
again this does not correspond to the hypcthe-iized role of an operant
procedure in desensitization in that desensitization does not involve
the occurrence of an active, positive notcr response. In hujnan de-
sensitization, finger raising serves as a signal rather than as a coapet-
ing response antagonistic to the phoMc response. Thus, to make the
operant procedure most comparable to the operant procedure that might
operate in desensitization, no clearly defined rewarded response v/as
employed.

Results from the Extinction Phase suggest that contrary to
our hypothesis the operant treatment is not superior to the desensitization
procedure. However, and as hypothesized, both treatments were signifi-
caJitly effective as compared to a no treatment condition. Before dis-
cussing possible implications of these results it must be pointed out
that the measures of treatment effect were tal:en fiom extincticji trials

^5

that did not continue to completion. The question arises as to vrhethor
the treatrieut differences demonstrated would have continued in the
same form if the trials had gone on to total extinction. Statistically
this v;as controlled to an extent by regarding the "days" factor in the
analysis of variance as having a random rather than fixed effect,
thus taking the five days measured as a saniple rather than as exhausti"9e
of the universe of "Days." It is most likely that data obtained from
all possihle extinction trials v;ould have yielded the same results as
the data actually' obtained. The most probable exception to this is
that the desensitization treatment might have been shovrn to be sig~
nificantly superior to the operant procedure employed in terms of
reducing number of phobic responses made, since this v/as nearly so
with the data obtained.

A major finding in this study is that an operant procedure,
vhich most closely approximated such a possible procedure in desensiti-
zation, was found to be slightly less effective (though this was not
statistically significant) than a desensitization procedure in the
reduction of phobic behaviors. All other sources of treatment effect,
not specifically 3.nd uniquely a part of the tv;o treatment procedures
employed, were equivalent in the two treatments. Such sources of
treatment effect v/hich v;ere equivalent in both treatments include
(1) extinction possibilities, (2) change from conditioned aversive
qualities to conditioned adient qualities of the hot end cf the shuttle
box through food presentation in treatment, and (3) opportunities for
making various motor responses. The CS tone vas of different stimulus

46

value for the two treatncnt groups, serving as a discriminative stimulus
for food presentation in the opeKuit procedure while serving as a C3
for response states involved in eating (presumedly incompatible with
previously conditioned anxiety response states) in the desensitization
procedure. Ho'jever, these differences in the tone's value, primarily in
its value as a conditioned reinforcer, are differences inherent in the
differences ■bet^/een the tv?o treatment procedures. And v;hereas the CS
tone's value as a discriminative stiraulus was much reduced during the
extinction trials (\;hen the food schedule hecame ad lih)' this a^ain
more closely approxlms-ted any possible operant procedure involved
in desensitization. Therefore, the desensitization treatment effect
may involve some possible operant conditioning treatment effect, but
it is not dependent upon suc'i operant elements. At any rate this study
supports the viev? that either the operant or the desensitization pro-
cedure is similarly effective in reducing phobic behavior. This find-
ing has some implications for further elucidation of the desensitization
process and for treatment of phobic behavior in general.

In regard to the nature of the desensitization process, contrary
to the viev; initially offered here, it appears that the desensitization
treatment effect involves more than either rev;arded confrontation
with the phobic object or reinforced induction of competing motor
responses. The operant treatment contained these tvo elements plus the
additional advantage of a more efficient use of rev;ard in using fading
and shaping procedures designed to more strongly condition motor response;
than could be accomplished by the continuous and simultaneous "reward"
offered as part of the desensitization treatment. The additional

4?

trcatnent effect eleraent, which is necessarily a part of the desensiti-
zation treatment to a greater extent than in the operant treatment,
could be a deconditioning of anxiety previously conditioned to the
stimulus represented hy the phobic object. This deconditioning might
occur as a result of tv;o processes; the wealcening or reciprocal in-
hibition of conditioned bonds between ajixiety affect and the phobic
stimulus, and a simultaneous strengthening or reconditioning of bonds
between the pliobic stimulus and pleasant affect (a process v;hich could
also serve to reciprocally inhibit anxiety). This might involve a
response state substitution to the same stimulus via classical con-
ditioning.

But rego,rdlcss of such conjecture, it does appear that phobic
behavior can be treated by clianging motor beh?.vior in the direction of
approach rather then avoidance, without dealing with anxiety directly.
Phobic beh^.vior treatment is also enhanced 'by reducing phobic anxiety,
perhaps by a process involving the substitution of conditioned adient
affect for previously conditioned aversive affect.

There is currently a continuing debate between behavior modifi-
cation learning theorists and analytically oriented theorists regard-
ing the necessity of dealing with internal "causes" in treatment of
neurotic symptoms (Ullmann and Xrasner, I965). Belated to this issue
is the differing view of phobias held by psychoanalysts and behavior
modifiers. Psychoanalysis traditionally regards phobias as generated
from anxiety which is displaced onto the phobic object, whereby avoid-
ance of the phobic object v.'hile being fearful of that object becomes a

48

means by which the anxiety is kept in check. Analytic treatment of
phoMas is accoiiplished through uncovering the sourco of the anxiety
in terns of the unconscious dynamics and intrapsychic conflicts in-
volved. After this underlying material is brought into full cognitive
and emotional a.'ai-encss then the anxiety is dispelled and the phobia is
reduced. In contradistinction to this view, the behavior modification
approach v/ould regard phobias in terms of behavior that originates
from prior aversive conditioning vrhich is best treated by elimination cf
the phobic behavior or modification of verbal behavior. Thus Eysenck's
(i960) dictum: "Get rid of the symptom and you have eliminated the
neurosis" (p. 9)«

Wclpe, in argijing for th.e progressive deconditioning or in-
hibition of the phobic anxiety in treating phobic behavior stands some-
vhere between the analyst, vho would resolve the underlying conditions
vhich give rise to the displaced phobic anxiety, and the behavior
modifier, v;ho would change the phobic (motor level) behavior. The
differences in these three approaches to the treatment of phobic behavior
can then be regarded in terms of the level of the phobic experience at
which attempted chsj'^e is primarily focused. The results of this
research suggest thjit either motor level behavior or more central
levels of response states can be the priaary change target in effective
treatment of phobic behavior. However, it must be remembered th?.t our
definition of phobic behavior is based on a learning theory model arid
not on the psychoanalytic "phobia." There may well be different types
of phobias corresponding to the dj.fferin3 definitions of phobias found
in psychocmalysis and learning theory.

^9

Treatment of photic behavior may best be accoKplished by in--
ducing changes at all levels of experience and behavior. But in
recent years there has been increasing support for the viev; that be-
havior, however conveniently studied at isolated and restricted levels,
is very much unified.

Social psychologists, particularly the cognitive dissonance
researchers, have vrell demonstrated ho;? changes in overt behavior can
bring about internal changes in cognition, attitude, and affect. The
operant anicsal in this study v;ho v/as rev;arded for confrontir^g threat-
ening stimuli probably experienced a fear reaction v/hich likely under-
went some extinction so that changes in affect followed aotor level
changes. Lilcev:ise, the subject v.'hose fear v;a5 diminished v/as enabled
to mal^e subsequent cotor level changes. Trend (1925) fou^nd it laost
helpful in treating phobias to make his phobic patients expose thea-
selvss to previously avoided and painful situations. The results of
this study are consistent with a unified view of behavioral change
such that chajige at any level of experience has important implications
for subsequent changes at other levels of overt and covert behavior.

The method presented here, of equitably comparing operajit with
desensitization methods, vrould seem worthy of further extension and
exploration. Hopefully, with mors sophisticated measuring techniques
and more refined procedures, future research will be able to demonstrate
hovf much the treatment of phobic behavior depends on the resolution of
anxiety engendering deep-seated conflicts versus reduction in specific
conditioned anxiety versus change in motor level behiavior. It might
also be possible to more precisely determine bow much desensitization

50

depends on the iiihioition of general anxiety versus specific conditioned
anxiety, and to detexaine hov; much desensit illation involveb instnunental
conditioning of either approach hehaviors or competing motor responses.
Such research is particularly neededwith human suhjects.

Although the purpose of this study was not to determine the
compai^ative efficacy of the tv70 treatments employed, such an evaluation
could he accomplished v/ithin the experimental frsmev/ork here intro-
duced. In all fairness to the operant procedure, such a study vould
require tlis reinforcement of a clearly defined, active motor response
(competitive to the phohic response). In the same vein, food deprivation
conditions should continue throughout the extinction trials in any
such treatirtent coDiparison.

Since operant procedures are effective in bringing ahout con-
frontation with a, phohic stimulus, desensitization with humans could
perhaps he made more effective by maximally employing verhal reinforce-
in ent which could he made contingent upon the subject's self-report of
having imagined an item in the hierarchy of threatening stimulus situ-
ations. Therapist reinforcement of the subject's actual confrontations
with the phobic object could also be combined with desensitization in
an attempt to bring about changes at many levels of behavior and ex-
perience in eliminating human phobias.

EEFSRHICI

Bentler, P. An infant's phobia treated vith reciprocal inhibition

therapy. Journal of Child Psychology and Psychiatry, 19^2, 3,
185-189.

Breger, L. and McC-raugh, J. L. Critique and reformulation of "learnins
theory" approaches to psychotherapy and neurosis. Psycholo°:ical
Ml?Jili» l^'^S, 338-358.

Carlson, W. J. ai'd Black, A. H. Traiimatic avoidance learning: note
on the effect of response prevention during extinction.
P sy chol Of.i cal ?.qv o r t s , 1959, i, 409-'il2.

Cautela, J. The Pavlovian basis of reciDrocal inhibition therapy.
Conditioned Reflex, I966, 1, 293-330.

Cocke, a. Evaluation of the efficacy of components of reciprocal in-
hibition nsycho therapy. Journal of Abnormal Psycholog:/, 19^3,

Cooper, A., Furst, J. E. , aiid Bridger, Yi. H. A brief comentary o.n the
usefulnci.s of studying fears of snakes. Jquj\''3l_ 9f Abnoriiial
Psychology, I969, 2i, ^13-^14.

Delude, L. A. and Carlson, N. J. A test of the conservation of anxiety
and partial irreversibility hypotheses. Canadian Journal of
Psychology, 196'l, 16, 15-22.

J^scncI;:, H. J. (3d.) Behavior therapy and the neurosis. Oxford:
Perga-ncn Press, ~1960.

Preud, S. Turnings in the v;ay of psychoanalytic therapy. In Cj)Il£cted_
£.5R§I.5.» '''ol* 3» 392-^402. London: Hogarth Press, 1925'^

Gale, Diana, Sturafels, Gloria, and Gale, S. N. A coaparison of
reciprocal inhibition and experimental extinction in the
psychotherat)eutic process. Behavior Research and Therapy,
1966, 3, 209-219.

Goldstein, A. J. A comparison of tv;o methods of overcoaing avoi_dance
behavior, Doctorafdissertation, University of Jlorlda, 19?7»

Eays, \], I.. Statistics for psychologists. New York: Holt, Rinebart
and Uinr'ion, Vj'E^,

51

52

Hogan, R. and Kirchner, J. Implosive, electJc, verbal and Mblio therapy
in the treatraent of fears of snal^es. Behavior Research and
Therapy, I968, 6, I67-I7I.

HuBsian, Ac Behavior therapy using hypnoj;is. In ¥olpes J., Salter, A.
and rtayna, L. J. The conditioning therapies. New York: Holt,
Rinehart and Winston, l<^&r.

Johnson, S. and Sechrest, L. Comparison of desensitization and progres-
sive relaxation in treating test anxiety. Journal of Consult-
ing and Clinical Psycholofiiy, I968, ^, 280-2^1^/""

Jones, Mary C. A lahoratory study of fear: the case of Peter. Journal
of Genetics and Psychology, 192^, ^,308.

Lang, P. and La^ovik, A. The experimental desensitization of a phobia.
Journal of Aonoraal and Social Psychology/-, I963, 66, 5^9-5^5'

Lazams, A. Group therapy of phohic disorders hy systematic desensiti-
zation. Journal of Ahnorrial and SociaI_ Psychology, I96I, 6^,
50-'i-5io,

Lazaras, A. and Abranovitz, A. The use of "emotive imagery" in the
treatment of children's phobias. Journal of Mental Science,
1962, 108, 191-195.

Lazaras, A. The results of beha.vior therapy in 126 ca.ses of severe
neurosis. Behavior Research _and Therapy, I963, _1, 69"?9.

Orne, M. On the social psychology of the psychological experiment:

with particular reference to d?inand ch-a.ract eristics and their
implications. American Psychologist, I962, 17, 776-783,

Pavlov, I. P. Conditioned reflexes. London: Oxford University Press,
1927.

Paul, G. Insight versus desensitization in usycho therapy. Stanford:
Stanford Press, I966.

Rachman, S. The role of muscular relaxation in desensitization therapy.
Behavior Research and Therapy, I968, 6, I59-I66.

Rachnan, S. and Hodgson, R. Studies in desensitization: IV. Optimum
degree anxiety reduction. Behavior Research and Therapy,

1967, i, 249-250.

Rosenthal, R, E:;n3crimenter effects in behavioral research. New York:
Apploton-Ceritury-Crofts, 19^.

53

Solomon, R. L. and Vi'ynne, L. C. Traumatic avoidance learning: the

principles of anxiety conservation and partial irreverriitility.
Psychological Hevie\f, 195^ » ii.. 353-3^5 •

Solyom, L. and Miller, S. Reciprocal inhibition by aversion relief
in the treatment of Dhobias. Bchavi or Re s ear ch and The rapy ,
1967, i, 313-32^f.

Stampfl, T. and Levis, D. Essentials of inplosive therapy: A learn-
ing theory based psychodynamic behavioral therat/y. Journal of
AbnorEal Psycholo.^y, 19^7, 72, 49^-503.

Ullmann, L. P. and Krasnsr, L. (Ms.) Case s tud i e s i n _b e hav i or

jiodAficati.on. Nev/ York: Holt, Rinehart "and ■■'■'ins ton," i"965.

Watson, J. B. and Rayner, Rosalie. Conditioned emotional reactions.
Journal of rjq)eriaental Psychology, 1920, 3j 1-14.

V^einberger, N. H. Effects of detainment on extinction of avoidance

respons es . Journal of CoEipai-ative and Physiological Psychology,
1965, 60, 135-138.

White, J. The use of learniTig theory in the psychological treatnent
of children. Journal of Clinical Ptychologv, 1959 > ±^%
227-229. ~" — .

Winer, B. J. Statistical principles in erperimental design. Nev;
York: McGrraw-HiTr, i'9c2.

WoIdc, J. SroeriEental neurosis as learned behavior. British Journal
of Psycholog:--, 1952, 43 > 192-206.

Wolpe, J. Psychotherapy by reciprocal inhibit io.n. Stanford: Stanford
Press, I95B".

Wolpe, J. The systematic desensitization treatment of neuroses. Journal_
of Nervous and Mental Disease, I96I, 1^2, 139-203.

Wolpe, J. and Lazarus, A. Pehavior therapy techniques. Nev/ York:
Pergaraon Press, 19^7 ~^

Wolpin, M. and Raines, J. Visual imagery, expected roles, and extinction
as possible factors in reducing fear and avoidance behavior.
Behavior Research and Theraw, 1966, 4, 25-37.

J^PENDICSS

55

iPPE^IDIX A

EXTINCTION PrlASE: DAILY AVERAGE UTENCY RECORDED TO THE NEAREST

FUl'IDREDTH OE A SECOND AIID SU^-CISD ITOiBER (GIVSI PAREHTHETIC.lLLl)

OE AVOIDANCE RESPONSES MADE BY THE CONTROL GROUP

OF TEN TRLA.LS OVER FIVE DAYS

Days

Animals

1

2

3

4

5

1

10,85 (8)

11.32 (8)

9.01 (9)

12.38 (9)

5'S(^ (10)

2

4.^?. (10)

5-3 (10)

8.73 (9)

8.76 (9)

6.46 (9)

3

5.69 (9)

1.82 (10)

5.S8 (10)

3.43 (10)

1.94 (10)

k

7.58 (10)

9.72 (10)

10.40 (10)

9.74 (9)

7.88 (iO)

5

5.79 (9)

7.98 (10)

3.93 (10)

7.04 (10)

2.48 (10)

6

6.89 (9)

12.7^ (6)

12.50 (6)

8.53 (10)

12.22 (9)

7

9.75 (8)

14.01 (6)

6.41 (10)

12.89 (10)

^'S5 (9)

8

10.57 (8)

14.65 (8)

13.0? (5)

10.82 (8)

11.10 (6)

9

6.33 (10)

9.86 (10)

9.90 (9)

8.82 (9)

6.60 (10)

56

APPEi^IDIX B

EXTINCTION PHASEt DAILY AVE.RAGS UTENCY RECORDED TO TPJ) KE.1PJST

HUIIDREJJTK 0? A SECOND AIJD SUIIMSD NUMBER (GIVEN PARSJTHETICAI-LY)

OE AVOIDANCE REfiPONSES M'lDE BY TKpI OPERA^^T GROUP

OE TEN TRIALS OVER EIVE DAYS

Days

Animals

1

2

3

4

5

1

1?.33 (2)

12.33 (6)

11.25 (8)

16.23 (4)

10.80 (7)

2

10.68 (8)

9.62 (8)

9.1''^ (10)

11.24 (8)

6,60 (10)

3

10.11 (9)

12.07 (7)

6.40 (10)

11.36 (9)

8.26 (8)

^

18.70 (1)

15.01 (7)

8.76 (9)

9.81 (8)

11.3^ (8)

5

15.59 (5)

8.98 (8)

8.33 (9)

10.08 (8)

7.53 (10)

6

20.00 (0)

18.42 (3)

10.68 (?)

10.79 (7)

8.20 (7)

7

20.00 (0)

20.00 (0)

20.00 (0)

20.00 (0)

20.00 (0)

8

13.50 (7)

13.22 (7)

12.96 (7)

8.12 (10)

12.98 (5)

9

11. 2^^ (7)

13.90 (7)

13.41 (6)

13.19 (6)

14.69 (6)

57

APPE:roix c

EXTINCTION TRIALS; DAILY AVER/.GS LATENCY RECORDED TO TKE NEAREST

HUNDRIIDTH OE A SECOND AI'ID SiJK.iED NUI-IBER (GIVEN PAHEilTHETI CMXY )

OF AYOIDMCE R2^P0NSE5 MADE BY THE DiSKNSITIZATION

GROUP 0? TEN TRIALS OYER FIVE DAYS

Days

Animals

1

2

3

4

5

1

13.17 (4)

17.69 (2)

9.01 (9)

12.38 (7)

18.90 (1)

2

19.57 (1)

16.88 (8)

12.30 (6)

18.55 (2)

18.33 (1)

3

19.53 (1)

8.59 (7)

13.60 (6)

11.02 (7)

10.86 (7)

4

19.53 (1)

17.05 (4)

15.71 (5)

14.06 (5)

8.74 (C)

5

20.00 (0)

2O0OO (0)

12.36 (8)

19.30 (1)

13.10 (4)

6

20.00 (0)

19.40 (?.)

14.20 (5)

16.34- (4)

12.00 (6)

7

7.69 (9)

5.89 (10)

7.16 (10)

11.59 (8)

8.72 (9)

8

15. t9 (3)

20.00 (0)

17.53 (2)

11.12 (7)

15.00 (4)

9

9.61 (7)

8.00 (9)

5.88 (10)

8.27 (8)

14.69 (9)

BIOGRAPHICAL SKETCH

Arthur H. VJells, Jr., V7a«> "born on October 9, 1939, at Birming-
han, Alabacia. He mo^ed to Florida in 19^9, where he attended puhlic
schools. He v;as graduated from Ft. Lauderdale High Scliool in June,
1958. In August, 1962, he received the degree of Bachelor of Arts
froa the University of Florida. Dyeing the s^oa^er of I96I, he worked
with maladjusted children under David Wineman while attending the
University of Michigan. In August, I965, he received his Master of
Rehahilitation Counseling degree from the University of Florida.
After working for one year in MitM, Florida; as a Reh-ahilitation
Counselor, he returned to the University of Florida to work to\,'ard
his doctorate in p-ychology. During his graduate studies in prycholog;.',
he was a Vocaticn-il Rehahilitation Administration fellow. In Au^st,
1969, he completed a one year's internship in clinical psychology at
the J. Hillis Miller Health Center. Since Septemher, 1969, he has
worked as a psychologist at the Northeast Florida State Hospital in
Macclenny, Florida.

He was honoraoly discharged from service in the U. S. Air
Force.

Arthur M. Wells is married to the former Ar^stasia E. Lawlor.
He is a menher of Psi Chi a.nd Phi Kappa Fhi.

58

Thri.s discertation vas prepared under the direction of the
chairfflan of the candidate's supervisory comiiittee and has been approved
"by all inefflhers of tliat coa-'nittee. It v/as sutmittcd to the Dean of the
College of Arts and Sciences and to the Graduate Council, and uas
approved as partial fulfillaent of the requirements for the degree of
Doctor of Philosopiiy.

March, 19?0

Dean, College of Arts and Sciem;

Dean, Graduate School

Supervisory Coiniiitteej