THE USE OF A SYMBOLIC MODEL AND

VERBAL INTERVENTION IN INDUCING

AND REDUCING STRESS

Thomas E. Narut
 ACKNOWLEDG.W·lENTS

1'he author is grateful to Dr. James D. Lowe, Jr.,

research director, for his supervision, advice and en­
couragement during all phases of this dissertation. In
addition to unselfish planning and coordination, he also
provided support with firmness resulting in the incentive
to begin and complete this study. Gratitude is also ex­
pressed to the other members of the dissertation committee;
Dr. John c. Koeppel, Dr. L. Erl Hehearg, Dr. Rays. Mus­
grave, and Dr. Thomas Yarnell, for their. interest and
constructive connnents •
.jpecial appreciation is extended to Captain William
G. Cumming, Jr., H;jC, U;:;,N, Head of the Clinical Psychology
Section, Bureau of Medicine and .::»urgery, Department of the
Navy, Washington, D. C ., who provided the opportunity,
finances, research environment, and duty station3 to allow
completion of the experiment. The same appreciation is
extended to Dr. Bruce Becker, Director of Internship
Training, and the Psychology Clinic staff at the U. S.
.Naval Hospital, National Naval Medical Center, Bethesda,
Maryland, for use of the apparatus, supplies and especial­
ly the time and facilities to complete the pilot studies.
A special note of thanks go·es to Commander Carl H.

ii
,Wagner, MSC, USN, Special Assistant for Hedical and
Allied Sciences, Office DCNM, Department of the Navy,
who helped generate the idea, offered constructive guide­
lines about the experimental design and interpretation of
the results. Gratitude is also expressed to Dr. Ch".rles
D. Spielberger, Director, Doctoral Program in Clinical
Psychology, Florida State University, Tallahassee, Flor­
ida, for his comments, information, and permission to use
the �tate-Trait Anxiety Inventories in this dissertation.
The author is especially grateful to Captain James
H. Holmes, HG, USN ., Chief of the Psychiatric Service, U. s .
Naval Hospital, Portsmouth, Virginia, for his understanding
and cooperation in allowing the time to complete this
dissertation. Special thank:s are also expressed to the
HAd-:cal 3ervice Corps officers in the Portsmouth and Uor­
folk, Virginia, commands who assisted in allocating the
physical facilities and subjects necessary to complete
this research.
F;nally, very special thanks and appreciation are
expressed to my wife, Carole, and children, Julie, Tommy
and Cathy. Their patience ., understanding, support and
sacrifices made this dissertation a reality, for without
them my effortD would be meaningless. In addition, an
endearine thank you to my wife for her many hours of de­
dicated typing resultinc in this final completed disser­
tation.

iii
 TABL}!; UF c.; ON'l'ENTS

Page
TITLE PAGE • • • • • • • • • • • • • • • • • • i

ACKNOWLEDG�l�NTS . • • • • • • • • • • • • • • • ii

LIST OF TABL�S • • • • • • • • • • • • • • • • • V

LIST OF FIGURES . • • • • • • • • • • • • • • • vi
CHAPTER
I. INTRODUCTION • • • • • • • • • • • • • • 1
II. HETHOD . • � • • • • • • • • • • • • • • 24
Subjects . • • • • • • • • • • • • • • • 24
Stress .Stimuli • • • • • • • • • • • • • 26
Apparatus . • • • • • • • • • • • • • • 28
Experimenter . . • • • • • • • • • • • • 34
Procedure • • • • • • • • • • • • • • • 34
III. RESULTS • • • • • • • • • • • • • • • • 43
IV. DISCUSSION • • • • • • • • • • • • • • • 61
v. SUH11ARY • • • • • • • • • • • • • • • 77
APPENDICES • • • • • • • • • • • • • • • • • • • 80
REFERENCES . • • • • • • • • • • • • • • • • • • 91
VI:rA • • • • • • • • • • • • • • • • • • • • • 97

iv
 LIST OF TABLES

Table Page
1. Summary Table for Analysis of Variance
of Hatched Experimental and Control Groups
to the Independent Variable of the Task
over T�ials for the S+.ate Anxiety Scale
Response Heasures (Form X-1) •.••••.•••.•.•.•.• 47
2. Significance Levels for A Priori .i tests
for Matched Subjects Between the Experi­
mental and Control Group of State Anxiety
Sc ores .......................... .a ... • • • • • • • • • • • • • 48
3. Hatched _t_ test Comparisons Between Pre-
and Postlog GSR Baseline Levels of the
Experimental and Control Group •.•.•.•••...••.• 50
4. s,.nnmary Table for Analysis of Variance of
Hatched Experimental and Control Group to
the Independent Variable of Task over Trials
for GSR lo� Response Heasures ....•............ 56
5. Significance Levels for A Priori� tests for
Hatched .:)ubjects J3etween the Exp erimental and
Control Group for log GSR Neasures ..••••.... •. 57
6. Pearson Product-l·ioment Correlations Between
Spielberger State Anxiety Normalized .I-Scores
and log GbR Measures on Trials I, II, III, and
Total Trials. .. . . . . . . . . . .. . . . . .. . . . . . . . .. . . . . . 60

V
 LI8T UF FIGl.JllliS

it'igure Page
1. Group State Anxiety response means after
exposure to each accident for the matched
eXperimental and control group.•.•.•••••.•.•.• 45
2. GSR group means comparing matched EG and
CG over pre- and postbaseline and during
the entire experimental film .•.•.•.•.•••.•.•.• 51
3. GSR group means for each of the five second
intervals before and after the accident im­
pact comparing the EG with the CG, and es­
pecially the effect of the IV on the EG in
Accident II ....•.•.•.•.•.•.•.•.•...•.•.•.....• 53
4. GSR means of the seven 5-second intervals
computed into a total mean for each group
(EG and CG), over trials, also a comparison
of pre- and postbaselines ..•.•.•••.•.•.•••••.• 54
5. BSR group means comparing matched EG and CG
over pre- and postbaseline and the entire
experimental film. BSR is in Ohms·-·•·•·•·-·• 59

vi
 CHAPTER I
INTRODUCTION

Psychological stress and its control are widely

recognized as central problems in human life. Histor­
ically, confusion about the definition and measurement
of stress occurred because individual researchers limited
themselves to using either a physiological, behavioral
or self-report measure exclusively. These self-imposed
limitations on the measurements allowed only sing_le
aspects of the stre�s response to be researched as was
evidenced by the classic works of William James, using
awareness of bodily changes; Wi1ndt with the introspective
report; Lange with vasomotor responses; Cannon-Bard with
hypothalamic activity; SAlye with adrenocortical activity;
and Wenger using visceral activity.
The main problem of these self-imposed limitations
on the measurement of stress was the inconsistent findings
reported. Each experimenter developed what he felt to be
the best criterion for measurement which resulted in a
fragmented and incomplete knowledge of stress.
Prior to the 1950 1 s, there was little or no differ­
entiation made betwe�n physical stressors which attached
the biological aspects of man and psychological stressors
which affected man purely because of their psychological

1
 2

�significance. Selye 1 s (1956) classic work concerning the

General Adaptation Syndrome (GAS) made such a differ­
entiation. The GAS referred to the physiological process
that developed when a person was exposed to a stressful
situation. Selye assumed that the body reacted to stress
in three successive stages: an alarm reaction, a stage
of resistance, and finally a stage of exhaustion. Ac­
cording to Selye the stress stimuli developed from two
main sources: physical conditions that could directly
harm the body, or psychological problems such as impending
fear or anxiety.
Selye 1 s biological and psychological dichotomy of
stress was enthusiastically assimilated into the litera­
ture and accepted as the new direction for future research
to clarify and define measurement of the stress response.
However, two important measurement problems were still
unresolved. These were: (1) the lack of agreement among
the various physiological response measures, and (2) a
lack of agreement between verbal reports and physiological
response measures to stress.
The problem concerning the lack of agreement among
various physiological response measures of stress was
demonstrated by Lacey (1959) in a series of studies.
Lacey used the response measures of skin resistance,
heart and respiration rates under numerous conditions of
stress including electric shock and extreme cold. He
 3

.found the intercorrelations among the physiological res­

ponse measures were essentially zero. In a review, Martin
(1961) cited a number of stµdies which tended to support
Lacey's findings. Martin also predicted and demonstrated
that intercorrelations among physiological measures taken
even under conditions of rest would be low and not signif­
icant. However, Lacey and Lacey (1962) concluded that
these low intercorrelations may be a function of the in­
dividual's own response pattern to stress.
The concept of "response specificity 11 (Lacey, Bateman
& VanLehn, 1953; Lacey & Lacey, 1958) refers to an in­
dividual having a preferred channel of physiological
responses to stress while other channels might not be
used. The preferred channel varies from individual to
individual, but seems to be reliable across different
stressors and reproducible over time (Schnore, 1959;
Lacey & Lacey, 1962).
Both Lacey (1959) and Martin (1961) agreed and con­
cluded that the near zero intercorrelations among physio­
logical response measures to stress was a problem in
methodology. They suggested the main cause of inconsist­
ent results appeared to be the lack of a clear but
separate criterion between the measurement of what were
the effects of biological versus psychological stimuli
to stress. In addition, Lacey (1959) concluded individual
response specificity needed to be controlled before near
 4
accurate measurements of stress responses could be re­
corded.
The problem concerning the lack of agreement between
verbal reports and physiological response measures of
stress also had its share of inconsistent results. Several
studies indicated that no affective arousal was measured
with verbal reports, while in reality, high autonomic
activity was being recorded as the stressful stimulus
was presented (Lazarus, Speisman, Mordkoff & Davison,
1962). The lack of agreement between verbal reports
and physiological response measures to stress can be
traced to other factors rather than being purely physio­
logical or methodological. Verbal reports were parti­
cularly susceptible, among other things, to the operation
of defensive distortion on the part of the subject
(Lazarus & Alfert, 1964).
The defensiveness of subjects was a main factor con­
tributing to the inconsistent relationships between
verbal reports and physiological response measures to
stress. In fact, the literature concerning subject
defensiveness became so controversial that nwnerous
publications resulted. There were monographs (Lacey,
1956; Lacey, Smith & Green, 1964), textbooks (Levitt,
1967; Spielberger, 1966; and Lazarus, 1966), and even a
special conference relating to subject defensiveness in
verbally reporting psychological stress (Anp ley &
Trumbull, 1967). The results ot these publications were
numerous and broad, but they concluded that if the stress
stimuli could be controlled, then defensive distortion
on the part of the subject might be reduced.
Two conclusions were made from the research pre­
viously cited (Lacey, 1959; Martin, 1961) in reducing
problems relating to inconsistencies in physiological and
verbal response measures of psychological stress. First,
further knowledge was needed about the concept or indi­
vidual response specificity. Second, and most important,
the stress stimuli needed to be controlled.
The growing body of literature on attempting to
control psychological stress made its largest contributions
after World War II. Research issues focused upon creating
and measuring two types or situations which could be
labeled stressful and provide models f or study; "natural­
istic stress," and "laboratory" or "simulated stress."
The earliest widely lmown research on naturalistic
stress was the work by Slater (1943) on English soldiers
being evacuated from the beaches at Dunkirk during World
War II. Biographical data were completed on the back­
grounds or soldiers who broke down while being exposed to
physical and psychological stress when waiting to be
evacuated. Psychiatric casualties were grouped into three
categories: severe, moderate and minor. Slater concluded
that psychiatric casualties were worrying, moody, sexually
 6

inadequate men who also had members in the immediate

family with previous psychiatric disturbances. The more
of these factors in the soldiers background, the less
stress he was able to tolerate, and the longer the time
he spent hospitalized.
The results or Slater's work were.supported by other
studies after World War II. Research after the Korean
.conflict on communist defectors among United States
servicemen was investigated by Farber, Harlow and West
(1957). They found, through using verbal reports, service
records, questionnaires and intensive interviews, that
various degrees of stress (e.g., threat of death, loss of
rood, and social isolation) imposed upon individual ser­
vicemen resulted in changes or behavior leading to various
forms of collaboration with the communists. However, all
the men subjected to this stress did not yield. Some
reacted quite differently to the various degrees of stress
imposed on them. Further study of the men who collaborated
indicated they had similar case histories as the English
soldiers in Slater's study.
From the research of Slater (1943) and Farber et al.,
(1957) it was apparent that naturalistic psychological
stress research had distinct limitations. It was usually
spontaneous, ruling out systematic observation. Extrane­
ous variables could not be controlled, and careful mea­
surement of the response variables were inconsistent.
 1

To correct for the limitations incurred in using

naturalistic stress situations, laboratory analogues
were developed and referred to as simulated conditions
to study psychological stress. A major study in this
area was by Karle and Bialek (1958). They constructed
and developed a psychological stress scale called the
Subjective Stress Scale (SSS) designed specifically to
measure highly defined stress situations. The scale
was then used in a unique study by Berkun, Bialek, Kern
& Yogi (1962) who ingeniously simulated stress situations
(e.g., a forest fire, an accidental explosion, and the
ditching of an airplane). There were twenty-two service­
men who served as subjects and were unaware of the
simulated conditions. Each was free to react as he saw
fit, but their efforts were blocked by equipment break­
downs, darkness, and false reports of conditions. When
the stressful period was 0ver each subject was given an
intensive clinical interview, physiological and chemical
tests, and filled out the SSS. The results on the SSS
were not significant and did not correlate with the actual
behavior of the men under stress. There was no doubt that
stress was being experienced as eight of the subjects
physically withdrew from the situation without permission.
The study by Berlrun et al., (1962) demonstrated the
need for improvements in the methodology to develop
psychological stress. r� illustrated dTamatically the
 6

need for controlled stress stimuli and for homogeneity of

subject samples. In addition, it demonstrated that de­
ception in stress research was an uncontrollable factor
which appeared to interact with and compom1d the final
results.
A study by Mechanic (1962) aimed at meeting these
methodological needs resulted in a more controlled ex­
-periment on stress. His work used a defined population
of graduate students about to take a standardized written
examination. Pre- and postmeasures were taken using the
Taylor Manifest Anxiety Scale. Results indicated that
preexamination stress was significantly higher than post­
examination stress. From laboratory studies similar in
design (Lazarus, Speisman, Mordkorr & Davison, 1962) it
was observed and recorded that psychological stress in­
creased in anticipation of the actual stressful event.
The problem then became how much in advance did the
anticipation of stress occur, and how a more defined
psychological measure of stress could be developed. To
answer such questions, more control and precision were
needed in the methodology to study stresso
The naturalistic method had the advantages of
realism without deception, but the control or variables
and results were inconsistento The research or Lacey
et al., (1964) and Spielberger (1966) emphasized that
physiological measures or stress responses were desirable
 9

for control and accuracy because they were immediate and

could be statistically standardized for comparisons. The
naturalistic method was not acceptable because it imposed
difficulties in obtaining the necessary physiological
response measures of stress.
In contrast, the laboratory analogue had the advan­
tage of controlling event occurrence, choice of subjects,
and using physiological response measures. A disadvantage
was that it used deception to achieve results and many
times employed threat or assault to induce stress with
human subjects. This deception resulted in inconsistent
data (Martin, 1961)�
From a review of the stress literature it appeared
the most effective approach to producing, controlling
and studying the effects of psychological stress was a
combination of the advantages cited by both the natura­
listic and laboratory methods. The major concern seemed
to be the identification of an adequate stress stimulus
which would allow objective measurement, not be threat­
ening, and still be realistic enough to the subject.
The bridge between the naturalistic and laboratory
method to produce psychological stress appears to be in
the works of Lazarus et al., (1962) who pointed out the
justification and advantages of using a motion picture
film as a stim�lus to produce stress. The authors cited
the advantages as follows:
 10

First, deception is 1llll1ecessary since the

impact of the film is natural and appears to take
advantage of the human tendency to identify with
the character and their experiences in a dramatic
protrayal. Thus, the subject may be caught up by
the film contents inadvertently, obliterating the
need for procedural misrepresentation. Films are
also extremely versatile in that all manner of
interpersonal situat-ions which might serve as
stressors are enlarged to include any that are
conceptualized as potential stressors. Further­
more, since the experimentally constricting and
irritating deception is no longer needed, the
subject can be exposed time and time again in
repeated measurement designs to a variety of
different types of film stressors and nonstressors.
Thus it is possible to test the range of stimulus
contents which disturb given classes of subject's
efforts at adaptation. Finally, since the subject
merely views the film seated comfortably in an
easy chair, no physical assault is involved, and
no confounding exists between the physical impact
of a noxious stimulus and its psychologically
meditated properties (Lazarus et al., 1962, p. 3).
Earlier research supporting Lazarus on some of these
points were reported by Schwartz (1956), and Aas (1958).
Other studies, supporting the specific advantages cited
by Lazarus, were substantiated in later research by Alfert
(1964}; Goldstein, Jones, Clemens, Flagg and Alexander
(1965}; Folkins, Lawson, Opton and Lazarus {1968); Kindler
(1968); and Ka.men (1969).
The theory evolving the use of film for the labora­
tory analogue of a naturalistic psychological stress model
was based upon the pioneer work of Dysinger and Ruckrnick
(1933) who studied emotional responses in children. Child­
ren and adult heart rates were �easured as they watched
movie scenes depicting dangerous situations and romantic­
erotic displays. The findings indicated that scenes of
 11

,danger, conflict or tragedy elicited the greatest re­

actions from children, while the romantic-erotic scenes
produced the greatest reaction from adults. This early
study suggested the hypothesis that by using a movie
film, stressful situations could be produced and studied
without the threat of real harm to the.subject. For this
reason films then began to be used increasingly as stress­
rul stimuli to study psychological stress {e.g., Alexander,
Flagg, Foster, Clemens & Blahd, 1961; Lazarus et al.,
1962, Handlon, 1962; Speisman et al., 1964; Lazarus &
Alfert, 1964; Goldstein et al., 1965).
The implicit assumption underlying the theory be­
hind these studies was that a person's reaction to a
stressful film provides an indication of how he would
react to a more direct threatening situation. This
assumption was formulated into a hypothesis and researched
by Alfert {1966). To test the hypothesis a film on wood­
shop accidents titled "It Didn't Have to Happen," was used,
physiological measures (GSR, heart and respiration rates)
recorded, and a psychological response measure {Nowlis
Adjective Check List for Mood) administered. The res­
ponses to the stress situations in the film were compared
to a direct ·stress situation {threat of shock}.
Results indicated significant correlations between
physiological measures and affective arousal, as well as
some differences in responses to specific stimuli. These
 12

correlations between the film and direct threat can be

interpreted to mean that a stress response to a filmed
threat situation could be used to predict other stress
responses in the face of direct threat. Thus, Alfert•s
work appears to be supporting the interpretation of using
a stress movie as a symbolic model. Alfert•s work also
reported the highest rate of responding to have occurred
prior to the impact of the accident.
Research with this same movie film by Brinbaum
(1964), and Lazarus et al., (1962) supported Alfert•s
findings and also concluded that it was the potential for
anticipation of thrbat or impending harm, rather than the
confrontation with the accident itself, that produced the
stress reaction. In summary, Lazarus and Opton (1966)
concluded that many problems which were not possible for
laboratory study could now be accomplished using the
motion picture analogue of psychological stress, especially
issues of central importance in the clinical field which
were discussed, but not controlled in the laboratory.
The use of motion picture films as stimuli has im­
proved intercorrelations among physiological response
measures to stress in the laboratory. Film stressors
provide controlled research with a standardized, contin­
uous, repeatable stimulus which is specific and results
in minimal variance. Any significant variance may now
be attributable to what is going on within that particular
 13

,person. The motion picture film methodology, which in­

cludes a continuous recording of autonomic levels of
activity, lends itself ideally to intra-individual anal­
ysis of the correspondence between measures. The re­
search of Lazarus, Speisman and Mordkoff (1963) dem�n­
strated this by using a stress film and finding a high
degree of correspondence between the autonomic measures
of skin conductance and heart rate. The results of
Lazarus et al., (1962) tend to support and justify the
assumption that a single autonomic measure (the GSR)
could be used to infer general arousal, if the film
stressor was specific and defined.
The film 11 It Didn't Have to Happen," was used in a
study by Halamstrom, Opton and Lazarus (1966} to further
support use of the GSR as a single autonomic measure.
They used a modification in scoring heart rate by log­
arithmically increasing the baseline and testing the mean
value in a time interval above a criterion. The modifi­
cation was called "peak rate" and resulted in significantly
higher intercorrelations with skin conductance and also
with a third variable, the onset of the impact of the
accidents in the film. The authors felt justified in
concluding that the correlation between heart rate and
skin conductance reflects the stress reaction only if the
stressor is defined and controlled.
In a series of studies on stress associated with
 14

,sport parachuting, Fenz and Epstein (1967), and Epstein

(1967) obtained continuous recordings of skin conduct­
ance, heart and respiration rates from experienced and
novice parachutists during a defined sequence of events
leading up to and following a jump. The results of their
findings supported that peak stress occurred prior to,
rather than at the jump. Another finding was that skin
conductance alone was a reliable indicator of stress be­
cause of the advantage of the irrnnediate response. Epstein
supports the theory and interpretation that different
levels of inhibition are involved for each autonomic
measure based upon its feedback from the physiological
system involved. Both heart and respiration rates showed
a significant discharge after stress while skin conduct­
ance did not.
Epstein's (1967) interpretation was also supported
by Geer & Klein (1969) who attempted to determine whether
GSR and heart rate to "aversive" photographs contained
both orienting and emotional components. Forty female
subjects were shown the photographs projected upside down.
The data was interpreted to indicate that, in addition to
the orienting response, an emotional component was also
present in the photographs. Further analysis of their
results revealed that the GSR was the most responsive to
the content of the stress stimuli while heart rate was not.
Thus, it can be concluded on the basis of the studies
 15

•previously mentioned that movie films can be utilized as

symbolic models to produce stress. The film on indus­
trial accidents ("It Didn't Have to Happen") is most
consistent in results, and skin conductance alone appears
to be an adequate physiological measure of stress, pro­
viding the stress stimuli are defined and controlled.
The problem of an appropriate psychological measure
which correlates with physiological response measures of
stress was studied by �ordkoff (1964). He used a movie
film as a stressor in an experimental design using
physiological response measures. Subjects rated their
affective state on a scale anchored by one of three
adjectives representing the three primary dimensions of
Wundt 1 s tridimensional theory of feeling. Mordkoff found
that regardless of what dimension was rated, the self­
report ratings tended to correlate somewhat with the ups
and downs of the physiological response measures and the
pilot work done on stress scenes in the film.
Lazarus and Alfert (1964) attempted to improve on
Mordkoff 1 s procedure by using the Bryne Repression-Sensi­
tization Scale (BRS) as the psychological response measure.
They found that high deniers as measured on the BRS re­
fused to admit disturbances verbally, but revealed them
autonomically. However, low deniers were prone to ver­
bally report feeling disturbed while revealing less
autonomic activity.
 16

Other studies using psychological response measures

such as the Nowlis Adjective Check List, Minnesota Multi­
phasic Personality Inventory (Speisman et al., 1964), and
Semantic Differential (Lazarus et al., 1962} tended to
support the findings of Lazarus and Alfert. Mordkoff's
study utilized an ef'fective design, but the use of Wundt•s
tridimensional theory of feeling did not yield signifi­
cance with replications because of the vague rating
criterion of the scale.
The inconsistent results of Mordkoff' and others were
interpreted to be due to appraised defenses employed by
the somewhat ambigious content portrayed in the film
(puberty rites and circumsion by an Australian aborigines
tribe). To correct and control for appraised stress,
studies turned to using the industrial accident film "It
Didn't Have to Happen," because or its explicit and de­
fined content, ease of ability for identification, and
generalization to real life situations of stress. Using
this film and the self-report scales the correlations
between physiological and psychological stress measures
were slightly higher and predictable. However, it was
apparent that a more suitable self-report scale needed to
be developed as a psychological response measure for
stress.
The work of Spielberger and Gorsuch (1966) provided
an answer to part of this problem through the development
 17

of a scale to measure two kinds of constructs: State

Anxiety and Trait Anxiety. The development of the scale
was completed when the preliminary manual was published
and reliability and validity established for the scales
(Spielberger, Gorsuch & Lushene, 1967, 1968). In 1969
Spielberger, Gorsuch and Lushene completed comparative
norms for their scales.
Spielberger (1966) emphasized the need to distinguish
between anxiety conceptualized as a transitory state or
conditions of the organism as a relatively stable person­
ality trait. State Anxiety (A-�tate) consists of feelings
of apprehension and heightened autonomic-nervous system
activity that vary in intensity and fluctuate over time.
Trait Anxiety (A-Trait) refers to individual differences
in anxiety proneness, that is, to differentiate tendencies
among individuals to respond with different levels of A­
State in situations perceived as threatening. Persons high
in A-Trait are also more disposed to see certain types of
situations as dangerous, particularly situations that in­
volve failure or some threat to the person's self-esteem
(�pielberger & Smith, 1966).
According to the scale developers, the A-Trait Scale
might be used as a research tool to select people who vary
in their di�position to respond to psychological stress
with different levels of state intenLlity. According to
this, one could then use the A-State Scale to determine
 18

.the actual levels of A-State intensity induced by an ex­

perimental procedure.
The State-Trait Anxiety theory formulated by Cattell
and Scheirer (1958, 1961), and Spielberger (1966), lends
itself to be used as an ideal psychological self-reuort
measure of stress produced by a movie film. In essence,
The A-Trait Scale could be compared to a GSR baseline,
that is, it refers to a person's individual differences
in anxiety proneness. It is specific and unique to that
person, much in the same way that Lacey {1958), Schnore
(1959), and Lacey and Lacey (1962), describe autonomic
individual response specificity.
It would appear that by using the movie film "It
Didn't Have to Happen," as the stress stimuli and skin
conductance as the physiological response measure, an
experiment could be designed which would identify high
responders to stress. This could be done by selecting
people who have high raw scores on the Trait Anxiety
Scale. If this hypothesis were supported, it would then
follow that measures of State Anxiety should correlate
with changes in skin conductance prior to the impact of
the accident, if the State Scale could be administered.
Support for this hypothesis is grounded in Spielberger's
theory of anxiety which states that people with high
Trait Anxiety are prone to display high State Anxiety
in certain stressful situations {Spielberger et al.,
 19

1969). Clearly, the movie .film, "It Didn I t Have to

Happen," appears to be a well defined and controlled
stress-producing .film. Should the hypothesis as de­
veloped be supported, then it would be one of the .few
measures of psychological stress to obtain agreement
between physiological and psychological response measures.
This type of agreement would further support and contri­
bute to the reliability and validity o.f using movie films
as symbolic models to further study psychological stress.
A current problem which has received attention has
been the experimental reduction of psychological stress
by using a movie .film as a symbolic model. Several
studies have demonstrated how induced changes in the
appraisal of a threatening event may "short circuit" or
reduce the stress response. Denial and intellectuali­
zation narratives employed in sound tracks (Speisman et al.,
1964), and introductory statements (Lazarus & Al.fart, 1964),
to a .film of primitive subincision tribal rites success­
fully reduced stress responses to the filmed threats.
Lazarus, Opton, Nomikos and Rankin (1965) have demon­
strated this principle of short-circuiting with similar
narrative orientations using the film, "It Didn't Have to
Happen. 11 It appears that reappraisal of' the stimulus
events through different narrations prior to the intro­
duction of the .film seems to be an ef�ective way to deal
with stress, but it is only one of many coping processes
 20

available (Lazarus, 1966). A study by Riess (1964) de­

monstrated that even supportive therapy before viewing
a stressful film could "short-circuit" or reduce the
stressful content or the film, and even more recently, it
was found that psychother.apy could be used as a technique
to reduce stress induced by a movie film (Bandura, 1967;
Spielberger, 1964; Folkins, Evans, Opton, & Lazarus, 1968).
Systematic desensitization techniques as espoused by Wolpe
and Lazarus {1966) have also reported successes by using
movie films as symbolic models to aid in reducing stress.
In a recent study, Folkins, Lawson, Opton and Lazarus
(1968) focused more intently on using systematic desensi­
tization and two or its components to test the ability of
reducing stress responses in a laboratory threat situation.
The subjects received one ot three types of training pro­
cedures before exposure to the stressful film "It Didn't
Have to Happen." The training procedures used were an
analogue of therapeutic desensitization, relaxation,
cognitive rehearsal, and a no training control. Stress­
ful responses to the film were measured by concurrent
self-report, heart rate, and skin conductance.
The results showed that the separate components of
therapeutic desensitization, relaxation and cognitive
rehearsal were as effective alone, as when combined in
the complete systematic desensitization program and, of
the three, cognitive rehearsal was the.most effective.
 21

,Making the assumption this experiment had some relevance

to the naturalistic theraputic concept, then the findings
highlight and emphasize the role of cognitive processes
in stress reduction. Furthermore, the use and application
of a movie film as a symbolic model to produce stan�ardized
psychological stress could result in a better understanding
of the dynamics of cognitive processes during stress.
Based upon the results of the review of literature
on the experimental reduction of stress, it is hypothe­
sized that a simple task such as asking a person a
question during a stressful period could result in re­
ducing stress significantly, to the point where it could
lend further support to a mediational process being in­
volved in reducing stress. It is also hypothesized that
this mediated process could be transferred to another
stress period without intervention from any external
sources, or more simply stated, the person would "learn"
a mediated process to help him control stress the next
time it occurs.
In summary, the purpose of this research proposal
was to construct an experimental movie film that could be
used to create and reproduce defined intervals of antici­
patory psychological stress. The main purpose was to use
this experimental film as a symbolic model in an attempt
to develop an improved methodology, induce and reduce
psychological stress, and objectively study stress and
 22

its correlates. The experimental film was also used as

a stress model to investigate whether GSR and State Anxiety
response measures could correlate during specified inter­
vals of stress. Specifically, it was designed to evaluate
if GSR responses could be reduced by the experimenter
initiating a verbal task during the onset or anticipatory
stress before the second accident impact, and to observe
if GSR decreases were supported by similar dec.reases in
State Anxiety. Specifically the following hypotheses were
formulated and tested:
1. The GSR response measures will increase during
specified time intervals near the impact of the
accident f-or each stress scene. Post scene
A-State scores will increase and correlate with
GSR response measures for each trial within the
Experimental and Control Groups. This corre­
lation will indicate a positive relationship
between the physiological and psychological
response measures to stress.
2. The Experimental Film will qualify as a valid
symbolic model based upon relationships among
three variables during each trial with the
Experimental and Control Groups. These vari­
ables will be the accidents in the film, GSR
response measures, and A-State scores.
3. During Trial I (Scene I), there will be no
significant difference or response measures
(GSR and A-State) between the Experimental and
Control Group.
4. During Trial II (Scene II), psychological stress
as induced by the symbolic model will be reduced
wi�hin the Experimental Group by intervention
of a task in a defined time interval during the
onset of anticipatory stress. This reduction
of stress will be reflected by decreases in GSR
response measures and A-State scores within the
Experimental Group as compared to the control
Group.
 23

5. During Trial III (Scene III), there will be de­

creases in GSR and A-State response measures in
the Experimental Group when compared to th e
Control Group. This decrease will suggest that
some mediational process may be functioning
resulting in a generalization of stress reduc­
tion from Trial II, to Trial III.
6. There will be a total significant difference
between the Experimental and Control Group re­
sponse measures, suggesting the task could have
an effect in reducing psychological stress.
7. Subjects who respond to questionnaires at the
end or the study with unusual expectations will
lack consistent relationships between GSR and
A-State response measures throughout the ex­
periment.
8. The Basal Skin Resistance (BSR) for all subjects
in the Experimental and Control Group will not
vary significantly over trials. This stabili­
zation of the BSR will tend to support that any
significant changes in GSR response measures
would be a function of the symbol model content
and not merely a shifting or autonomic response
levels within the subject.
 CHAPTER II
METHOD

Pilot Studies
Prior to the actual experiment, te�tative procedu£�s
were formulated and f'ifteen subjects (Ss) participated
in three pilot experiments to answer such questions as:
What was the optimum criterion for the selection of Ss;
What film would serve best as a psychological stressor;
Where would be the most significant point in the film to
introduce the mediational task; What type of' task would
be the most effective independent variable (IV); Would
the Galvanic Skin Response (GSR) be an adequate dependent
variable (DV); and, how would the response be defined?
No statistical procedures were used, as the purpose of'
these pilot experiments was to answer basic procedural
and apparatus questions.

Subjects
Two hundred and thirty-seven male enlisted naval
personnel, who were volunteers f'rom surrounding naval
facilities in Norf'olk, Virginia, agreed to be tested for
possible selection as Ss. From this group thirty-f'our
were selected based upon the following criteria:
1) Sex; as Hartin (1961) sunnnarized findings to
substantiate differences in GSR measures between males

24
 25

,and females. These sex differences were supported in

pilot work and to rule out such differences only male Ss
were used.
2) Race; as res�lts by Dreger and Miller (1968) in­
dicated racial differences in GSR response measures,
therefore, only American born caucasian males were used
in order to simplify and standardize the matching of Ss
within the EG and CGs. Only right handed Ss were used
due to physical limitations imposed by the apparatus.
3) Age; Ss between the ages of 18 through 24 were
selected as research using the GSR supported this age
group had demonstrated optimal GSR reactivity, and com­
parison norms were available (Kaiser & Haessler, 1968;
Kimmel, 1968).
4) Trait Anxiety Inventory Scores (STAI A-Trait
Form, X-2); Ss with raw scores near the mean of thirty­
six were selected. According to the authors of the scale
(Spielberger, Gorsuch & Lushene, 1969) this appeared to
be the optimum point to select Ss who responded with
different levels of state anxiety in situations perceived
as threatening (OtNeil, Spielberger & Hansen, 1969).
5) Interview; five minute patterned interviews were
used to rule out Ss who were overtly phobic, unstable
or had witnessed or experienced accidents similar to those
portrayed in the experimental film {Krause, 1961).
The criteria for selecting Ss provided a well defined
 26

homogeneous sample to constitute two separate matched

groups of sixteen each. These Ss ranged in age from 18
through 24 years with a mean age of 20.88 years and a
standard deviation of 1.92 years. The raw scores on
the Trait Anxiety Inventory ranged from 32 to 39, with
a mean of 35.43 and a standard deviation of 1.44.
Thirty-four �s were available at the beginning of
the study, but only thirty-two were used. The other
two were available as alternates in case of procedural
errors or misinterpretation of the instructions. One
pair of Ss had to be eliminated because of a broken
splice in the experimental film and an alternate matched
pair was substituted. All thirty-four Ss were given com­
pensatory time off from work for their participation in
the experiment.

Stress Stimuli
Original Film. The original W1modified stress
stimuli were a 12 minute and 49 second black and white,
sound, 16mm industrial safety film, "It Didn't Have to
Happen ., " purchased from the International Film Bureau of
Chicago, Illinois (Order number 2-lFB-72).
The fil.m portrays three wood shop accidents. In the
first accident a worker lacerates the tips of his fingers
on a motorized planner and the scene focuses on the
bleeding. In the second accident another worker manages
 27

,to amputate two joints of his middle finger in a milling

machine with dull blades and the scene focuses on the
stump spurting and dripping blood. The third accident
occurs because a careless worker allows a circular saw
to shoot a board through the air and drive it throu3-�
the abdomen of a fellow worker and the scene focuses on
the innocent victim bleeding, writhing and finally dying
on the floor. These three movie scenes have previously
been shown to produce rises in both physiological and
subjective indicators of stress reactions {alfert, 1964;
1966; Brinbaum, 1964; Lazarus, Opton, Nomikos & Rankin,
1965; Folkins, Lawson, Opton & Lazarus, 1968; and, Pilot
studies by author, 1970).
Experimental Film. For the experiment the original
version of the film was spliced into a near-continuous
psychological stress film using a procedure described by
Mordkoff (1964) and used as accepted standard procedures
for stress model research by Lazarus, Tomita, Opton and
Kodawa (1965). In addition to splicing, other modifica­
tions were made. The sound track on the film was elimi­
nated to standardize conditions for introducing the IV,
as studies by Lazarus and Alfert (1964), and Kamin (1969),
reported rises in physiological reactions as measured by
the GSR in other stress films without the sound track.
The film was spliced into three accident scenes (Scene I,
II, and III) to allow adequate time for �s to evaluate
 28

and record their feelings after each scene on the stress

scale.
The completed near-continuous psychological stress
film was referred to as the Experimental Film {EF) and
was fifteen minutes in length: 9 minutes viewing time

(Scene I, II, and III), and 6 minutes evaluation time

{Blank I, II, and III). The sequence of the film was as
�ollows:
TRIAL I {T-I); start stress scene I, (3 minutes)
followed by blank lead I (2 minutes) to evaluate scene
I, followed by:
TRIAL II (T-II); start stress scene II, (3 minutes)
followed by blank lead II (2 minutes) to evaluate scene
II, followed by:
TRIAL III (T-III); start stress scene III� (3
minutes) followed by blank lead III (2 minutes) to eva­
luate scene III, followed by the end of the film.
Pilot studies by the experimenter indicated this
sequence was adequate for the experimental procedureo

Apparatus
This study was conducted in a sound proof 9 X 12
room with a constant temperature range between 68° and
75 Fahrenheit and humidity control between 40 and 60%
allowing for standardization and control of the apparatuso
The experimental apparatus consisted of a 16mm movie
 29

_projector which projected the Efi' to a screen size of 3 X 4

for visual stimulus standardization (Reynolds, 1968).
All instructions for the experiment, including interven­
tion of a verbal task, were prerecorded and played for
each� on a tape recorder to standardize and reduce ex­
perimenter bias (McGuigan, 196)).
Physiological Measurement. A Bio-Physical Research
Instrwnents Model 201, Galvanic Skin Reflex Amplifier was
used to obtain Galvanic Skin Reflex (GSR) and Basal Skin
Reflex (BSR). GSR and BSR skin resistance were measured
by passing a 20 micro-ampere current through coated elect­
rodes attached to the palmar surface of the middle segment
of the left middle finger and the lateral surface of the
left forearm.
GSR and BSR outputs were measured and recorded on a
Texas Instruments Rectilinear Recording Milliammeter. Two
ink line tracings were made; GSR and BSR along with event
markings, and paper speed was three inches per minute for
all �s. To facilitate and coordinate the inked outputs
of both GSR and BStt with the EF, the event marker was
activated by the� to mark off and define the following
intervals:
Adaptation Period: a two minute interval for sta­
bilization and calibration of the GSR recorder pen near
the zero range.
Test Period: a two minute interval to allow each S
 30

.to complete the 3tate Anxiety Inventory ( STAI A-0tate It,orm

X-1) in order to determine his psychological baseline.
Pre-GSR Rest Period: the start and end of a three
minute rest period prior to starting the EF (Pre-GSH Base­
line).
Experimental Period: the start and end of the EF
including all stress and evaluation scene intervals.
Post-GdR Hest Period: the start and end of a three
minute rest period after the EF lPost-G�R Baseline).
Skin resistance measurements yielded two dependent
variables: GSR or the 11 phasic" response referring to the
transient, orienting reaction that immediately follows
a stimulus event, and the B3R or "tonic" response which
refers to the continuous, background level of ongoing
autonomic activity which is not time-locked to the specific
stimulus.
Both the G�R and B�R measures were obtained and re­
corded by using a resistance measure. In the measurement
of G0R and B0R, a correction or reciprocal conversion to
conductance is usually applied to interpret the data.
Since the Bio-P.hy.sical Research Instruments l\lodel 201,

Galvanic Skin Reflex Amplifier uses a two condenser­

coupled system of constant voltage and current, there was
no need to express the change in skin resistance into con­
ductance units. The two condenser-coupled system results
in resistance changes being converted directly into
 31

fractional changes of conductance making it possible to

use either kilohrns for resistance or micromhos for con­
ductance units. The kilohms scale was selected as it
results in more manageable data and a minimum of decimals
are involved in statistical computation (Hare, 1968).
Furthermore, kilohms are now being used in the more current
GSR literature for ease of standardization and comparison
,(Ka.men, 1969).
Since GSR measures were recorded and inked in skin
resistance units (SR), the variable GSH was defined as the
mean SR changes increasing from a zero baseline for each
� during a specified 30 second interval during each stress
scene, twenty seconds before the impact of the accident
and ten seconds after that impact (criterion of a response).
This criterion was defined by pilot work and appeared most
realistic because the major increase of stress occurred
prior to the impact of each accident.
The BSR was the SR established for each S over the
complete experimental film. Any deviation above the zero
baseline exceeding 200 ohms indicates, according to the
manufacturer, that there was a change of autonomic level
within the s, or a malfunction within the circuitry of the
amplifier or. recording pens. A stabilized BSR was inter­
preted to mean that changes in the SH measures were phasic
to the particular stimulus being presented to the S and
were not dependent upon or a part of a fluctuating
 32

autonomic level within each�-

Psychological Measurement. The State-Trait Anxiety
Inventory (STaI) is composed of two separate self-report
scales for measuring two distinct anxiety concepts, Trait
Anxiety (A-Trait, Form X-2) and State Anxiety (A-State,
Form X-1). Both scales were developed by Charles Spiel­
berger, Richard L. Gorsuch and Robert E. Lushene (1969)
and published by Consulting Psychologist Press, 577
College Avenue, Palo Alto, California.
The A-Trait Scale consists of twenty statements that
ask people to describe how they �nerally feel by marking
one of four choice&: (1) Almost never, (2) Sometimes,
(3) Often, and (4) Almost always. The (X-2) form was de­
signed to measure Trait Anxiety which refers to relatively
stable individual differences in anxiety proneness; that
is, to differences between people in the tendency to res­
pond to situations perceived as threatening with elevations
in A-State anxiety.
The A-State Scale (Form X-1) also consists of twenty
statements followed by marking one of four choices: (1)
Not at all, (2) Somewhat, (3) Moderately so, and (4) Very
much so; but the instructions require Ss to indicate how
they feel at� particula� moment in time. This form
(X-1) was designed to measure State Anxiety which refers
to feelings of apprehension and heightened autonomic ner­
vous system activity that varies in intensity and
 33

,fluctuates over time. Copies of both scales (X-2 and X-1)

may be found in Appendix A and B.
The Trait-State Anxiety Scale, Forms X-2 and X-1,
were scored according to the instructions set forth by
Spielberger, Gorsuch and Lushene (1969) under test scoring
and interpretation in their manual. The range of possible
raw scores for each form varies from a minimum of twenty
to a maximum of eighty points. Each of the twenty items
has a value from one to four. The point value for each
item is added to give the total raw score which is trans­
ferred to a table of norms for the appropriate age group
(18 through 24 years of age, the equivalent norm for high
school graduates and entering college freshmen). Tables
are provided for converting raw scores into normalized
T-Scores allowing interpretation and comparison of results.
The Trait Anxiety Scale was used prior to the ex­
periment only to select Ss. The State Anxiety Scale was
administered four times for each S; the first was to de­
termine the degree of stress of being wired to the appa­
ratus and to determine a psychological baseline for com­
parison with the same scale administered after Trial I,
II, and III. The result of the State Anxiety trials were
computed and correlated with the corresponding GSR response
measure (the mean of each 30 second impact interval) for
each trial.
.Experimenter
In a study by McGuigan (1963) it was reported that
more than one (E) during the course of an experiment
could be an uncontrolled variable and effect final results.
Therefore, only one male_!, the author or this paper,
conducted the experimento He was thirty-six years old
and a candidate tor a doctoral degree in psychology at
the University of Southern Mississippi, while on active
duty as a clinical psychologist at the Naval Hospital in
Portsmouth, Virginia.
The! administered the stress scales, selected the
Ss, read directions� connected and disconnected leads
attached to the.§. from the apparatus, showed the experi­
mental film and answered questions. The E and apparatus
were located within the same room, but behind the.§_ to
avoid distraction and focus attention to the experimental
filmo This arrangement of apparatus was found not to
effect the results and was supported in a personal com­
munication by Kline (1970), from a series of ongoing
studies.

Procedure
The §_s were selected by going into three groups of
one hundred enlisted naval personnel and asking for male
volunteers to participate in an experimento Two hundred
and thirty-seven agreed and were read the following:
 35

We are attempting to evaluate people's re­

actions to a movie film using an experimental
technique which utilizes both physiological and
psychological measures. The total time involve­
ment in the research will be about 40 to 60 min­
utes. The first part consists of answering the
inventory being passed out to you. Please put
your name and age at the top of the page. The
purpose for this information is to sort you into
groups. All information will be used for research
and treated as group data. All individual infor­
mation will be kept confidential. Now, please
read the directions silently as I read them to you.
Please do not start on the inventory until I finish
reading the directions.
"DIRECTIONS: A number of statements which
people have used to describe themselves are given
below. Read each statement and then blacken in
the appropriate circle to the right of the state­
ment to indicate how you generally feel. There
are no right or wrong answers. Do not spend too
much time on any one statement but give the answer
which seems to describe how you generally feel."
.Please start.
(Pause to allow time to complete form.)
Thank you. Now, those of you who would like
to participate further in this study please indi­
cate so by writing yes or no at the bottom of your
completed inventory and then pass it to the front
of the room.
I will return next week and ask certain people
who agree to come for a five minute interview.
During this interview they will be given further
information and I will answer what questions I can.
Upon final completion of the study I will be avail­
able to discuss all aspects with those who may be
interested. Thank you again for your time.
The ;nventory administered was the STAI, A-Trait
Scale (Form X-2) and Ss were selected for the interview
on the basis of the criteria (1) through (4) described
previously under "Subjects." Wb.An the interviews were
completed thirty-two Ss were selected f'pr the experiment
 36

by using A-Trait raw scores and age as the matching

criteria. Each S was matched with another S of the same
age and a similar A-Trait raw score. This resulted in
sixteen pairs of matched .§.s, which formed two groups.

-
Each s was individually coded, to later identify the
matched .§_s, and then randomly assigned.a number; one
through thirty-two. The Ss were ranked in the one

.through thirty-two ordero Next, the Table of Random

Numbers was entered and sixteen �s were selected to form
Group I, while his matched counterpart formed Group II.
The �s were again randomized within their respective groups
and an unbiased coin flipped to d�signate them and CG.
The order of running Ss through the experimental procedure
was based upon the one to sixteen numerical sequence and
each.§. was not aware to which group he belonged. The :EG
and CG were alternated, one.§. at a time for counterbalancing
and each§_ assigned a time to undergo the experimental
procedure.
The Ss were told a film would be shown during which
physiological measurements of the skin would be taken,
and that during pauses in the film they would be asked to
fill out an inventory. The §_s were assured there would be
no surprises or shocks and that specific instructions
would be given concerning the procedure when they returned.
Before each s arrived at the experimental room ror his
scheduled time, all apparatus and instrumentation were warmed
 37

up for at least five minutes, and maintenance and cali-

'bration checks made prior to and after each §_s partici­
pationo Before starting, the general procedure was
explained again and the s reassured that there would be
no surpriseso
The GSR electrodes were fastened to the S before
entering the experimental room allowing time for hydrationo
At the same time an attempt was made to establish rapport
with informal conversation to find out his expectations
concerning the GSR apparatus, EF and the procedure (sug­
gested by Hinton, 1970). The silver electrodes were cleaned
with alcohol and sapded with jewelers rouge. The electrode
sites were washed in cool water, blotted dry, and wiped
with acetone moistened gauze. A commercial electrolytic
paste was applied to a depth of at least 1/16 inch at each
site. The finger electrode was approximately 2.ocm2 and
the arm ground electrode was 3" X 4•" Each electrode was
held in place by masking tape and an elastic bandage. This
procedure, recommended by the manufacturer, was found to
give best results in the pilot studies, and was consistent
with the recent literature (Montague & Coles, 1966; Venables
& Martin, 1967; Hare, 1968)0
After the electrodes were fastened, the ,2 was led
into the experimental room and the jack plugged into the
GSR amplifier. The.§. sat in a comfortable chair and was
allowed a two minute Adaptation Period to the apparatus
 38

�(Krause, 1961; Martin, 1961). The left armrest had a

clipboard attached with a built-in constant light source
focused on the printed instructions and scales. This
arrangment allowed almost effortless completion of the
State Anxiety scales by the �s.
Next, the two minute Test Period corrnnenced and the
S was instructed to complete the first State Anxiety
Scale to determine his reaction to the apparatus and
obtain a psychological baseline. As the� completed the
State Anxiety Scale minor adjustments of the apparatus
were made for zero baseline of the GSR and BSH measures.
At this point a trial was explained to familiarize the s
with the time interval sequence when completing the A­
State Scales. The� answered only questions relating to
the� understanding the procedure, but not about the film
content.
During the next three minutes (Pre-GSR Rest Period),
the S was instructed to relax without falling asleep. He
was told to fix himself in the chair comfortably, and a­
void moving the left arm and finger to which the elect­
rodes were fastened. The lights in the room were dimmed
while the movie projector ran, but no stimulus image was
projected. This allowed further adaptation of the S to
the apparatus and provided a stable time interval where
ohmic range settings of the G�R could be adjusted for
each s. At the end of this time interval the �'s GSR
 39

measures were recorded along with all pertinent data

necessary for comparison with the Post-GSR Rest Period
taken at the end of the experiment.
On the clipboard in front of the S was a packet of
printed material containing the instructions for the ex­
periment and three copies of the A-State Scale (Form X-1).
The� was asked to read the instructions silently along
with a tape recorder.
You have volunteered to participate in a
psychological study in which you will be shown
a 15 minute silent, black and white movie film
about accidents with woodworking. During the
movie physiological measurements of your skin
will be made. Two things are very important:
lt'irst, no matter what the content of' the film
try to identify with what is going on. Some of
the scenes you might find unpleasant, but keep
watching them and giving them your full attention
as during one of the scenes you will be asked a
verbal question about that scene. When asked the
question please give the answer verbally so the
researcher can record it. Please do not move or
turn your head, just continue to watch the film.
The question will not be repeated, so be alert,
pay attention to the film and you should be able
to answer it. 5econd, it is important you remain
seated comfortably and not move your left arm;
this is asked to insure proper recording of the
results.
Any questions? (Pause).••••••••••••••••••.
Remember, watch and try to identify yourself
with the film, no matter what the scene content,
try to watch it, do not move your lef't arm, and
rill out your questionnaires during each blank in­
terval following the scenes. There will be three
scenes, three blank intervals and three question­
naires to complete. Now( do you have any questions
before we start? (Pause) .••••••••••••••••••••••
It was reemphasized that once the film started no
 40

_questions would be answered. The E switched on the movie

projector light which automatically started the film and
activated the event marker starting the Exp erimental
Period.
E;xperimental Groun (�G); followed the procedure and
directions just described, but during Scene II, the ver­
bal task (IV) was introduced (How many objects do you see
in this worker's top right pocket?). The start, length
and interval of this period was important and began ex­
actly 16 seconds before the impact of the accident. 1'he
verbal task was prerecorded on tape and synchronized with
the accident impact in Scene II. The start was signaled
by a splice in the film to the E who activated the tape
recorder. The 16 second prerecorded tape was synchronized
with the film splice as follows: Heaction time of� and
tape recorder to sound 2 seconds, question starts and
ends ('!'ask} 4 seconds, and � answering period 10 seconds
lresponse interval from the end of the question to the
impact of the accident}. These time intervals were de­
rived from pilot work on the experimental film and based
upon the graphs of studies by Lazarus and Opton (1967).
Control Groun {GG); essentially followed the same
procedure and directions described for the (�G), but did
not receive any task (IV) during any scene.
As part of the procedure, both the �G and CG were
administered the A-State Scale (.r'orm X-1) during each of
 41

.the three blank lead intervals (I, II, & IIIJ following
each stress scene. After each stress scene ended (I, II,
& III), the instructions from the tape recorder were as
follows:
In front of you there is a questionnaire
which I would like for you to fill out. Please,
do !!Q1 fill in the information at·the top, but
read the directions with me silently as they
are read to you.
"DIREC'l1ION::3: A nmnber of statements which
people have used to describe themselves are given
below. Read each statement and then blacken in
the appropriate circle to the right of the state­
ment to indicate how you feel right now, that is,
at this moment. �here are no right or wrong
answers. bo not spend too much time on any one
statement but give the answer which seems to des­
cribe your present feelings best."
When the scale is completed it will be picked
up by the researcher and the next scene will be
shown. This procedure will be repeated for Blank
interval II and III.
When each S finished the final scale after Scene III,
the Experimental Period ended. The final three minute
time interval lPost-GSR Hest Period) commenced and re­
laxation instructions were given identical to those for
the Pre-GSR Hest Period. The post baseline period allowed
for stabilization of the GSR for pre- and postfilm com­
parison. At the end of this period the£ was disconnected
from the apparatus and asked to fill out a questionnaire
(Appendix CJ as suggested by Lazarus, .'.)piesman, Hordkoff
and Davison (1964). The purpose of the questionnaire was
to obtain a rough index of £'s attention to the experi­
ment film and to generate ideas for further research with
 42

the film. When the� completed the questionnaire it was

the end of his participation in the study.
Questions were answered only about the film content,
but not the purpose of the study. In response to direct
questioning the �s were told the study was aimed at find­
ing out if there were any correlates between skin re­
sistance and the different scenes in the movie. It wan
reemphasized not to discuss the experimental procedure
or film content with anyone else who might be connected
with the study for one week. Each S was afforded an op­
portunity to express his feelings about the film and all
aspects of the ex�eriment, but this information was not
part of the data.
 CHAPTER III
RESULTS

The data to be presented were analyzed using an

ipsative approach to compare and contrast responses of the
same individuals over different trials. Traditionally,
theoretical issues concerning the measurement of physio­
logical and psychological response measures to stress were
controversial not only· on methodology, but also on the
type of experimental design used. Supporters of the
ipsative approach (Opton & Lazarus, 1967) felt it was
more sensitive than the normative approach in detecting
physiological and psychological correlates of stress
response changes over trials when using movie films.
Nomikos, Opton, Averill and Lazarus (1968) suggested
that a series of graphs be plotted with the original data
in responses across trials to visually illustrate rela­
tionships between the variables. These graphs also
allowed comparisons of physiological and psychological
response measures during each accident of the film.
For continuity and simplicity the results were
presented in three sections. The first section was the
psychological response measure (A-Stat e) plotted with
graphs, followed by a two way a�alysis of variance using

43
a repeated measures design (Winer, 1962), and the
differences tested using a priori! tests for matched
§_s. In the second section, autonomic rdsponse measures
(GSR) were also graphed and statistically analyzed in
the same manner. The third section presented the
correlations between A-State and GSR response measures
for each of the three trials in the film.

Psychological Measures
The A-State scores were calculated for each Ji
within his group (F.G or CG) and transformed into
normalized !-Scores using the Norms for College Fresh­
men published in the State-Trait Anxiety Inventory Test
Manual, Form X (Spielberger et al., 1969), P• 10).
Group means were computed for the baseline and over each
of three trials during the EF (Appendices D & E).
Figure 1 depicts an incremental rise of the EG and
the CG means from their baselines through Trial I. From
Trial I to Trial II, the EG mean decreased while the CG
mean increased and from Trial II to Trial III, the ID
mean decreased further while the CG mean increased.
A difference between the group means over trials on the
graph is obvious, but to ascertain statistical signifi­
cance, an analysis of variance using the two factor
repeated measures design was computed.
 4.5

.56
CG (li=l6)
55
54
CIJ
53
0
C)
Cl)
52
I
E-i 0
s:l 51 I\
•r-1 I \
I \
CIJ
(I) 50 I \
CIJ I \
s:: I \
0
P-, 49 I '
m I

48
(I) I
�
I \
(I)
I
M \
"1 47 I
I
\
I \
p,
..p
46 \
\

·g
Q)
\
\
45 \

(I) 44 \
\
o,
m ''
43 ''
..p
Cl)

''
42 'o EG
(£! = 16)
41

0 1 Baseline '1'-1 II

;:state Anxiety Heans

Fig. 1. Group .::,tate Anxiety response means after
expo�ure to each accident for the matched
experimental and control group.
 46

As Table 1 indicates there was a significant dif­

ference (p<. .02.5} between the EG and CG A-State response
measures suggesting the task (IV) had an effect on one of
the groups. Significance (.E <. 01) was also attained over
Trials, meaning one of the groups' A-State scores changed
over Trials. The significance {_p < .001) of the (IV x
TRIALS) interaction suggested that one of the two groups
scores changed at a different rate from the other. Nor­
mally a multiple comparison technique is used to detect
where these differences exist. -
�ince the F-Ratios were
all significant these differences were examined using a
priori! tests as indicated in the hypothesis.
At test for matched compari3ons was used since the
Ss selected were originally matched in pairs. Table 2
shows no significant difference (p>.10) between the EG
and CG measures during Baseline and Trial I, significance
(p < . 01) for Trial II, and �.,ignificance (p <. 001) for
Trial III.

Autonomic MeaE,g�
Two measures of skin resistance were taken in this
study; 1) galvanic skin re:Jistance (GSR) which were phasic
change8 and 2) basal skin resistance (B�R), which were
tonic or autonomic response baseline changes. Prior to
startinG the EF, both GSR and B�R response levels were
calibrated to a zero baseline for each s. This procedure
 47

TABLE l

Summary Table for Analysis of Variance

of Matched Experimental and Control Groups to
the Independent Variable of the Task over Trials f'or
the State Anxiety Scale Response Measure (.lt'orm X-1).

Source df MS F

Between Subjects 31
Exnerimental Condition (IV) 1 852.02
Error (b) 30 133.73
Within Subjects 64
Trials 2 34.07
IV x Trials 2 646.70
Error (w) 60 6.79
Total 95

-:} p � . 025
�H:· p< . 01
-::-��-:;- ] < . 001
 48

TABLE 2

Significance Levels for A Priori t tests

for Hatched Subjects Between the Experimental
and Control Group of State Anxiety Scores

Trials t significance

Baseline 1.15 p > .10

Trial I 1.65 p > .10

Trial II 3.31 .E < .01

Trial III 4.45 p < .001

 49

_was used to allow a simple analysis using a t test for

matched group comparisons between pre- and postbaselines
for the EG and CG response measures (Nomikos et al., 1968).
The t test baseline is preferred to Lacey and Smith's
ll954) autonomic liability score when the G�R and BSR can
both be calibrated to a zero baseline for each� before
the experiment. If the t test is not significant at the
.01 level the interpretation is made that the pre- and
postbaseline measures are statistically the same and no
covariance adjustments are needed to interpret response
changes during the film. A nonsignificant matched t test
also satisfies the .,,law of initial values" lBenjamin,
1963), which states if a Sis exposed to a particular
stress condition, the magnitude of his autonomic responses
is probably not a function of the prestre3s level. Table
3 shows nonsignif icance (p > . 01) for both the �G and CG
response measures, meaning that the GSR changes during
the .1!:F were true changes and needed no corrections or
adjustments.
The graphed data were presented in kilohms for stan­
dardization and comparison. Figure 2 depicts the mean
G�R response for the EG and CG plotted in one minute in­
tervals over the entire experimental period. Comparisons
between pre- and postbaselines, show that the EF does
result in GSR·increanes in anticipation to each accident
scene, and that the EG response measure decreases in Trial
 so

TABLE 3

Matched t test Comnarisons

� B�tween
Pre- and Postlog GSR Baseline Levels
of the Experimental and Control Group

GROUP significance

CON'l'RUL 1.0.5 P > • Ol

2. 71 p > .01
 PRE .sTRE;ss NK STRESS
I LANK ci�R�d� ni.Ju�K POST
:1 3ASELINE :SCEIJE I �CEI'JE II r II �cz:;E III III i3A�ELINE
0
l l l3
�
�
•ri
�
2.2
. I I .I
•ri
2.0
(l) CG --- N:16
(.)
1.8
s::
ci
EG ------ !I= 16
..µ
:)
1.6
•r-i
:J
(l)
1 .4
r:r::
s:: 1.2
•r-1
y
7i 1.0
.
:,
I \
(.) I \
•r-1 s I \
s:: I •
� \ I\
I
:> \ I \
� I
\
Film
I \
.6
Cli \
a
l Starts
I ' \
I \
-4 ....., I \
\
rFilm ends

,, ' I \
.2 t' \ I
',
,... -, -·-
---------·,,
'·-- - --·.............. .......,... '•--- -- ....
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Time in 1 minute intervals
\rt
Fig. 2. G�H group means comparing matched EG and CG ove� pre- and t-J
postbaseline and during the entire experimental film.
 52

II and Trial III, when compared with the CG measure.

Figure 3 presents, in more detail, each accident
scene in five second intervals; twenty seconds before,
and ten seconds after each accident scene impact (the
criterion for the GSR responses analyzed). Means for
each of the seven 5-second intervals w�re computed for
each S during each of the accident impact scenes, and
�hese means graphed as EG and CG. The graph shows, in
considerable detail what happens during each accident
scene between the two groups.
Figure 4, depicts each of the �s• seven 5-second
interval means computed into a total GSH mean level res­
ponse per�. Each �s• total GSH mean level resuonse was
then computed into a Experimental or Control Group mean
for each Trial {I, II and III) and pre- and postbaseline
time interval. The graph shows a stable pre- and post­
baseline between the EG and CG means over time. There is
an increase of both group means in Trial I, and a decrease
in the EG mean during Trial II, while the CG mean increases.
In Trial Ill the EG and CG means are shown to decrease,
but the difference between the group means is large.
An attempt was made to analyze the G�H data (�s mean
level response in kilohms) using an analysis or variance.
Homogeneity was achieved, but the EG and CG distributions
were not normal. The analysis yielded nonsignificance
and the variance was difficult to account for.
 ACCIDENT ACCIDENT ACCIDENT
3 �0 I II_ III
O'l
2.8 •
,,.,. ..........
I \
\
]0
r-1 \
2.6 ,,
•r-l ,. \
CG--­
� \
I
2.4 EG-----
I __... \
� f
.,,,,...-- \ •
•r-i I
2.2
(D 2.0 I
(,)
·I
C I l
C1S I
1.8
..µ
C'J 1.6
•r-1
t1 I ANSWER
(I) I
1.4
0:::: I
1.2 '
•r-l
I ... .....,,. . .. _ .....,
I
� I
1. 0
C/l ,, , ' ...
__
.8
'
/TASK START� -. .,......... ....
.6 ( IV)
-4 ..
I \. '',,
.2 I
IHPACT I II-'.IPACT
I
... ,. IMPACT
I ,
1
''-
0
-20 -15 -10 -5 0 5 10 -20 -15 -10 -5 0 5 10 -20 -15 -10 -5 0 5 10
Time in seconds from Impact
Fig. 3. G�R group means for each of the five second intervals before
\J1.
and after the accident impact comparing the EG with the CG,
and especially the effect of the IV on the EG in Accident II. \.,.)
 54

CG • •
2.0 EG 0-------0
1.9
1.8
1.7

r,
1.6
1 5
I
\
.E 1 .4
\
r-1
•ri
1.3 I \
s::
1.2 I \
•r-1
1.1
I \
G)
I \
1.0 I \
Ill
.9 I \
I \
�
(1)
.8 \o,
I
s::
•ri
�
.7 I ' ....
'1l
.6 I
.5 I 'o
\
I
.4 I
\
\
.3 I
.2
.1

0
PRE TRIAL TRIAL TRIAL POST
dASELINE I II III BASELINE
GSR GROUP HEANS
Fig. 4- G.jfl i1eans of' the seven $-second inter-
vals computed into a total mean f'or each
group (EG and CGJ, over trials, also a
comparison of pre- and postbaselines.
 55

To compensate for not meeting the assumptions for

analysis of variance, the GSR kilohms were transformed
using a logarithmic conversion. The principle reason for
using logs was that biological systems tend to obey log­
arithmic laws {Montague & Coles, 1966). The log to the
base 10 of X, was selected as it was the most commonly
used in GSR research (Fenz & Epstein, 1967; Hare, 1968;
Hinton, 1970). The transformation yielded homogeneity,
normalized the distributions, and resulted in an additive
model. These conversions are presented in Appendices F
and G.
The summary table for the analysis of variance is
presented in Table 4. As indicated, there was a signifi­
cant difference (p <. 025) between the EG and CG GSR res­
ponse mea�ures suggesting the experimental task (IV) re­
sulted in a change between groups. There was also signifi­
cance (p<.001) over Trials meaning the groups changed as
a function of the Trials. A significant (IV x TRIAL0)
interaction (p < . 001) indicated that one of the groups
GSH 1 s changed at different rates from the other. These
differences were tested with a priori! test� for matched
compari:,;ons within each trial {'rable 5). There was no
significant dirference (p>.10) between the EG and CG
measures within Trial I, �ignificance (p<.001) for Trial

I� and significance (p <.Ol) within Trial III.

 56/57
TABLE 4
Summary Table for Analysis of Variance
of Matched Experimental and Control Group
to the Independent Variable of Task over
Trials for GSR log Response Measures

uource df M.:> F

Between Subjects 31
•::-
Experimental Condition (IV) 1 1.4561 5. 5746
Error (b) 30 .2612

Within Subjects 64
{H}
Trials 2 .7871 1124.4285

47_5. 428_5
{H}
IV x Trials 2 .3328
l!;rror (w) 60 .0007
Total 95

-::- p <. 025

►:H:- p < . 001

TABLE -5

Significance Levels for A Priori t tests

for Matched Subjects Between the Experimental
and Control Group for log GS.H Hea.sures

Trials t .':>ignificance

Trial I .9859 p > .10

Trial II 4.19 p <. 001
Trial III 3.36 P< .01
 58

As Figure 5 indicates, the group mean BSR did not

vary above or below the 200 ohm interval specified by
the manufacturer at any time during the procedureo
Since there were no changes a statistical analysis
was not possible.

Correlations
Table 6 presents the Pearson Produc t-Moment Corre­
·1ations between the A-State Normalized T-Scores and log
GSR measures for each.§_ (1 through 32) for each trial
(I through III). The correlation for Trial I yielded
marginal significance (p_�.10), Trial II and Trial III
were significant (I?..(. .0,5) and (p_ < .02), as was the Total
Trials (I?.< .05).
 ?H:b; POST
r3ASELlN.t!; EXPERIMENTAL FILM BAS.l!:LINE
70
60
C/l CG N:16
50
EG .,._______ E =16
40
•r-1
Film-.{
fl:: 30 Starts 14 Film Ends

20

10
... ---•----•----•-------- ----
----.. --- =--·--·-
- - --a ----
- ---- - •-- - • - ··• - - • ��·
0 ------- --- A -=---•-.
.•- --- - • •- .. -•-
t-..P"----------------- ZER O B SELINE

-10

-20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Time in 1 minute intervals

VI.
Fig. 5. B�R group means comparing matched EG and CG over pre- and post­ -.c
baseline and the entire experimental film. BSR is in Ohms.
 60

TA13LE 6

Pearson Produce-Moment Correlations

Between Spielberger State Anxiety Normalized
,!-Scores and log G.::iH Measures on
Trials I, II, III, and Total Trials

TRIAL
Source N I II III TOTAL TRIAL::i
(E= 96)
�xperimental � :- •:� ::--:�
. 3.s/ . 41 • 23•
.:I!.�'--'�
X 32 • 29
"
fl
"'

Control

-::- p -<. .10

,,,,
�.-,.. p < . 05
p < .02
 CHAPTER IV

DISCUSSION

The data presented in the results section have, in

the main, supported the experimental hypotheses. The
first hypothesis regarding GSR response measures increasing
during specified time intervals prior to the impact of each
accident was supported by the graphical analysis in Figures
2 and 3. The A-State scores also increased similarly with
the GSR response measures (Figure 1) lending additional
support to the hypothesis. The correlation in Trial I,
between A-State and log GSR response measures failed to
reach significance at the .05 level, however, the corre­
lations were significant beyond the .05 level for Trial
II, beyond the .02 level for Trial III, and beyond the 005
level for Total Trials. These correlations were presented
in Table 6, and confirm the first hypothesis.
The second hypothesis s tated that the Experimental
Film would qualify as a valid symbolic model based upon
the relationships of three variables; A-State, GSR and
accident scenes for each of the three Trials. This hy­
pothesis was supported with graphs presented in Figures
1, 3, and 4,· and statistically by the correlations in
Table 6. These data support the interpretation that the
61
 62

EF functions as a valid symbolic model to induce psycho­

logical stress.
The third hypothesis stated there would be no signi­
ficant differences in response measures between the ED
and CG during Trial I. This hypothesis was fully sup­
ported with the analysis of variance for each response
measure depicted in Tables 1 and 4, by the i tests in
Tables 2 and 5, and visually by the graphs in Figures 1,
2, 3, and 4. In none of the log GSR and A-State dat a was
significance achieved,.which is interpreted as ther e being
no differences or changes between F.G and CG measures during
Trial I.
The fourth and main hypothesis was strongly supported
by most of the data in the results section. Essentially,
in Trial II, psychological stress as induced by the
symbolic model was significantly reduced for the Fl1 by
intervention of the task (IV) during a defined interval
at the onset of anticipatory stress. This decrease in
the F.G mean was especially noticeable when comparisons
were made with the matched CG mean as shown in Accident
II of Figure J. The IV was introduced 15 seconds before
the accident impact and was the point where the ED mean
began to decrease and the CG mean increased. These same
changes between EG and CG response measures were reflected
in Figure 1, 2, and 4. Table 6-lends additional support to
the main hypothesis by yielding a significant (.E,<o05)
 63

correlation between A-State and log GSR response measures

during Trial II, through both t he EG and CGo Tables 2
and 5 depict matched§_ :t tests between the ID and CG for
the A-State (E_,< .01} and log GSR (;e_<. .001) response
measures, demonstrating that there was a statistical
difference between the EG and CG.
The fifth hypothesis stated there would be slight
decreases in GSR and A-state response measures during
Trial III for the EG when compared with the CG. This
hypothesis was supported in Figure 3, by graphs. Statis­
tically the decrease in the ED measure was supported with
significant t test� for both the A-State scores in Table
2 (,E_.(,.001}, and for the log GSR 1 s in Table 5 (,E_.t:..Ol)o
In addition, the correlation between A-State and log GSR
(Table 6) was highest in Trial III (£=•41, ,E_<:,.02).
These statistical data strongly support the fifth hypo­
thesis as do the graphs in Figures 1, 2, 3, and 4.
The sixth hypothesis stated that there would be a
total significant difference between the En and 00 res­
ponse measures throughout the EF. The A-State measure
as graphed in Figure l shows the F.G and CG mean differences
across Trials, and the total difference supported by sta­
tistical significance beyond the .025 level (Table l)a
Figure 4 depicts the GSR mean difference between the En
and CG measur�s across Trials, and the total difference
supported by statistical significance beyond the .025
 64

level (Table 4). The Total Trials correlation (Table 6)

'was significant beyond the .05 level supporting the re­
lationship of the A-State and log GSR response measures
and affirming sound confirmation to the sixth hypothesis.
The seventh hypothesis did not have to be tested as
no� responded to the questionnaire with unusual answers
or had unusual expectations about the study.
The eighth and final hypothesis that the BSR for both
the En and CG would not change significantly over trials
was fully supported by Figure 5. This figure shows minimal
changes and a statistical analysis was not possible.
Six major conc:uusions appear warranted from the
results; (1) the experimental film qualified as a valid
symbolic model to induce defined intervals of repeatable
psychological stress; (2) both GSR and A-state response
measures can correlate to provide a physiological and psy­
chological relationship to measure and study stress induced
by the film; (3) the stress response can be modified by the
intervention of a specified task prior to the accident im­
pact; (4} this reduction of stress appears to be gener­
alized to the following trial in an attempt by the subject
to control his reaction to the next stress response; (5)
the GSR appears to be an adequate physiological response
measure to detect phasic changes of stress, provided the
stress stimuli are defined and repeatable, and finally;
(6) the A-State Scale appears to be the moat appropriate
�sychological response measure to detect changes in anxiety
 6$

of certain stressful situations as it permits statistical

'analysis of response changes for each subject over time.
The near continuous, silent experimental film con­
sisted of three distinct accident scene� that resulted in
GSR response measure changes when shown to subjects. During
most of the EF the GSR measures were relatively stabls and
increased only in anticipation to each accident, then sta­
bilized when it was over. These same changes in GSR mea­
sures were reported by Lazarus and Opton (1966) using the
original unspliced version of the film.
Splicing the film into three separate accident scenes
did not appear to reduce variability in GSR responses; in
fact, it offered a visual pre- and postbaseline comparison
for each scene (Figure 2). The purpose of each splice was
to provide an interval after each scene allowing the subjects
time to complete the A-State Scales. Figure l shows the A­
state scores for the EG and CG changing in relation to each
accident sceneo In examining the graphs (Figure 1 and 2) it
can be observed that there is a relationship between the GSR
and A-State response measures within each accident scenea
This relationship is interpreted to support the use of the
EF as a valid sj'lnbolic model to induce defined, reproducable
intervals of psychological stress.
Correlations between GSR and A-State response measures
were achieved within each trial during the entire EF. In
general, both response measures correlated (.£<-05) through­
out trials during the entire EF. This correlation was
 66

important because it statistically supported the relation­

ship between GSR and A-State response measures during the
accident scenes and justified using the EF as a symbolic
model. It also was the first time a significant correla­
tion was demonstrated between a physiological and psycho­
logical response measure of stress using the film, "It
Didn't Have to Happen."
Previous experiments involving multiple response
measures (Brinbaum, 1964; Mordkoff, 1964; Lazarus, 1966;
Nomikos et al., 1968; Dysinger & Ruclanuck, 1969) were rel­
atively unsuccessful in obtaining significant correlations
between physiological and psychological response measures to
stress using films. The major problem was each study at­
tempted to correlate at least two autonomic and one psycho­
logical measure of stress, thus reducing the probability of
obtaining a significant correlation. The two autonomic mea­
sures most frequently used were Heart Rate (HR} and Skin
Conductance (SC). Results have been clearest in the case of
SC. Fenz and Epstein (1967) studied the stress patterns of
novice versus experienced parachutists. They concluded that
SC alone was the best physiological response measure of
stress to attain a correlation with a psychological measureo
Examining the correlation reported during Trial I
shows it failed to reach significance at the 005 level,
but did at the .10 level. The .10 correlation suggests
a relationship between GSR and A-State response measures
 67

during Trial I. Lazarus (1966) described similar results

in his re�earch and concluded that during the first scene
the subjects were naive and did not lmow what to expect;
however, once the first scene was over, the correlations
of other scenes increased. The interpretation of Lazarus
was supported in this study by answers-to the question­
naire in Appendix C. Ten subjects responded with some
�orm of reply indicating they were unprepared for what to
expect during the first accident scene, but felt more pre­
pared for the remaining scenes.
another explanation for the .10 correlation in Trial
I, could have been a latency factor. The GSR was a con­
tinuou!3 re��pon.se measure recorded during the entire EF

and analyzed, in detail, during each accident scene,

while State Anxiety waG a single unit measure recorded
during the blank intervals following each accident scene.
A similar latency problem was reported by l·lalmstrom {1965)
and Mordkoff (1964) in their research with stress using
movie .films. Although the marginal relationship between
GSR and A-State measures in Trial I was initially thought
to be a curvilinear one, neither a correlation ratio nor
a scatter plot of the two dependent variables supported
this supposition.
The mo�.Jt reasonable interpretation explaining the
marginal correlation durin� Trial I, appears to be a com­
bination of subject expectancy, naiveness and u.nf'amiliarity
 68

of the stress film content. Once the subject experienced

'Trial I and became familiar with the procedure, the corre­
lations for Trial II and III improved. The correlation dur­
ing Trial II was significant beyond the .05 level, while
Trial III achieved significance beyond the .025 level.
The correlations reported strongly point out that
when a psychological stressor can be ·defined and con­
trolled, a relationship can be demonstrated between a
physiological and psychological response measure to
stress. This finding contributes to the methodology lit­
erature a valid symbolic model of stress, and a procedure
which allows various parameters of physiological and psy­
chological stress responses to be observed, investigated,
and studied.
The present study also resulted in a method to modify
a stress response during Trial II, and demonstrated these
results graphically and statistically with a difference
between :m and CG response measures. The introduction of
a verbal task (IV) during Trial II, at a defined point
prior to the accident impact, resulted in the En's GSR
and A-State response measures decreasing significantly
(.P.< .025) when compared with the CG measures. This de­
crease in response measures was consistent for all sub­
jects within the EJ., while increases of these measures
occurred for their matched coW1terparts in the CG. The
difference between En and CG measures was interpreted
 69

to mean that psychological stress induced by the EF' used

in this study, can be modified through the intervention
of a specified task prior to the accident impact.
Previous studies have demonstrated other methods to
modify a GSH response and attain results similar to those
in this study. Geer and Klein (1969} modified GSR mea­
sures by changing the intensity of light used to flash
slides. Sniesman et al., (1964) modified GSR measures
using a movie film and providing different instructions
to subjects before the film started. Riess (1966) used
supportive therapy, and Folkins et al., (1968) used an
analogue of Wolpe's systematic desensitization to modify
GSR measures.
The evidence cited from the preceeding studies sug­
gests that any change to the stimulus complex could re­
sult in some GSR measure change. However, it would be
difficult for these changes to effect other nonautonomic
response measures, specifically the A-State measures.
Since the A-State response measures were recorded later
in time (two minutes after the accident scene impact) it
would have been difficult to record any changes in the
stimulus complex other than those of the stressful ex­
perience (�pielberger et al., 1969). The fact that GSR
and A-State response measures did correlate significantly
(p <. 05) during Trial II, further supports the interpre­
tation and conclusion that psychological stress during
 70

Scene II was modified mainly as a function of the experi­

'mental task (IV}, and possibly associated with these
changes, some residuals of changing the stimulus complex.
1'he results in Trial IIl appear to be related to
Trial II. During Trial II, there was a significant dif­
ference (p <. 001) between the EG and cq- A-Stete respon:::s
measure which remained significant (p<.01) during Trial
.III. The A-State measure continued to increase for the
CG, and conversely decreased for the EG to a point below
the original baseline and Trial I comparison means. Spiel­
berger in his test manual (�pielberger.et al., 1969) re­
ported the mean of A-State scores after a stressful movie
was 50.03, with a standard deviation of 12.48. 1be mean
of the EG for this study at the end of Trial III was
41.87 with a standard deviation of 7.91, while the matched
CG had a higher mean of 55.43 and a standard deviation of
6.96. This difference between the EG and CG means, was
significant beyond the .001 level and suggests a gener­
alization effect was operating within the EG. The most
suitable explanation for this generalization effect appears
to be in the concept of the cognitive mediational response
process. (l•'enz & .l!;pstein, 1967).
The concept of the cognitive mediational process was
supported by the research of Fenz and Epstein (1967) who
studied stress responses of novice and experienced para­
chute jwnpers approaching a jump. Their results indicated
 71

.that novice jumpers demonstrated stress responses right

up to the actual jump. They concluded experienced jumpers
who had repeated eXposures to threat were able to recog­
nize the early rise in stress as an automatic signal of
danger and thereby initiate control of such periods of
anticipatory stress. Fenz and Epstein.(1967) theorized
that cognitively mediated processes served as inhibitory
reactions which prevented arousal from becoming excessive,
thus providing a highly adaptive mechanism for the mastery
of threat.
During Trial II the EG had the opportunity to be ex­
posed to a method to control psychological stress and was
able to apply it to Trial III. The CG did not have this
opportunity and their response measures showed increases
throughout all trials. Thus, the EG was able to utilize
the concept of the cognitive mediated process to inhibit
or control stress responses during Trial III.
The cognitive mediational process interpretation was
also supported by eleven of the sixteen subjects in the
EG through their answers to the fourth question in Appen­
dix C. Each of the eleven, in some fashion, reported
that they attempted to control stress during Trial III,
by counting ·objects in the vicinity of the workers in
the film. The CG did not answer question four in any de­
tail except to state they followed the instructions.
In conclusion, the results of this study support
 72

the interpretation that some type of generalization took

place in the ID from Trial II to Trial III to reduce
psychological stress. In agreement with Fenz and Ep­
stein's (1967) interpretation, the author feels the modifi­
cation of the stress response in this study was a direct
fW1ction of the subjects using a type of cognitive mediated
process to control stress. In short, the IV taught the
EG a technique to make the film content less stressful.
During Trial II and III both the EG and CG demon­
strated a decrease in GSR measures (Figure 4}, but yet
the difference between the two groups was significant be­
yond the .01 level. As noted previously, GSH measure
fluctuations corresponded to the content in the film.
However, there was a progressive decrease in the GSH mea­
sures for both groups near the end of the film (after
approximately 20 minutes of being connected with the GSH
apparatus). The majority of stress studies using GSH
measures reported a similar progressive decreases after
subjects were connected to the G�H apparatus for fifteen
minutes or more. Lazarus et al., (1962), and Geer and
Klein (1969}, interpreted this progressive decrease in
the G�H measure as adaptation. Since this study used
subjects matched into EG and CG, and the GSH measure de­
creases were progressive toward the end of the EF for
both groups, then it appears the explanation of adaptation
to the GSH apparatus is most reasonable.
 73

Another possible explanation for the progressive

decrease of the G�tt measure during Trial III might have
been the content of the stress scene. This was the only
scene that dealt with a worker being killed and not in­
jured. i111ere were several comments by subjects in both
groups concerning the death scene. Many felt it was not
upsetting, was unrealistic, and of all the scenes appeared
as through it were staged. Host of the subjects in both
groups commented on the realism of the accidents in Trial
I and II, and were able to identify with the characters
in the film. Several commented they felt they could hear
the sound of the saws and drills even though it was a
silent film.
Both explanations of adaptation and the death scene
connnents appear to account for the progressive decrease
in the GSR measures during Trial III. However, the sta­
tistical significance beyond the .01 level in log GSR
response measures between the EG and CG demonstrates
there was a difference between the groups. This signifi­
cant difference between the EG and CG lends additional
support to the conclusion that decreases in the GSH res­
ponse measures for the EG during Trial III was a cognitive
mediated process generalized from Trial II.
The results of this experiment supported the useful­
ness of the GS� as a physiological response measure in
studying highly defined, repeatable stress situations.
 74

The main advantages of using the GSR measure was the

immediacy of response and rapid after discharge, making
analysis of the response record simple and reasonably
accurate (equipment permitting the option to calibrate
e,ach subject at his own zero baseline).
The disadvantages of using the GSR measure stem
largely from simple methodological problems such as pre­
paring the subject for electrode application and con­
trolling body and arm movements. 1bese were resolved by
using a comfortable chair and establishing rapport with
the subject. The problem of individual response specifi­
city was resolved by using individual ohmic range settings
from 2,500 to 5,000 ohms for each subject. This method
for correcting psychophysiological measures for individual
responsivity in ohmic ranges was also supported by the
research of Lyken, Rose, Luther and Holey (1966). The
ohmic range settings made statistical analysis somewhat
difficult, however, the logarithmic conversion solved
this problem.
The Spielberger Trait-State Anxiety Inventories were
the psycholoGical response measures used in this study
(Spielbere;er et al., 1969). 'l'hese scales provided mea­
sures which were adaptable to assessing psychological
stress in movie rilms. The scales were well defined,
easily administered and scored, and had established norms
which yielded signiricant correlations with GSR response
 15
,measures during accident impact.
The A-Trait Scale (Form X-2) was used to select sub­
jects prior to the experiment and provided information to
aid in development of the hypothesis. The results of this
study indicate that the A-Trait Scale might be comparable
to a GSR baseline in that it provides a type of psycho­
logical zero point from which comparisons can be made.
The trends in this research also support the interpreta­
tion that the A-Trait Scale could possibly be used to
select high autonomic reactors for future GSR research
when using a symbolic model.
Based upon the data in this study, the A-State Scale
{_ti'orm X-1) was the most impressive response measure. 'l'he
scale was administered to the same subject four times
during a twenty minute interval, or on the average of
once each five minutes. The results showed the CG res­
ponse measures increased successively over trials, while
decreases were recorded for the EG measures. This finding
was interpreted to mean that only a minimal practice ef­
fect was functioning on successive administrations of the
scale. In hi3 manuel, �pielbereer et al., ll969) reported
that the A-State 8cale could be administered repeatedly
with little 6r no practice effect without randomizing or
changing the order of items. This statement was supported
by the results in this study.
1he A-Trait and A-State �cales appear to be adaptable
 76

to many facets of research studying psychological stress .

They can be used on a test-retest basis and still detect
differences between groups, and they have norms available
which lend themselves to ease in interp1•etation and statis-
tical computation Finally, there were no obvious disad-
vantage� noted in using the A-State uca�es to study p8�.
chological stress using a movie film ·as a stressor.
In conclusion, the purpose of this study was to con­
struct a movie film which could be used as a symbolic
model to induce and reproduce defined intervals of p3ycho­
logical stress. Another purpose was to control the stress
stimuli in an attempt to improve the methodology of
studying stress and it's correlates. This purpose was
achieved. Succe:1s was also reported in correlating a
physiological and psycholoBical response measure during
accident scenes in the film. Finally, it was demonstrated
that both physiological and p2ychological response measures
to stre�s could be reduced by initiation of a verbal task
prior to the onset of anticipatory stress.
 CHAPTER 5
SUMMARY

Psychological stress and its control are widely re­

cognized as central problems in human life. A revi�w of
the literature indicates the most effective approach to
produce, control and study the effects of psychological
stress is to use a movie film as a stressor. A growing
body of research indicated that not only can a film be
used as a symbolic model to reproduce defined intervals
of stress, but also that control of such intervals could
result in the experimental reduction of stress. The re­
sults of other studies demonstrated the problems involved
in measuring stress by using physiolo�ical and psycho­
logical response measures. This study was designed to
construct a movie film that would create and reproduce
intervals of anticipatory psychological stress. The main
purpose of this study was to use the movie film as an ex­
periMental model to develop an improved laboratory meth­
odology to record increases and decreases of psycholocical
strecs and to study its correlates using the CSR measure
and ..:ipielberger �tate Anxiety ;:,cclc scores as dependent
variables.
The followinG hypotheses were tested:
1. The G�rl response measure will increase durinc
3pecified time intervals near the impact of the
accident for each stress scene in the film.

77
 78

2. The Experimental Film will qualify as a valid

symbolic model based upon relationships among
three variables during each trial within the
Experimental and Control Group.
3. During Trial I (Scene I), there will be no
significant difference of GSR and State Anxiety
response measures between the Experimental and
Control Group.
4. During Trial II (Scene II), psychological stress
as induced by the symbolic model will be reduced
within the Experimental Group by the interven­
tion of a task in a defined time interval during
the onset of anticipatory stress.
5. During Trial III (Scene III), there will be de­
creases in GSR and state Anxiety response mea­
sures in the Experimental Group when compared
to the Control Group.
6. There will be a total significant difference
between the Experimental and Control Group
response measures.
7. Subjects who respond to questionnaires at the
end of the study with unusual expectations will
lack consistent relationships between GSR and
State Anxiety response measures throughout the
experiment.
8. The Basal Skin Resistance for subjects in the
Experimental and Control Group will not vary
significantly over trials.
Thirty-two male volunteers served as Sa for this
study. These Ss were matched and then randomized into
two groups of 16 each using Spielberger Trait Anxiety
Scale scores and age as the matching criteria. The Ex­
perimental Group was shown the Experimental Film depicting
three accident sceneso During Scene II, a verbal task was
introduced exactly 16 seconds before the accident impact.
The Control Group followed the same procedure� but did
not receive any task. Both groups were administered the
 79

state Anxiety Scale after each accident scene and a ques­

tionnaire when the film was over.
The results supported each of the hypotheses and six
conclusions were drawn; (1) the Experimental Film qualified
as a valid symbolic model to induce defined intervals of
repeatable psychological stress; (2) b9th GSR and state
Anxiety response measures could correlate to provide a
_physiological and psychological relationship to measure
and observe stress as induced by the film; (3) the stress
responses can be modified by the intervention of a speci­
fied task prior to the accident impact; (4) this reduction
of stress appears to be generalized to the following trial
in an attempt by the§. to control the next stress response;
(5) the GSR appears to be an adequate physiological mea­
sure to detect phasic changes of stress, and finally; (6)
the State Anxiety scale appenrs to be the most appropriate
response measure to detect induced changes in psychological
stress.
 81

APPENDIX A

TRAIT ANXIETY INVENTORY

STAI FORM X-2

NAME -----------------------DATE------
DIRECTIONS: A number of statements which
people have used to describe themselves are :t:-
given below. Read each statement and then 1--'

circle the appropriate number to the right s0 �

0
rn
of the statement to indicate how you gener­ ca
CT
CT
ally feel. There are no right or wrong
answers. Do not spend too much time on any z
(b
one statement but give the answer which seems <
to describe how you generally feel. (D (b
to

1. I feel pleasant..•• •• ••• ..• • •.. • •.•...... 1 2 3 4

2. I tire quickly. . •.. .•..•• • • •• • ••• • .•... •• 1 2 3 4
3. I :reel like crying ..••.•.•.....•.•....... 1 2 3 4
4. I wish I could be as happy as others
seert1. to oe . . .••••••••••••••••••.••...•••• 1 2 3 4
I am losing out on things because I can 1 t
make up my mind soon enough..•.•.•.•..•.. 1 2 3 4
6. I feel rested . ........................... 1 2 3 4
7. I am "calm, cool, and collected u ......••. 1 2 3 4
B. I feel that dirficulties are piling
up so that I cannot overcome them........ 1 2 3 4
I worry too much over something that
really doesn't matter ....•...•....••••••. 1 2 3 4
10. I am happy ................................ 1 2 3 4
11. I am inclined to take things hard.......• 1 2 3 4
12. I lack self-confidence............... .••. 1 2 3 h
13. I feel secure ...........•...•.......•..•. 1 2 3 4
14. I try to avoid facing a crisis or
difficulty.. •. . .•.•• •.... •. . ..•. •.•. . . ..• 1 2 3 4
 82

APP�NUIX A--Continued

15. I feel blue. • • • • • • • •• •• • • •• . •••••• • •• • ••• 1 2 3 4

16. I am content. . • • •• • . • • •• • • • • • • •• • • • • • •• • • 1 2 3 4
17. Some unimportant thought runs through
my mind and bothers me ••••••••••••••••• •• 1 2 3 4
18. I take disappointments so keenly that
I can 1 t put them out of my mind •••..••..• 1 2 3 4
19. I am a steady person .••••••••••••••••.••• l 2 3 4
20. I become tense and upset when I think
about my present concerns.••••••••••••••• 1 2 3 4
 APPENDIX B

STATE ANXIETY INVENTORY

STAI FORH X-1

NANE------ ---------------DATE------
DIRECTIONS: A number of statements which
people have used to describe themselves are
given below. Read each statement and then
circle the appropriate ntunber to the right
of the statement to indicate how you feel
right now, that is, at this r.ioment. There
are no right or wrong answers. Do not spend
too much time on any o�e statement but give
the answer which seems to describe your pre­
sent feelings best.
1. I re e 1 Calm. . . . .......... .. ..... .. . . . . .. . . 1 2 3 4
2. I feel secure..••••..••.•••••.• .••• ..• . •. 1 2 3 4
3. I am tense.••.• .• . .• • • • .•• .• . . . . . . • •• •.•. 1 2 3 4
4. I am regretful......... . .. .•••. . .•• •. . .•. 1 2 3 4
5. I feel at ease..•.••..••.••.••...••••.•.. 1 2 3 4
6. I feel upset .•.•..•...••...••......•..... 1 2 3 4
7, I run presently worrying over
possible mis:fortunes..................... 1 2 3 4
8. I feel rested ........................... . 1 2 3
9. I feel a!lX:LOU S .. . . . . . . . . • • • • • • • • • • • • • • • • • • 1 2 3 4
10. I feel comfortable . ....................... 1 2 3 4
11. I feel self-confident ..•.•••..•.......•.. 1 2 3 4
12. I feel nervous.•••.•••••••.•.•••••••••••. 1 2 3 4
13. I am jittery ............................ . 1 2 3 4
14. I fee1 11 h 1g
. h s t rung II ••••••••.••••..•••••• 1 2 3 4
15. I e.m relaxed.•••••••••••••••••••••• •· ••••• 1 2 3 4
 84

APPENDIX B--Continued
�
0 <
CD
CD
0 1-:f
(/)
0
> FJ
c1"
;l
._.
CD

s,sn
(I) 0
t<I
._.>
p'
CD
c1" 0 0

16. I feel content ••••••••••••••.•••••••••••• 1 2 3 4

17. I am worried .•••••••••••••••••••••••••••• 1 2 3 4
18. I feel over-excited and rattled .•••••••• • 1 2 3 4
19. I feel .joyful ..... � ...................... 1 2 3 4
20. I feel pleasant .......................... 1 2 3 4
 85
APfENDIX C

Sub. No • ___Name______________Age__Date

Instructions: Please answer these seven questions

on the following pages as fully as you can. Remem­
ber, your comments will be anonymous and try to be
completely frank. Answer all questions in order and
do not look ahead. Go ahead and turn the page.

1. What were your reactions to the movie?

2. What were the things in each of the three scenes that

stood out in your mind?
SCENE I

SCENE II

SCENE III
 86

3. Write down as much as you can of what went on in the

movie.

4. Did you in any way to your knowledge try to avoid any

of the scenes? If yes, how? If no, how?

5. How did your expectations before the study match up

with what actually happened?

6. vi.hat do you think the research is all about?

?. Comments.
 87

APPEi"\fDIX D

Individual Subject's Baseline, Trial I, II and III

.Normalized T-Scores as Measured by the Spielberger
State Anxiety Scale (Form X-1)

'11.H
IAL

Subject
GROUP Number Baseline I II III

2 57 55 52 51
4 41 43 36 36
6 44 51 48 41
8 '
57 50 44 43
10 41 36 38
12 40 � 44 36
EXPER II,JENTaL 14 52 62 38 44
16 43 41 33 29
18 44 48 48 41
20 25 57 36 33
22 46 51 48 46
24 44 48 48 41
26 51 61 51 55
28 50 55 44 44
30 54 60 55 57
32 47 50 48 43

Hean 46.00 ,51.68 43.56 41.87

Standard Deviation 7.88 6.30 6.86 7.91
 88

APPENDIX E

Individual Subject's Baseline, Trial I, II and III

Normalized _!-Scores as Heasured by the Spielberger
State Anxiety Scale (Form X-1)

TRIAL
Subject
GHOUP Number Baseline I II III

1 48 48 56 57
3 44 44 54 61
5 38 44 50 52
7 47 48 50 55
9 52 64 64 67
11 47 46 46 50
13 48 52. 57 61
1
CON l1ROL 15 36 44 47 61
17 32 36 41 48
19 51 54 52 57
21 46 48 57 55
23 34 41 46 51
25 32 36 40 38
27 44 54 60 62
29 41 50 52 52
31 46 55 56 60

i'-1ean 42.87 4 7. 75 51.75 55.43

Standard Deviation 6.57 7.23 6.64 6.96
 89

APPENDIX F

Individual Subject's log GSR Hesponse

Measures in Group Means over Trials
(log to the base 10 of X)

TRIAL
Subject
GROUP Number I II III

1 2.26 2.47 2.38

3 2.21 2.39 2.39
2.13 1.90 1.60
,7 2.32 2.52 2.22
9 2.06 2.15 1.65
11 2.44 2.63 1.62
13 2.02 2.23 1.77
CON'l'ROL 15 2.33 2.38 2.37
17 2.26 2.31 2.13
19 2.01 1.95 1.81
21 1.60 2.08 2.09
23 2.15 2.42 2.42
25 2.02 1.92 1.92
27 2.17 2.11 2.02
29 1.60 1.39 1.60
31 2.32 2.42 2.42

Hean 2.118 2.204 2.020

Standard Deviation .23 .31 • 31
 90

APPENDIX G

Individual Subject 1 s log GSR Response

Measures in Group Heans over Trials
(log to the base 10 or X)

TRIAL
Subject
GHOUP Number I II III

2 2.37 1.17 1.69

4 1.95 1.39 1.30
6 2.32 2.04 1.97
8 2.39 1.60 1.74
10 2.31 1.81 1.00
12 2.22 2.36 2.02
14 1.69 1.30 1.69
EXP..t;tt IE�N'l'AL 16 2.33 1.81 1.65
18 2.51 1.90 1.84
20 1.81 1.60 1.54
22 1.74 2.19 1.47
24 2.24 1.95 1.90
26 2.16 1.77 1.81
28 2.22 1.81 1.84
30 1.77 1.60 1.39
32 2.22 1.60 1.81

Hean 2.140 1.743 1.660

Standard Deviation .26 .31 .26
 REFERENCES

Aas, A. Mu����tio� Fantasies and Autonomic Responses.

Oslo, Norway: Oslo University Press, 1958.
Alexander, F., Flagg, G. W., Foster, s., Clemens, T., &
Blahd, W. Exp erimental studies of emotional stress,
I. Hyperthyroidism. PRychosomatic Hedicine, 1961,
£3., 104-114.
Alfert, E. An idiographic analysis of personality differ­
ences between reactors to a vicariously experienced
threat and reactors to a direct threat. Journal of
Exoerimental Research in Personality, 1966, 2, 200-
207.
Appley, H. H. & Trumbull, R. Psychological Stress: Issues
in Research. New York: Appleton-Century-Crofts,
1967.
Bandura, A. Behavioral modification through modeling pro­
cedures. In L Krasner and L Ullman (Eds.) Re­
search in Behavior Hodification. New York: Holt,
Rinehart and Winston, 1967, 310-340.
Derkun, 1-1. M., �::iialek, IL H., Kern, R. P., & Yagi, K.
Experimental studies of psychological stress in man.
Psychological Honographs, 1962, 1£.. ( 15, Whole No. 534).
i)irnbaum, R. M. Autonomic reaction to threat and confor­
mation conditions of psycholof,ical stress. Unpub­
lished doctoral dissertation. �erkeley: University
of California. Dissertation Abstracts, 1964, 25 (3),
2040-2041.
Drec;er, R. H. & Hiller, K. s. Comparative psycholor.;ical
studies of necroes and whites in the United States:
1959-1965. Psychological Bulletin Honograoh Supple­
ment, 1968, lQ, 1-58.
Dysinger, W. S. t: R11ckrnick, C. A. The emotional respon::rns
of children to the motion-picture situation, 1933.
Reported in Albert Dandura, Principles qf 3chavior
Modification. New York: Holt, Hinehart, and Winston,
1969.
91
 92

�pstein, s. Toward a unified theory of anxiety. In Maher,

B. A. (Ed.) Progress in Experimental Personality and
Research. New York: Academic Press, 1967, Vol. IV.
Farber, I.E., Harlow, H.F., & West, L. J. Brainwashing,
condition and DDD (Debility, Depenuency, and Dread).
Sociometry, 1957, 20, 271-285.
Fenz, W. D. & Epstein, s. Gradients of physiologic�l
arousal in parachutists as a funct_ion of an approa�h­
ing jump. Psychosomatic Medicine, 1967, 29 (1),
33-51.
F· olkins, C .. H., Lawson, K. D., Opton, E. H., Jr., & Lazarus,
R. S. Desensitization and the experimental reduction
of threat. Journal of Abnormal Psychology, 1968, 13.
(2), 100-113,
Geer, J. H. & Klein, K. Effects of two independent stresses
upon autonomic responding. Journal of Abnormal Psy­
chology, 1969, 74 (2), 237-241.
Goldstein, N. J., Jones, R. B., Clemens, T. L., Flagg,
G. W. & Alexander, F. G. Coping style as a factor in
psychophysiological response to a tension arousing
film. Journal of Personality and Social Psycholof,y,
1965, 1, 290-302.
Handlon, J. H. Hormonal activity and individual responses
to stress and easement in everyday living. In R.
Roessler and N. S. Greenf'ield (Eds.) Physiological
Correlates of Ps;rcholor;ical Disorder. Hadison:
University of Wisconsin Press, 1962.
Hare, R. D. Psychopathy, autonomic functioning and the
orienting response. Journal of Abnormal Psychology
Honograuh Supplement, 1968, lJ., l-2J+.
Hinton, D. �. An investigation of vicarious conditioning,
generalization, and extinction and their relationships
toEysenck 1 n extraversion factor. Unpublished doc-~
toral dissertation. Hattiesburg: University of
Southern Mississippi, 1970.
Kaiser, C. & Roessler, R. Stability and habituation of
non-specific G�R•s. Perceptual and Motor Skills,
1968, ll (2), 495-498.
Kamen, G. H. Effects of a stress-producing film on the
test performance of adults. Journal of Projective
l1echni ues and Personality Assessment, 1969, JJ. (3),
1

281-28.�
 93
Kerle, R. H. & Bialek, H. M. The construction validation
and application of a subjective stress scale. United
States Arm Leadershi Human Relation Research Reports:
Presidio of 1/lonterrey, California, .F'ebruary, 19
Kimmel, H. D. GSR amplitude instead of GSR magnitude:
Caveat emptor! Behavior Research Methods and In­
strumentation, 196S, 1 (2), 54-56.
Kinder, M. I. The effects of prior information, desensiti­
zation, and denial on physiological reactivity to a
stressful motion picture. Dissertation Abstracts,
1968, g§. (11-13), 4758.
Kline, V • .t3. rrracks that violence leave. Life, 1970, £2
(4), 57-58.
Krause M. s. The measurement of transitory anxiety. Psy­
chological Review, 1961, 68 (3), 178-189.
Lacey, J. I. The evaluation of autonomic responses: 'l'o­
ward a general solution. Annals of the New York
�cademy of Sciences, 1956, 67, 123-164.
Lacey, J. I. Psychophysiological approaches to the eva­
luation of psychotheraputic process and outcome. In
E. A. Rubinstein and H.B. Parloff {Eds.) Research
in Psy,chotherany. Washington, D. C.: .l-1.merican
Psychological Association, 1959, 160-208.
Lacey, J. I., Bateman, D. .I!:. & VanLehn, R. Autonomic
response specificity: An experimental study. P.§X ­
chosomatic i·iedicine, 1953, 1.2., 8.
Lacey, J. I. & Lacey, D. C. The law of initial value in
the longitudinal study of autonomic constitution:
Reproducibility of autonomic responses and response
patterns over a four year interval. Annals of the
New York Academy of Sciences, 1962, 98, 12.57.
Lacey, J. I. & Lacey, B. G. Verification and extension
of the principle of autonomic response stereotype.
American Journal of Psychology. 1958, 11., 50-73.
Lacey, J. I .., Smith, R., & Green, H. Use of conditioned
autonomic responses in the study of anxiety. In
C. }'. Heed, I. l!.:. Alexander, and �- s. Tomkins (Eds.)
Psychonathology. New York: John Wiley and Sons,
1964, 275-288.
 94

L· azarus, R. s. Psychological Stress and the Conine Process.

New York: HcGraw-.tlill Book Company, 1966.
Lazarus, R. �. & Alfert, E. The shortcircuiting of threat
by experimentally altering cognitive appraisal.
Journal of Abnormal and Social Psyrhology, 1964, ftz.,
195-205.
Lazarus ., R. �. & Upton, E. H., Jr. The study of psycho­
logical stress: A summary of theoretical formula­
tions and experimental findings. ·In C. D. Spiel­
ber�er (�d.) Anxiety and Behavior. New York:
Academic Press, 1966.
Lazarus, R. s., Upton, E. H., Jr., Nomikos, H. s., & Hankin .,
N. u. The principle of shortcircuiting of threat:
Further evidence. Journal or Personality, 1965, .:U,
622-635.
Lazarus, R. s., Speisman, J . c., Mordkoff, A. H., a. lJavi­
son, L. a. A laboratory study of psychological stress
produced by a motion picture film. Psychological
Ho��:ranhs, 1962, ]i?. (34, Whole No. 553).
Lazarus, H. :s., 11omita ., M., Opton, E. N., Jr., & Kodama,
M. A cross-cultural study of stress reaction patterns
in Japan. Journal of Personality and Social Psycho­
logy. 1966, .1Q, 4J.i4-449.
Levitt, E . .I!:. The Psycholoc:v of Anxiety. Indianapolis:
Bobbs Harrill Company, 1967.
Lyden, D. 'l'., Hose, R., Luther, 13., & Noley, M. Correcting
psychophysiological measures for individual differ­
ence� in range. Psychological Llulletin, 1966, 66,
481-484.
Malmstrom, E. J., Opton, E., & Lazarus, R. d. Heart rate
meaDurement and the correlation of indices of arousal.
Psychosomatic 1-Iedicine , 1965, 27 (6), 546-556.
1-Jartin, 13. The assessment of' anxiety by physiological be­
havioral measures. Psychological ,3ulletin, 1961, .2.§.,
23L�-255.
I-IcGuir,an, F. J. The experimenter: A neglected stimulus
object. Pnychological Bulletin, 1963, f&, 421-428.
Hechanic, D. .Students Under Stress. l'�ew York: The Free
Press of Glencoe, 1962.
 95
Montagu, J. lJ. & voles, E. H. Hechanism and measurement
of the galvanic skin response. Psychological Bulletin,
1966, §S.. l5), 261-279.
Mordkoff, A. H. l'he relationship between psychological
1

and physiological response to stress. Psychosomatic

Medicine, 1964, �' 135-150.
Nomikos, M. s., Upton, E., Averill, J. H., & Lazarus, R. ::i.
Surprize versus suspense in the production of stress
reaction. Journal of Personality and Social Psycho­
logy. 1969, § (2), 204-208.
U 1 Neil,H.F., Spielberger, C. D., & Hansen, D. N. The
effects of state-anxiety and task difficulty on
computer assisted learning. Journal of Educational
Psychology. 1969, !ill, (5), 343-350.
Opton, E. H., Jr., & Lazarus, R. 3. Personality deter­
minants of psychophysiological response to stress:
a theoretical analysis of an experiment. Journal of
Personality and �ocial Psychology, 1967, 6 (3), 291-
303.
Reynolds, J.C. The effect on viewer distance on film
induced anxiety. Dissertation Abstracts, 1969, 29
(10-A), 3341. Indiana University, 1968.
Ries G, W. li'. 'lhe effect of psychotherapy-like interviews
upon adaptation to psychological stress. In C. v.
�pielberger (l!.id.) Anxiety and r3ehavior. New York:
Academic Press, 1966, 262.
Rosenthol, R. The effect of experimenter bias on the re­
sults of psychological research. In B. A. Haher
(�d.) Pro�ress in Exuerimental Personalit Research,
Vol. I. New York: Academic Press, 196 •
Selye, H. The .Stress of Life. New York: HcGraw-Hill
Dook Company, 1956.
Schnore, N. i.1. Individual patterns of physiological
activity as a function of task differences and degree
of arounal. Journal of Exuerimental Psychology,
1959, 5.§, 117-121.
Schwartz, B. J. An empirical test of two Freudian hypo­
thesis concerning castration anxiety. Journal of
Pernonality, 1956, �, 318-327.
 96

,Slater, E. The neurotic constitution. Journal of Neuro­

iogical Psychiatry, 1943, 6, 1-16.
Spielberger, C. D. Anxiety and Behavior. New York:
Academic Press, 1966.
Spielberger, C. D., Gorsuch, R., & Lushene, R. State­
Trait Anxiety Inventory. Preliminary Test Manual.
�alo Alto, California: �onsulting Psychologists
Press, 1969.
Spielberger, C. JJ., Lazarus, R • .::i., Mordkoff, A. H., &.
Vavison, L, A. 'l'he experimental reduction or
stress based on ego derense theory. Journal of
1�bnormal and Social Psycholop;y. 1964, 68, 367-380.
�pielberger, C. D., & �mith, L. H. Anxiety (drive), stress,
and serial-position efrects in serial-verbal learning.
Journal of Exnerimental Psychology, 1966, ]g, 589-595.
Venables, P. H., & Hartin, I. .Skin resistance and skin
notential. In P. H. Venables and I, Hartin {l;!;d.)
I'-lanual or Psychophysioloeical Methods. Hew York:
John Wiley and �ons, 1967, 53-102.
Winer, B. J. Statistical Princinles in Exoerimental De­
sign. New York: .i,icGraw-hill JJook Company, 1962.
1·lolpe, J-., & Lazarus, A. A. Behavior Therany Techniques.
Oxford: Percamon Press, 1966.
 VI'l'A

Name: Thomas E. Narut

Birthplace: Chicago, Illinois
Date of Birth: February 27, 1935
Education: B.A., Southern Illinois University, 1961
M.A., Southern Illinois University, 1963
Grants: Division of Vocational Rehabilitation, Federal
Grant in Rehabilitation Counseling, 1961-1963.
United States Navy, Duty Under Instruction in
Clinical Psychology, University of Southern
Mississippi, 1967-1969.
Professional Experience: Counselor Trainee at Southern
Illinois University and Anna �tate Mental Hos­
pital, Anna, Illinois, 1961-1963.
Sociologist Field Investigator at Menard �tate
Penitentiary, Chester, lllinois, Summer, 1962.
Employment Supervisor (Industrial Psychology)
at A. �- �taley Manufacturing Company, Decatur,
Illinois, 1963-1965.
Clinical Psychologist for United States Navy at
Parris Island, �outh Carolina, 1965-1967.
Graduate 0tudent Clinician at University of
�outhern Mississippi, Psychology Clinic, 1967-
1969.
Clinical Psychology Intern, U. s. Naval Hospital,
National Naval Medical Center, Bethesda, Maryland,
1969-1970.
Head, Psychology Llranch, U. �- Naval Hospital,
Portsmouth, Virginia, 1970 to Present.

97