"“ravt ww-r-r-IVi-vvw-uwv --—-

 

SEX DIFFERENCES TN

PATTERNS OF [MOTION

ee of Ph. D.

gr

for the De

' . ‘ nation

NWERSITY
E NUTTALL

1.974

MICHIGAN STATE U
IOAN RA

 

 

 

 

 

 

 

 

 

ARV

Illl‘jl l‘llllllllllulllll ‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

thesis entitled
Sex Differences in
Patterns of Emotion
presented by

Joan Rae Nuttall

has been accepted towards fulfillment
of the requirements for

Ph.D. dpgephIPexoholofiy

 

 

ABSTRACT
SEX DIFFERENCES IN PATTERNS OF EMOTION
by

Joan Rae Nuttall

Previous research has shown that females begin themes with
negative emotion (deprivation), and move to positive emotion
(enhancement), while males produce the opposite emotional sequence.
The female response has been called a positive D/E pattern, and
the male response, a negative D/E pattern. Dissatisfaction with
traditional sex role expectations on the part of both males and
females has been associated with D/E scores which are most extreme
in the pattern typical of each sex. The present study was con-
ducted to assess the influence of instruction set, sex role dis-
satisfaction and biological sex on D/E patterns, and to investi-
gate the possible relevance of assumed similarity between self
and picture hero.

Sixty males and sixty females received either dramatic or
control instructions to determine if earlier findings were an
artifact of instructional set. Subjects were also divided into
low, moderate and high dissatisfaction groups on the basis of a sex
role questionnaire. Among females, D/E scores showed a marginal
increase in positivity with sex role satisfaction.

Subjects were also asked to choose the pictures which were

“7

Joan Rae Nuttall

£5? "most" and."least" similar to themselves. Males were expected to

0
U

tell positive D/E stories to the "most" similar pictures because
of high identification with the picture hero, and negative stories
to the "least" similar pictures because of dissociation of self
from the picture hero. No differences were found in D/E score on
the basis of similarity alone. However, a significant sex X sex
role dissatisfaction X similarity interaction was found, in which
the moderately dissatisfied subjects of both sexes produced the
D/E pattern appropriate for their own sex on pictures they judged
as being "most" similar to themselves. They told the pattern
expected of the opposite sex on pictures they judged as "least"
similar to themselves. This pattern was most clear among males.
Highly dissatisfied subjects told stories to pictures they judged
as "most" similar to themselves which obtained D/E scores expected
of the opposite sex.

A further analysis of the protocols involved scoring the
turning point in the action of the story for autonomy—heteronomy.
The majority of subjects told autonomous stories, and there was
some evidence of a mild picture effect.

Stories were scored for the nature of the final outcome.
Results showed that males ended their stories with more collapse
while females showed more evidence of ultimate hero resurgence.

The discussion emphasized that D/E sex differences are not
an artifact of instructional set, but that the relationship between
D/E scores and sex is not a simple one. D/E scores reflected

within-sex variations in degree of dissatisfaction with sex roles.

Joan Rae Nuttall
The behavior of the moderately dissatisfied subjects was inter-
preted as strain to conform to their own sex role model, and
the reversal among highly dissatisfied subjects reflected their

rejection of the sex role appropriate for their own sex.

SEX DIFFERENCES IN PATTERNS OF EMOTION
by

Joan Rae Nuttall

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Psychology

1974

ACKNOWLEDGEMENTS

I thank Kathy and Vik, my raters, who persevered when no
end was in sight. we spent many long evenings together, and
I wish you well.

I thank Jim, my husband, for being there to listen to my
own positive D/E fantasies, especially when I had not yet
reached the "pivotal incident" in completing this project.
Thank you for comforting me, and feeling good with me.

The members of my committee, Andrew Barclay, Eugene
Jacobson, Robert Zucker and William Crano, helped me organize
my thoughts and polish the pages found herein. I appreciated
your patience in waiting for some long intervals without overt
signs of progress.

And, I thank my subjects, who volunteered for that strange

experiment entitled, "Personality EXpression."

List of Tables. . . . .
Introduction. . . . . .
Empirical Evidence. . .
Stimulus Cues . . . . .

Within Pattern Analyses

Overview and the Pilot Study.

The Main Smdyo o o o o o 0

Method. . . . . . . . .
Results . . . . . . . .
Discussion. . . . . . .
Bibliography. . . . . .

Appendices

.A - Thematic Stimuli

B - The Scoring Systems.

0
I

U
l

Instruction Sheets .

Factor Analysis Data

OF CONTENTS

iv

10
25
37
1+5
51
55
66
86

10}

109
117
130

142

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

10.
11.

12.

111.

15.
16.

17.

LIST OF TABLES

Summary of stimulus cue value data. . . . . . . . . .
Mean D/E scores for pilot study . . . . . . . . . . .
Factor structure of stimuli from stimulus cue study .
Rater reliability/agreement for all story measures. .
Mean story length (words) for each picture. . . . . .
Correlations between story length and D/E score . . .

Mean D/E scores for instructions and sex (story-rate
and story-only cells) . . . . . . . . . . . . . . . .

Mean male and female D/E scores for all cells and
the mean across pictures. . . . . . . . . . . . . . .

Mean R—C scores for all pictures. . . . . . . . . . .
Distribution of sex role scores . . . . . . . . . . .
Mean D/E scores of sex role dissatisfaction groups. .

Three-way classification of sex role dissatisfaction
scores. 0 O O O O O I O I O O O O O O O O O 0 O O O O

Nban D/E scores for sex role groups . . . . . . . . .
PI classification for D/E groups. . . . . . . . . . .
Proportion of type of PI of sex X sex role groups . .
Frequency of PI classification. . . . . . . . . . . .

Mean D/E score of stories told to pictures judged
most and least similar to self. . . . . . . . . . . .

iv

31
1+7
65
68
69
7o

72

74
74
75
76

77

77
8o

81
81

82

Table 18.

Table 1 9 .

Table 20.

Table 21.

LIST OF TABLES (cont'd.)

Mean D/E scores of sex role groups on most and
least similar to self stories. . . . . . . . . . . . 83

Mean D/E scores fer similarity levels for
moderately dissatisfied subjects . . . . . . . . . . 84

Factor structure of stimuli from stimulus cue study:
Remainirlg pic-twee C I O O O O O O I O O O O O O O O 1A2

Mean ratings of pictures on semantic differential. . 150

INTRODUCTION

The outcome of a thematic story is traditionally considered
one of the most important aspects of protocol production. The
relative strengths of needs and press are ultimately inferred
from the hero's success or failure and the conditions under which
either fate is likely to occur. The precise meaning of an out-
come is nebulous without a corresponding analysis of the preced-
ing events, which not only provide a context for the outcome,
but are considered.meaningful in themselves as expressions of
the author's emotions, conflicts and needs (Murray, 1943).

Whereas content analysis is concerned.with motivational in-
fluences, structural aspects of the thematic apperception test
(TAT) protocol are considered reflections of thinking style and
the general modus vivendi of the author (Rapaport, 1942). Gener-
ally, structure has proven a more fruitful criterion for elicit-
ing sex differences than content, and has illustrated potential
individual difference variables dealing with ways of organizing
one's world, tendencies to attribute emotions and thoughts beyond
the immediate stimulus situation (weisskopf, 1950) and tendencies
to think in affective terms (Kagan, 1961).

.A structural analysis of emotion assumes that the manner in

which subjects deal with emotion in written protocols has meaning

2
relatively independent of story content. The pattern of emotional
movement in a story, i.e., the number and quality of references
along a time line, is designed to tap consistencies in the trajec-
tories of events imposed on the stimulus card. These trajectories
are "emotional paths" which do not have to vary appreciably with
the specific content of thought. To the extent that intra-individual
consistencies do exist in these trajectories, a structural analysis
of emotion can be said to measure affective "habits."

The idea that emotional experiences occur and change over
time according to a learned sequence implies that emotions are
not only experienced but are retrievable as memories separate
from the events that originally produced them. This idea is
described by Tomkins (1962) in his concept of emotional chaining.
The conditioning of emotions in chains occurs through repeated
contiguity of events which are unconditioned stimuli for affective
responses. .After the chain is formed through classical and higher
order conditioning processes, triggering of the initial emotion(s)
leads to the later emotions rather automatically. Through these
chains, which.may consist of only two links, emotions can become
functionally autonomous in Allport's (1937) sense of motive
independence.

Although many factors operate in emotional chaining and
several chains may fonm simultaneously on different levels, an
illustration will make the process more clear. Both Infant A
and Infant _B_ are distressed due to hunger. The mother of A quickly

provides the external stimulus needed to relieve distress.

3

This child learns over time that relief follows distress. Presum—
ably, of course, relief is the outcome in other need areas as well,
so that the pattern becomes established as a general phenomenon.
The critical aspect of the learning in the example is the formation
of the child's expectation that when he is distressed, relief
occurs.

Let us assume that the mother of Infant B waits a very long
time before attending to her child's needs. Bilearns that distress
leads to more distress, to more intense crying and to the ultimate
termination of the signals from exhaustion or inability to get
attention. In this case, distress leads to expectations of
failure and.more intense negative emotion.

The assumption here is that these learned expectations will
affect the cognitive interpretation of events which precipitate
distress later in life, as well as influence assumptions of personal
competence to overcome the emotion-producing events. For child
B, the precipitating events are catastrophic since his or her
experience has been that distressful occurences are seldom overcome.
The experience of child.A, however, should lead to the expectation
that the precipitating event can be overridden.

The individual's emotional expectations are here suggested
to play a part in determining the overall subjective probability
that a person will engage in any of a set of behaviors to obtain
a goal. Whereas past experience in success and failure and the
individual's perception of the success probability of a certain

behavior are commonly accepted determinants of whether or not

h

a person will perform a behavior (Atkinson, 1958; Rotter, Chance,
and Phares, 1972), emotional expectations, built up as underlying
assumptions about the way the world is, are simultaneously experi-
enced with personal success and failure. .As Infant Azlearns that
relief follows distress, he also learns that "I bring relief,"

or even "I am relief." Feelings of personal competence or incom-
petence and emotional chains thus reinforce each other. The emo-
tional "habits" that become developed, then, yield a particular
trajectory because the events with which they were originally
intimately tied actually occurred in that order.
Deprivation-Enhancement a§_Emotional Trajectory

Trajectories can differ in the quality of emotion, the quantity
of references to emotion, and the location in time of the references.
In order to describe variations in trajectories completely, all
three of these characteristics must be measured. The quality
and quantity are content-determined additive scores that necessitate
only general positive and negative scoring categories. The temporal
problem, however, demands some kind of dissection of the story
into temporal units, i.e., some reference point from which positive
and negative affective references can be tabulated and assigned
values to contribute to an overall trajectory score.

May (1966) introduced a scoring system which fulfills these
requirements, called "deprivation-enhancement" (D/E). Deprivation
refers to various types of physical and psychological discomforts,
while enhancement includes positive emotions, success and eminence.

The pivotal incident (PI) is the anchor point in time which

5
determines the weight to be assigned to positive and negative
references. The PI can be identified as the event which leads
to the final outcome in the story, the turning point in the action
(May, 1970). There are two models of emotional chaining which
serve as criteria in the scoring of D/E, although a large domain
of possible trajectories obviously exists. A.positive D/E sequence
is defined as a progression from deprivation to the PI to enhance-
ment, and the second model is the reverse, initial enhancement to
PI to deprivation. The latter is arbitrarily assigned a negative
D/E score. Any.D/E score represents the pattern of emotion refer-
ences rather than the absolute presence of either positive or
negative emotion per sq. Thus, a weight of +1 is assigned to
each deprivation reference occurring before the PI, AND also to
every enhancement unit following the PI. A.weight of -1 is
assigned to every enhancement unit prior to the PI AND to every
deprivation unit after the PI. The difference between these two
patterns is the final D/E score for any given protocol. Hence,
the D/E score assigned to a story represents the relative adherence
of the story to each of the criterion trajectories. This way of
viewing a D/E score is fruitful because very few "pure" D/E stories
exist, in the sense that the only scorable references that exist
are consistent with either pattern alone. Most stories involve
varying degrees of cancellation of scorable references on each

Side of the PI.

6

s a.Sex-Linked.Variable

Deprivation-Enhancement
The greater emphasis of positive D/E over negative D/E has
occurred most frequently and to the greatest degree among females,
both of college age (May, 1966, 1969), and through the middle
elementary grades (Cramer and Bryson, 1973). In these studies,
males consistently produced stories emphasizing the negative D/E
pattern, or stories which were significantly less positive than
those written by females. May (1966) suggested that the differences
in D/E emphasis parallel both physiological cycles and socialization
practices which differentially impinge on the sexes through develop-
ment, and hypothesized that D/E can be used as a projective measure
of feminine and masculine gender identity.
.According to this framework, a positive D/E score evolves
from life experiences which are most consistently feminine in this
culture. The important similarity across these experiences is
the repetition of a sequence of events wherein inhibition of
aggression and direction of activity inward leads to societal
approval and parental love. The female acceptance of pain and
suppression of self mediate the ultimate possession of joy, love
and opportunity. Concomitant with this sequence of actual events
is the development of the emotional chain of negative—to-positive
affective expectations. .According to N y (1966), the sequence
of inhibition followed by reward is also mirrored in the physio-
logical fUnctions of sexual intercourse and giving birth. The
contribution of these physiological processes is dubious, as will

be pointed out in the empirical evidence review; The fundamental

7

point is that the female is constantly reliving the deprivation-
to-enhancement cycle, which molds her assumptions about the ulti-
mate desirable consequences of psychological and physical discom-
fort, and which also imply the "price," to be paid in advance,
for enjoyment.

The factors in personal experience associated with the
negative D/E pattern (from enhancement to deprivation) are iden-
tical to those which produce the positive D/E pattern, i.e.,
socialization stresses and awareness of bodily processes. The
physiological experience of tumescence and detumescence and the
constant pressure to succeed and the fear of failure which males
experience in this culture lead to an emphasis on being successful
and on winning. Awareness of downfall, however, always accompanies
such emphasis on personal competence, either as a fear of failure
or just a realistic admission that an upward thrust cannot continue
indefinitely. This fear of downfall is pervasive, if not intense,
and becomes manifest in male fantasy as the ultimate defeat in a
negative D/E story.

One premise of this paper is that the pressure to succeed
which males experience in this society might well be as valid
if used to justify positive D/E scores. Pressure to succeed is
a negative state (and is scored as such in the deprivation-enhancement
system), and the ultimate successful outcome is mediated by movement
from the initial state of deprivation to one of enhancement. One
of the questions explored in this study involves the conditions

under which males might be expected to relate each type of story.

 

 

8

May (1966) has drawn theoretical parallels between positive
D/E and feminine masochism (Deutsch, 1944), and between negative
D/E and the classic myth of Icarus (Hamilton, 1942). In the female
case, the distinction between the deprivation portion of the fantasy
and the enhancement is blurred by the analogy to masochism since
the latter implies positive rewards arising from.the deprivation
itself. Questions can also be raised about the necessity of the
enhancement phase of the fantasy at all. May bridged the gap
between masochism and a sequence from deprivation-to-enhancement
by agreeing with Deutsch that female fantasy can be divided into
two parts, "...the negative tension and the subsequent release"
(Deutsch, 1944, p. 255). What appears to be occurring in this
line of argument is a declaration that the negative and positive
phases of a fantasy sequence are not separable, even though they ap—
pear to be, followed by the erection of an artificial barrier to
insure that such a separation can be achieved. The proposition
that some enhancement exists in the deprivation may well be valid,
but the author fails to see any greater explanatory power commanded
by invoking self—punitive tendencies than by the simple conditioning
of expectations of success to the goal-directed activities.

In the case of negative D/E, the prototype is the Greek myth-
ological hero, Icarus, who plummeted to his death after flying
too close to the sun with wings made of wax. Whereas the classical
myth is an interesting example of the negative sequence, the analogy
appears to have been an unfortunate one because of the existence

of an actual clinical syndrome of Icarianism.which was proposed

9

and.most fully elaborated by Murray (1938). Icarians are rela-
tively rare clinical types, who produce very bizarre and extra-
vagant fantasies marked by unrealistic goals and ambitions. They
commonly have fantasies of being female (Murray, 1938; Fried, 1972).
They are described as being very contemptuous of reality demands,
immature and impulsive. The downfall which besets the hero of
true Icarian fantasy is often unrealistically overcome, as through
reincarnation.

From.May‘s perspective, Icarianism properly reflects negative
D/E because of the widespread experiences of most males which
suggest a similar sequence, while he does not necessarily imply,
in the author's point of view, that most males espouse the remain-
ing characteristics of Icarianism. This interpretation of May's
viewpoint receives support from some empirical work reported in
a later chapter, and avoids the confusing implication that negative
D/E is simultaneously a mirror of basic sex differences in the
population and characteristic of the fantasy patterns of one of

the most feminine and extreme groups among males.

EMPIRICAL EVIDENCE

Replications of the D/E patterns include developmental studies,
construct validation studies using clinical groups, and replications
using both college and non-college individuals. This chapter
will begin with May's original studies and then expand to include
the clinical and developmental findings.

D/E studies have used TAT-like stimuli devised by May (1966)
or selected.pictures from the standard TAT set. Originally, May
developed four pictures, only two of which have been consistently
used. These are the first two pictures in Appendix A, coded and
described as follows.

.A. .A pair of trapeze artists performing
B. Bullfighter in full costume

The remaining four stimuli in the appendix were pictures select-
ed by the author for a pilot D/E study conducted in the Spring, 1973,
and.which were included in the initial picture pool for selecting
the stimuli for the present experiment.

E. A man and woman walking down a busy
downtown street

F. .A man and woman sitting together in
a quiet room

G. A.man and woman in eveningwear at a
fancy restaurant

H. .A teen-age couple gazing into each
other's eyes

11

There has been no consistent use of any standard TAT cards
in D/E research. Only Card 1 (boy with violin) has been used
more than once. Usually, protocols given to TAT cards have been
scored for D/E because they were already in existence, as with
various clinical groups. The majority of the results presented
here were obtained with Pictures A.and B, and the use of other
pictures will be noted when appropriate. The detailed scoring
method for D/E appears in.Appendix B.

May (1966) administered his four original pictures to males
and females from.two eastern colleges. The mean D/E score for
females was +.804 and -.235 for males. The sex difference could
not be explained as general optimism of females or pessimism of
males, since the total number of deprivation and enhancement units
did not differ between the sexes, nor was the difference a secondary
consequence of verbal fluency sex differences. The correlation
between length of story and D/E was +.l9 for females, and +.14
for males. Finally, the possibility that females simply preferred
"happy endings" was discounted since females related significantly
more deprivation than enhancement units prior to the PI than did
males.

.A replication of this original study was designed to investi-
gate personality characteristics of subjects who wrote extreme
positive and.negative stories. be'(1969) reported that extremeness
in D/E score, in the direction predicted for each sex, correlated
with the subject's sex role satisfaction, as measured by a lengthy

personal interview. .A negative correlation was found for males

12

between D/E score and sex role satisfaction; hence, these males
who were most dissatisfied with the male sex role told the most
negative D/E stories. May reported that during the interviews,
these males expressed envy of the female's greater warmth and
freedom not to compete. They felt they could not achieve societal
expectations for males.

The relationship between the females' interview data and
D/E score was not significant, but they related stories expressing
resentment about traditional female role expectations, and a dislike
of passivity. During the interviews, the dissatisfied females
expressed envy of the male's more active role in society, while
the satisfied females did not feel they were being pressured into
a traditional female role, but rather that society's expectations
were natural for them.and meshed.with their own personal desires.

From.May's perspective, the most dissatisfied individuals
tell the most clear patterns for their own sex because they feel
an acute strain to overcompensate for their dissatisfied feelings,
and this tension makes the cognitive patterns more extreme than
they might be otherwise. This conclusion was reached after May
made the assumption that D/E was a basic sex difference, and con-
stituted an attempt to incorporate the sex role satisfaction finding
into that framework. This only makes sense if one assumes that
most females fall into the misfit category and most men resemble
the insecure, tense males described in the interview material,
since this is the only way that the two D/E patterns can be

"characteristic" of the sexes. Furthermore, the fact exists that

13
the sex difference in D/E has been repeatedly found in previous
studies without categorizing subjects on sex role satisfaction.
This can only mean that the majority of subjects were dissatisfied,
or that the proportion of the sample that was dissatisfied produced
such extreme scores in each direction as to override the nondif-
ferentiated scores of the satisfied subjects.

The crux of the problem is this: if D/E is a basic and imr
plicit part of being male or female, why did the more satisfied
members of each sex fail to exhibit clear-cut examples of the
patterns? May (1969) completely avoided the fact that his "satisfied"
subjects did not produce significant sex differences in DVE, and
concentrated on how different the D/E scores were of the dissatisfied
subjects. 'Yet, the very fact that the satisfied subjects did not
produce significantly different D/E scores in the 1969 study clarified
the D/E variable. Rather than being a measure of a basic sex dif—
ference, D/E measured dissatisfaction with one's own sex. The
empirical data showing variations in D/E score with sex role
dissatisfaction provided evidence that May's theoretical ideas
about the origin of D/E sex differences are somewhat incorrect.
be'had.hypothesized that both physiological fhnctions and sociali-
zation experiences produced sex differences in the D/E variable.
Since physiological experience is constant across levels of
sex role satisfaction, physiology must be a less important deter—
minant of emotional pattern than social experience.

However, Winter (1969) reported findings which contradict

the suggestion that physiological functions have little to do

14
with variations on D/E scores. She obtained protocols from.women
while breastfeeding, and compared them to women the same age who
had breastfed, but who had recently weaned their infants. She
found that women who were currently breastfeeding produced stories
significantly more positive in D/E than the comparison group.
These results support May if one assumes that the currently breast-
feeding female is more cognizant of her distinctly feminine
biological activity than is the mother who has weaned her infant.
The importance of the ongoing nature of the breastfeeding activity
in yielding more positive D/E scores is puzzling for May's thesis,
because it implies that the memory of physiological activity no
matter how vivid, was not sufficient to maintain more positive
patterns. The women in the comparison group had weaned their
infants only two weeks before the study. If physiological experience
has an effect on D/E only as it occurs, the explanatory power
of the sex differences in sexual processes does seem severely
limited.

.A second possible explanation of Winter's data should be noted.
Both experimental and control groups in Winter's study were con-
forming to traditional sex role expectations, at least temporarily.
From May's interpretation of the sex role interview data, one
might hypothesize that the woman who is currently breastfeeding
may feel a temporary increase in dissatisfaction with the female
role and resentment of the increased demands made upon her. The
increase in dissatisfaction would then produce more positive D/E

scores 0

15

Further evidence suggesting that biological functions are
less important than sex role expectations was found in the
developmental studies. Three major experiments with children
were reported by May (1971), Saunders (1970) and Cramer and Bryson
(1973). They included samples of seventy-five, one hundred and
ninety, and eighty—nine children, respectively, and spanned an
age range from nursery school through eighth grade.

Both Saunders and Cramer and Bryson tested for age differences
as well as sex differences, and both reported a significant main
effect for age for females only. The main effect for sex reached
significance only among the older children, i.e., children falling
in an age range from eight years to eleven years, nine months.

Both studies reported significant sex differences for each picture
(Cramer and Bryson used TAT Cards 1 and 17BM as well as.A and B.),
as well as for overall mean scores.

The existence of sex differences in D/E in the child studies,
then, appears to be due to the increase in positivity of D/E scores
of females over time. The male pattern revealed no such age change.
Younger males either produced slightly negative scores (Saunders,
1970; Cramer and Bryson, 1973), or mildly positive scores (May,
1971). In either case, protocols written by older boys did not
increase significantly in the negative direction. Comparing the
scores obtained by male children with those of adult males from
May (1966) shows that both means are the same and slightly negative.
Since May only refers to molar bodily processes, such as puberty,

sexual relations and childbirth, as being analogous to the positive

16
D/E sequence, the occurrence of the female pattern by eight years
of age suggests that cultural norms are more important determinants
of D/E than physiology.

The fact that the female group is largely responsible for
the stable sex difference in child studies on D/E is unusual since
investigators of sex role identification and sex role preference
have reported that females are more unpredictable than males.

Brown (1957) developed the IT Test as a projective measure of

sex role identification and reported that female children are quite
divided in choice between male and female-associated objects and
behaviors until around nine or ten years of age, when a stronger
feminine identification appears. The male children showed very
early preference for masculine activities (Kagan, 1964).

The development of IT Test preferences closely resembles the
development of D/E differences, and occurs at about the same age
among females. Male children show an early and stable plateau
on both.measures. The fact that these variables develop so
similarly suggests that D/E may be more a measure of sex role iden-
tification than of gender identity. Whereas gender identity is a
term usually used to refer to the cognitive self-categorization
as a male or female, identification involves more of an incorpor-
ation of the behavior patterns and attitudes of members of the
chosen sex. The relationship between gender identity and sex role
identification is unclear, and constitutes an arena for a lively
debate within developmental psychology (Kohlberg, 1966; Unschel,

1966; Kagan, 1964; Freud, 1961), but most points of view basically

17

agree that the child has fonmed the necessary cognitive and
behavioral connections for both gender identity and sex role
identification by the time he enters school (Money, 1972; Kagan,
Hosken and watson, 1961). The absence of D/E sex differences and
the lack of consistency in IT Test preferences in girls until at
least third grade are incongruent with these theoretical points.
Kohlberg (1966) offered one hypothesis that IT Test inconsistency
among females is attributable to societal preferences for males.
Especially at the younger ages, children operate at a level of
cognitive development whereby a girl's choice to play with male
toys "makes" her a boy. The shift to feminine choices in later
childhood occurs when the girl realizes that females have subtle
advantages in the culture and she incorporates these advantages
as values. The IT Test has been criticized for its susceptibility
to social desirability, the influence of which could mask female
sex role identification of girls in the early grades. The absence
of D/E sex differences in the early school years might be attributed
to the young child's inability to meet minimal verbal fluency
requirements in the story-telling task. The increased verbal reper—
toire of children in the middle years might be necessary to produce
the individual variations in D/E score that reflect sex role identi-
fication, even though the identification had existed several years
before.

Ray‘s concept of gender identity, as measured by D/E, appears
to be a hybrid of the traditional concepts of gender identity and

sex role identification. To the extent that D/E would reflect

18
the person's experience of his or her own body, it is an inner
"awareness of self." D/E also reflects socially acceptable roles
in society, and to that extent it would be mediated by modelling
and reinforcement processes. In this way, D/E could be considered
an emotional correlate of identification.

In practice, May's own work suggests that the subject's level
of satisfaction with that sex role designated by biological sex
is more important than pure gender identity in predicting D/E
scores. N y (1970) predicted that homosexual males should relate
more positive D/E stories than heterosexual males because "...of
a more traditionally feminine role both in childhood experiences
and in adult fantasy" of homosexual men. The hypothesis was sup—
ported. The homosexual group yielded the most positive score of
any male group tested.up to that point, with a mean score of +.7l.
Yet, one cannot conclude that the homosexual men in this sample
lacked male self—categorization.

Furthermore, the conclusion does not seem.warranted that
those males in the 1969 study, who were classified as dissatisfied
by the interview data, and who related the most negative D/E stories,
had.more intense male gender identity than the satisfied group.
‘May did not describe those males who were "satisfied" as well-
adjusted, which has been the implication so far in this discussion.
Rather he referred to their lack of strain in being male as a
"bland and flaccid acceptance of their gender and the social expec-
tations that go with it" (May, 1969, p. 468). This placidity was

then associated.with a weak or intermediate D/E score. Presumably,

19
those males intermediate between "flaccid acceptance" and
"conscious strain" are most representative of the male population,
as are those females'halfaway between actively disliking the female
role and total contentment with it. Still, the reasons why those
males and females who are most incongruent with the sex roles
assigned to them by biological sex should.most clearly exhibit
the D/E patterns predicted for their own sex remain unclear.

One should note in passing here thathay's prediction of
more positive D/E scores from the homosexual males than the
heterosexual males supports the author's interpretation of the
relationship between negative D/E and Icarianism. Homosexual
tendencies are salient aspects of the Icarian personality, and
should logically lead to greater negativity if the entire Icarian
dynamic is applied to May's system. Hence, May's analogy to Icarus
must be specifically limited to the rise-andrfall quality.

One further finding from.the study with homosexual males is
important in view of the approach taken here that positive and
negative D/E patterns are models for a variety of subpatterns.

.A detailed analysis of the protocols focused on the sex of the
hero. Stories were categorized into "she," "he" and "they" groups.
Since only Picture.A (the trapeze artists) was used, each subject
had a choice between the male or female, or possibly a nebulous
"they" hero. The positivity of the D/E score increased linearly
with the femaleness of the hero, for all subject groups. The
reason why the homosexual male group received significantly more

positive EVE scores than heterosexual males was the greater

20
preponderance of "she" stories produced by that group, while
heterosexual males wrote more stories with the male artist as
the focus of attention. Also, heterosexual males wrote negative
D/E stories to "they" stories, while homosexual males wrote
positive D/E stories for nebulous heroes. A comparison of the
mean D/E score for homosexual males and heterosexual males on each
of these three groups of stories showed that the differences were
not negligible. The mean D/E score for homosexuals when the hero
was male was -2.00. When the story was classified in the nebulous
hero group, the mean was +.7l, and +1.80 when the hero was female.
The same means for the heterosexual males were —3.14, -3.33 and
+1.67, respectively.

The final study done by May (1970) was a construct validation
experiment with samples of hospitalized hysterics and obsessive—
compulsives, who wrote stories to Picture A.and took the Embedded
Figures Test (EFT) and the Block Design subtest of the Wechsler
.Adult Intelligence Scale as measures of field independence. May
hypothesized that the hysterical subjects would produce more posi-
tive D/E scores because hysterical traits correspond to exaggerated
stereotypes of feminine traits. The obsessive-compulsive traits
emphasize the male stereotype. Almost all the obsessive-compulsive
persons were males and the majority of hysterics were female.
Although mean D/E scores were in the predicted direction, the
variance was very large. The EFT also failed to differentiate
the sexes. waever, when the EFT scores (average number of seconds

required to locate the figure in twelve complex designs) of males

21
and.females divided into High and.Low D/E score groups were
compared, the results supported the predictions for field
independence. Namely, those females scoring most positively on
D/E were significantly more field dependent than females with
low positive D/E scores (n = 12, n = 10, respectively). The
results for men were in the opposite direction, as predicted.
The results for Block Design performance were Similar.

The final study in this section was based on theoretical
predictions which this author has already argued are incorrect
interpretations of May's position, yet the study contains some
information which is useful in this paper. Fried (1972) administered
Picture A (trapeze) to groups of diagnosed Icarian males (IM),
non-Icarian males (NIM) and sadomasochistic females (SMF), and
rescored their TAT protocols for D/E. He hypothesized that the
IM group should produce the most negative D/E scores, since they
were clear examples of the Icarian personality type. Less negative
scores should be produced by NIM subjects and the SMF group should
obtain positive D/E scores. In addition to scoring for D/E, Fried
also assigned a separate score for Resurgence-Collapse (RFC).

A detailed description of the R—C scoring system occurs in.Appendix
B. Briefly, RFC assigns an intense or mild score to the absolute
ending of a story, since Fried believed that the specific occurrence
of resurgence or collapse was the essence of what D/E was designed
to measure. Fried.hypothesized that SMF subjects should.write
stories receiving more Resurgence (R) scores and.males should

obtain more Collapse (C) scores.

22

The results showed that the female subjects did obtain the
most positive D/E scores, while the IM.and.NIM groups either did
not differ from.each other, or the IM.subjects received more positive
scores. Occasionally, the difference between IM and NIM group
means, for a given picture, was greater than that between males
and females.

The individual picture patterns did not support May's predictions
nearly as well as the overall means. Four of the standard TAT
cards (4, 11, 13MF and 14) produced no D/E differences at all,
while Pictures.A (trapeze) and B (matador) were associated with
the most clear-cut support of May's predictions. Both IM and
SMF subjects received.more Q_ratings than NIM males, and did not
differ from.each other, while IM males received the largest number
oflfl scores. These RFC findings show that positive D/E scores
do not automatically imply that resurgence is involved, nor do
negative D/E scores imply collapse.

Fried concluded that D/E sex differences are too dependent
on specific stimulus cards to be considered a general phenomenon.
He pointed out that selection of subject samples to underscore
factors that May considers basic to D/E did not magnify the D/E
effect, nor did it overcome stimulus cue factors. Fried suggested
that stimuli which elicit D/E have characteristics in common,
such as danger and depression, and that they are "loaded" in favor
'of Icarian themes, such as the flying themes elicited by Picture A.
When D/E differences did occur, Fried proposed that they re—

sulted from the alternation between the two complementary processes

23
of story—telling, the receptive and counteractive. These correspond
to acceptance or inhibition of the emotion in a story, and can
produce D/E differences because learned sex differences in empathic
ability exist. .A female is initially receptive of the dark, depres—
sive mood of the stimulus. She is more willing to operate on an
emotional level. Later in the story, the female comes to grips
with the emotion and.works through to a solution. This reversal
from passive reception to active working is the counteractive phase.
Since males learn to suppress feelings, they begin with the counter-
active phase. Continuous exposure to the stimulus, however, compels
them to recognize the depressing character of the situation, and
the "unhappy" conclusion is devised. Fried suggested that fantasy
is the same for both sexes, i.e., an alteration between reception
and counteraction, and whichever tendency occurs first in a parti-
cular case ultimately yields to the other. The stimuli which elicit
D/E focus on content areas where females are particularly receptive
and males are not; hence, these pictures consistently differentiate
the sexes by exerting control over which of the phases of fantasy
production occurs first.

In sum, Fried's study supports May's thesis in an overall
sense, but neither the D/E nor R—C analysis gave unqualified support
to the predictions. Fried considered his data more damaging to
May than this author is prone to do, because Fried relied heavily
on the literal interpretation of May's analogies to Icarianism
and.masochism. Given the combination of characteristics found

in true Icarian personalities, including homosexual tendencies,

2L1

Fried's result that IM subjects occasionally related more positive
D/E stories than NIM males was quite in accord with May's study
comparing homosexual and heterosexual males.

The stimulus card issue is an important one, and a chapter
is devoted to it later. Whereas several studies have investigated
the meaning of TAT cards, there have been no published studies
on the nature of Pictures A and B. A factor analysis of these
two pictures is part of a preliminary study to determine the semantic
structure of the pictures used in this experiment, and will be
discussed in the first section of the Results chapter. Maintaining
control over the stimulus cues in the pictures, is a more realistic
way of proceeding than denouncing D/E because it fails to overcome
stimulus differences, as Fried (1972) suggested. Such generality
is never found with psychological variables studied through the
thematic medium. The author doubts if Fried would reduce all
these measures to pseudo-variable status. Secondly, the criterion
that picture cues be irrelevant to the variable is equivalent to
demanding that changes in principle stimuli to which a subject
responds have no effect on the dependent variables. This criterion
is a stiff one for any branch of psychological research, and may

well be an undesirable one.

STIMULUS CUES

Clear stimulus-correlated differences in fantasy patterns
and relatively widely varying thematic baselines between pictures
necessitates a full understanding of the stimuli used to elicit
D/E patterns. Murstein (1963) suggested that the stimulus charac—
teristics are by far the most important determinants of the response
to a TAT card and this opinion is upheld by statements of other
TAT investigators (Eron, Terry and Callahan, 1950; Dollin and
Sakoda, 1962). Fortunately, attempts to scale the TAT stimulus
cue values for differences in pull for a variety of need states,
structural characteristics and emotional tones have been fairly
successful, and illustrate that the TAT is quite amenable to standard
scaling procedures. Given the large amount of story variance
potentially controlled by stimulus cues (up to half, according
to Murstein, 1963), such scaling attempts warrant constant use
to avoid confounding of stimulus cues with other experimental
manipulations.

The issue of scaling stimulus pull is complicated by inter-
actions between card values and subject variables (Clark, 1952;
Lubetsky, 1960). Two heavily researched examples of this inter-
action are relationships between stimulus pull and need state or

motivational arousal (Epstein, 1961), and between stimulus

26

characteristics and sex of the subjects (McClelland,.Atkinson,
Clark and Lowell, 1953). In the final analysis, a TAT response
can be considered a product of picture cues, subject variables
and background characteristics, such as instruction set.

The kind of complexity inherent in these interactions can
be exemplified by a brief discussion of the problems involved
in a stimulus cue and sex of subject interaction. .A long standing
assumption in projective theory has been that picture characters
which are similar to the subject in sex and age facilitate pro-
jection (Murray, 1943; Combs, 1946). Whereas this belief has not
been proven false, it has undergone considerable sophistication
and modification. Males have consistently illustrated greater
projection to same-sexed than opposite-sex heroes (Murstein, 1963),
but the reverse is not always true of women and depends considerably
on the particular variable being measured. In the literature
on need for achievement (nAch), for example, high scores have
been obtained by women more readily when the scene included a
male rather than a female (Veroff, Wilcox and Atkinson, 1953),
even when the female character was depicted in an achievement-
oriented scene. Evidence suggests that this result can be modified
by the quality of the achievement scene and the education level
of the female subject (Veroff, 1961). Part of the problem with
assessing this issue is the large between-card differences in
achievement cues even within the sexes, and the lack of precise
knowledge of just how equivalent the parallel forms of some pictures

are when the people depicted in them are the opposite sex.

27

Overall, studies of projection of both sexes to figures of
both sexes shows that females often project more to male figures
than to female figures, while males seldom show as much projection
to female heroes. McIntyre (1954) used similarity of answers
given to the Minnesota Multiphasic Personality Inventory for the
self and for characters depicted in short films as an index of
projection and found no evidence for greater projection onto same—
sexed figures for either males or females. In a somewhat different
design format, Silverstein (1959) corroborated McIntyre's findings.
Both authors reported that a more important variable in determining
the degree of projection exhibited by either sex was the similarity
of the depicted scene to their everyday lives.

Both Lubetsky (1960) and def/harms, Morrison, Reitman and
McClelland (1955) reported that females exhibited greater projection
to the male heroes than to the female heroines. Lubetsky noted
that projection of female subjects remained constant regardless
of the age of the male hero, whereas projection was negatively
correlated with age when the heroine was the central character.

In a study specifically of women's responses to the TAT, Eron

(1953) found that the majority of stories written to cards depicting
both a male and a female (notably TAT Cards 4 and 13MF), centered
around the male figure. Two judges categorized each story for

sex of the central character and found, with 94% agreement, that
26% of all stories written to Card 13MF centered on the female.

Eron concluded that a list of apperceptive norms for female subjects

may ignore story hero differences, while this is not true for

28

male subjects.

Perceptual prominence of male and female characters in pictures
is conspicuously absent from discussions of hero choice. The female
character in Card 13MF is not as perceptually salient as the male,
while the female in Card 4 is more dominant than in 13MF, but
still somewhat secondary. Eron's data does suggest that female
subjects more often choose the female character when that character
is prominent in the stimulus, but that they definitely do not
"search out" a perceptually obscure female character and formulate
their stories around her. Males may exhibit a similar behavior,
which is overshadowed by the larger number of male-dominant TAT
cards. This would mean that males are not more sex-specific in
their choice of heroes, but merely responding to the dominant figure.
An example of a TAT card which features a dominant female is Card 2,
and the male themes for this picture suggest that the female
character is the main figure in the stories, i.e., themes of con-
flict between educational aspirations and parental conflict are
popular.

Whether or not sex differences are real in hero choice, there
are replicated stimulus cue effects and stimulus cue X sex of subject
interactions for measures of emotional quality. Emotional tone has
been one measure of emotional quality, usually rated along a five—
point continuum from very happy to very sad (Eron, Terry and
Callahan, 1950; Eron, 1953). The ratings have been determined
either directly, using ranking procedures and rating scale methods,

or more indirectly through the kind of apperceptive norms associated

29

with each stimulus picture. The direct methods require subjects
to make decisions concerning the properties of the cards themselves,
and either to rank them on some dimension or rate them on the cri-
terion of interest for relevance, direction and degree. The
indirect approach assumes that the emotional tone of the stories
told to each stimulus reflect the subject's perceptions of the
stimulus cue characteristics, and relies on measures of the emo-
tional tone obtained through judges' agreement in rating the Stories
on a relative scale of emotional level.

Results obtained using direct rating procedures suggest that
subjects can agree quite unanimously on the objective nature of
the stimulus characteristics, despite the wide range in ambiguity
of the pictures (Bijou and Kenny, 1951). Correlations between
rankings of males and females on a variety of objective dimens;ons
range from +.75 (Bijou and Kenny, 1951) to +.96 (Newbigging, 1955).

Despite high similarity in ratings, significant sex differences
have often been found in emotional tone of stories written to these
same pictures (Eron, Terry and Callahan, 1950; Sarason and Sarason,
1958; Newbigging, 1955). All these studies report that females
tell sadder stories than men. Sample or picture differences cannot
account for the discrepancy between ratings and stories on emotional
tone, because both measures have been taken with the same subjects
(Newbigging, 1955).

Newbigging's results were obtained under instructions to tell
"happy" stories. The response time for males increased as the happi-

ness rating of the picture decreased, and males also showed

30
tendencies to avoid definite outcomes in their stories under the
sadder stimulus conditions (see also Lindzey and Heinemann, 1955).
These results suggest that males experienced greater difficulty
with the more negative stimuli, and that females did not remove
themselves from the nature of the stimulus to the degree that
males did. Possible reasons for this difference include the
hypothesis that females are more stimulus—bound than males, or
that they are more willing to accept the negative emotional situa—
tion than males. Fried's (1972) reception-counteraction analysis
of the story—telling process is congruent with the latter. The
case for stimulus binding is supported by normative data on the
TAT, since large samples of college men and women judged the
stimuli to be sad and depressing. The sadder female protocols are
more accurately reflecting the stimulus nature of the material.

The fact that males relate happier stories than females does
not imply that the pictures rank differently for each sex, when the
rankings are determined from story analysis. Table 1 illustrates
the consistency in judgments of pictures, using both direct and
indirect ranking methods. Only those stimuli which have demon-
strated the capacity to elicit D/E are included in the table.

The nature of the agreement between the studies listed closely
parallels intuitive logic. Stimuli judged least sad in several
studies, for example, were the same stimuli which Lebo's subjects
reported liking the most. Pictures judged least sad in the Eron,
at al. (1950) study elicited stories with positive emotional tones

in studies by Eron and Garfield (1948) and Eron (1953).

31

 

.oposoossopsH

.mgamn mho>

.mmmcflmaeg so mPMHUmEMmPGH

 

.mLSpOHQ pmmflaaam

.mmme 35m mmHsSom
.mmxmm neon how hands hhm>

.mhdpofim pmoflmmmm

.mwxom neon an mmflsopm m>flpmwoz
.Qdosw pmmpoem exp mo stEmz

.Qdosw pmopomm one mo soQEmz

.moflsopm magmas: pmeoq

.thVOHQ has wo
emanm o>HPflmom pmmemopw

.osmn o>flmmma hho>

 

.omm mpo> .pmm zeo> .mmflsopm pew hso>
.moxmm upon
90% mmmsflgmmn no opmflvmsuwch
.cmsoz so.“ memfiUmEopsH .Cma
pom mouse pmoflmmdn one mo oso
Apoosaav AeoossooHv ApoossosHv

Amomev oooxsm nos season

Aommev ssssaaso one assoe .sosm

mofiwdpm

Awsmsv sosm one sashesoo

or

\x

we

"m\q wasps
1oflam poz

zmwe
as

mama

2mm

a
um\q
mosesosfim

HHSEHpm

 

spam ooHo> one moassspm so ssossom

F m4fl46

32

.Ammmr

.oflopmsdzv oogpoafipmsfi maaopop ms 9.“ 3365.3 .mdodwfinfim you me 0.. peso pone memowwdm

moosdom honeo 89C seaweed/m

Chosen Phenom one .90me CH pensions so: was or p.30 .39

n

.mwm we: modem macaw no poms mSHSEHPm one o

 

 

Tr YT QéOHPmEomcH oz .oPMfiUogoch
m+ e1 .mfiflmflnfim so pmmflmflm .opmflooEmpsH
_.+ o .mpgmsnfim :0 30A .wsflofla swam
e+ a-.o .spsoMsoso so sod .mssssa swam
m1 m- .opssoossopsH .oxsuneo swam
I. e1 . mpmflp men 3.5,. . weeds och och
.seeowsnso
oe+ a. so swasuosoeoossopoH .oeasnso swam
.hpfism
e+ a. lapse so someooEsopsH .mseesa swam
F+ s1 .ssssmeQEo so sou .mosgsa swam
9:03.50 mcoe HdQOHpofim
Apomsflofiv 30933 Apomsav
Ammmev oosm Ammmev endow ono soeem Ammmev onoq

m mMUSPm

or
S.

”m\q moses
nosam poz

ems?
as
same
a

2mm

um\n
mosesosam

adsEHPm

 

A.o.pdoov r mamas

33

The last column on the second page of Table 1 summarizes
Eron's (1953) results for "emotional shift." Emotional tone
ratings were obtained by a general rating of the whole story, but
a separate rating was assigned to the outcome itself, and together
these ratings were called "shift."

Emotional shift is a primitive D/E measure. The most common
female pattern, as shown in the Table, was a shift from a slightly
negatively toned story to a slightly positive outcome. 0f the
stimuli listed, only Card 13MF ends with strong negative affect.
No data exist for males, since the Eron study was done with females
only.

The difference in emotional tone of stories told by males
and females suggests possible hypotheses concerning the nature of
the D/E variable. If males tell happier stories than females do,
even though both sexes agree on the objectively "sad" nature of
the pictures, the differences in initial starting points of stories
may account for the first half of D/E differences. Nbles begin
with enhancement and females begin with deprivation.

The importance of reviewing studies on TAT card cues lies in
their use as guidelines to assess the nature of stimuli eliciting
D/E differences. Control of the cues is desirable not only as a
technique for controlling variance in pull, but also to provide
clues as to the nature of real life situations which may trigger
D/E kinds of cognitive sequences.

The task at this point is to isolate similarities among the

standard TAT cards which have elicited D/E differences. If such

54
similarities can be found, the trapeze and matador pictures should
rate highly on the same salient dimensions, and perhaps even more
highly to be commensurate with the degree of D/E differentiation
of which they have proven capable. An attempt to find patterns in
semantic meaning among pictures will be the focus of a preliminary
study.

For high confidence to be placed in the patterns which might
come out of this preliminary study, not only should the trapeze
and matador pictures rate highly along the same dimensions as the
TAT cards which have successfully elicited D/E differences, but
stimuli with little power to elicit D/E should not rate highly.

The problem with establishing discriminant validity in this case
is the severely limited number of TAT cards which have been used
in D/E studies.

The first step in accomplishing the task of assessing stim-
ulus cues is a close examination of those TAT stimuli which have
elicited the strongest patterns. Cards 1 and 3EM.have been most
differentiating, while Cards 2, 4, 14, 13MF and 17EM have elicited
smaller pattern differences (Fried, 1972; Cramer and Bryson, 1973).
The D/E differences in these cards is relative, so that the female
mean to Card 13MF in Fried's data was almost at the zero point
(+.27), while the male mean was strongly negative (-2.00 and -1.42
for the NIM and IM groups, respectively).

The studies in Table 1 provide some information on the semantic
meaning of stimuli which have elicited D/E, for ratings of happiness,

emotional tone, liking and ambiguity. The stimuli cover the entire

35
range of happiness—sadness and of liking, while most of them seem
low on ambiguity. These studies strongly suggest, therefore, that
the stimulus cards do not have to be liked (or disliked), nor do
they have to be either happy or sad in order to elicit D/E dif—
ferences. A study by Goldfried and Zax (1965), which was the most
inclusive of all the experiments on stimulus cues, supported these
conclusions. These authors used the entire set of TAT cards,
except Card 16, and obtained subject ratings for each card on ten
adjective dimensions. Male and female ratings were kept separate,
but the patterns in the ratings were similar for each sex.

The objective ranking data on semantic dimensions from Gold—
fried and Zax' study and the others show that the stimuli from the
TAT which have elicited D/E sex differences fall into distinct
categories. The first group includes Cards 14 and 17EM. Both
stimuli are pleasant, liked, and usually elicit positive or
neutrally-toned stories from both sexes. Both cards elicit thema
concerning power relationships with the world, success and failure,
ambition and control over environmental forces.

The remaining cards in the "D/E TAT" -—-Cards 1, 38M, 4 and
13MF --- yield an impression of sadness. All four rate significantly
on the same dimensions, although not in a uniform direction. These
continua are aggressive—passive, impulsive—controlled and pleasant—
unpleasant. Whereas Card 1 and 13MF and "impulsive," both 33M
and 4 receive "controlled" ratings. Cards 1 and 38M are "passive,"
and Cards 4 and 13MF are "aggressive." This second group of stimuli

appears to be quite sad in tone and to emphasize concerns with

56
personal aggressiveness and impulsiveness.

In sum, both sad and relatively less sad stimuli have
elicited D/E differences, but all the stimuli seem concerned
with issues of personal control or personal effectiveness in
the world. The degree to which Pictures A (trapeze) and B
(matador) can be meaningfully described by the same dimensions

is an empirical question.

WITHIN PATTERN ANALYSES

The theoretical position espoused in the initial chapter of
this paper is that the positive and negative D/E patterns are two
model emotional chains which represent, to varying degrees, a
larger number of possible emotional chains. Evidence supporting
this viewpoint showed that meaningfu1 within sex variations can
be discerned (May, 1969). Interview material on sex role satis-
faction differentiated the high negative scoring males from those
who received scores closer to zero. The females in May's experi-
ment who received the most positive scores related more themes of
dislike of passivity, of resentment over the female role, and envy
of men. Females who scored more moderately on D/E related themes
in which the story characters were less concerned with sex roles
and accepted societal expectations.

Although a dissatisfied female would seem to possess a very
different kind of sex role attitude and belief about what her
gender identity implies for her than the more satisfied female
would, the data suggest that this difference is reflected only in
relative amounts of positivity on the D/E scale. Actually, the
number of deprivation and enhancement units is the only measure
that has been scrutinized, so no information about other possible

differences exists.

58

The basic characteristic which differentiated satisfied and
dissatisfied females was their attitude toward passivity. The
dissatisfied woman told themes in which she expressed dislike of
having to wait for men to do things for her, while the satisfied
female expressed a belief that there is a certain virtue in being
passive, in accepting rather than initiating, and in not "getting
worked up" over things.

The most likely candidate for measuring different beliefs
about passivity is the PI, since the PI is that point in the story
where the subject must make a decision about the character's role
in.mediating the outcome. The subject must decide whether the
story hero will engage in some action to bring about the outcome,
or if some event will befall the hero which automatically creates
the outcome.

Attributions to personal causes imply that the impetus for
behavior was generated from.within the actor. Presumably the actor
intended the behavior, and hence a unit relation occurs between
the actor and the event with its consequences. From an observer's
point of view, a unit relation often occurs because the person
whom the event affected is assumed to deserve his fate (Heider,
1958). This unit relation was the foundation for the later elabora-
tion of the "just world" hypothesis (Lerner, 1965), which may be
defined as the belief that individuals deserve the events which
befall them,

Of course, there are degrees of personal causality, in which

relative amounts of responsibility can be assigned to personal and

39
environmental forces in producing effects in individual cases
(Heider, 1958). ‘Variables such as effort and luck can determine
the nature of an attribution (weiner, Frieze, Kukla, Reed, Rest
and Rosenbawm, 1971), but Heider (1944) also suggested that indi—
viduals have "attributional styles" from which they consistently
create certain causes to make sense of incoming information With
the least amount of disturbance to their cognitive system. Kelley
”(1972) also suggested that immediate attributions are largely a
function of the kind of attributions one has made in the past.

The importance of personal causation and just world beliefs
in this paper is their utility in analyzing the proposed differences
within the D/E pattern between females who are satisfied and dissat-
isfied with the female role, and the extent to which the female
author may see herself (or the story hero) as instrumental in
achieving the final state of enhancement.

PI's of satisfied and dissatisfied females are assumed to re-
flect differences in their attributional styles. When the thematic
story begins on a note of deprivation and the hero engages in some
personal act which causes enhancement, the author implies that (1)
deprivation is an undesirable state, and (2) the hero can, and does,
succeed in overcoming the negative affective state in favor of a
more pleasant set of circumstances. The female who feels more come
fortable with the traditional role should manifest her acceptance of
passivity in her attributional style and her PI choice. She should
relate more stories with heteronomous PI's, even though the emo—

tional pattern of her story, like that of the dissatisfied female,

40

is still positive in D/E score.

The implications of the heteronomous PI for the author's
attitudes about deprivation and enhancement differ from those of
the autonomous PI. The hero who exhibits passivity in the face of
deprivation accepts the deprived state, although not necessarily
as inherently desirable or for masochistic reasons. ‘A theme with
a passive hero and a heteronomous PI leading to enhancement resembles
the traditional female fairy tale. This story type is epitomized
by the standard female "Cinderella syndrome," in which some lowly
working girl is one day carried away by a prince charming (who
just happened by). The state of initial deprivation is glorified
by the assumption of just world. Those who endure are worthy of
praise, proven by the simple fact that a highly beneficial chance
event has befallen them. The central character in heteronomous
D/E stories is passive, awaiting the whims of the peripatetic
prince, and is rewarded for the acceptance of the deprivation which
made it all possible.

Two different D/E patterns, then, are proposed to exist in the
female population. The pattern where deprivation is terminated by
a personally caused PI is proposed to characterize the female who is
dissatisfied with traditional female passivity, and a positive D/E
pattern with a heteronomous PI is characteristic of women who find
the female role more acceptable.

A.measure for assessing sex role satisfaction will be used that
is based directly on the dimensions which May (1969) used to describe

dissatisfied.males and females. The dissatisfied males focused on

41
their dislike of the pressure to succeed, while females emphasized
resentment of the passivity expected of them. The questionnaire
used in the present study is included as part of Appendix C, and
is labelled as P7. A detailed presentation of the scoring system
appears in Appendix B.

Applying this analysis of PI differences to the male pattern
yields a less clear cut distinction between personally caused and
heteronomous PI's. According to just world, the occurrence of a
chance event leading to deprivation implies that the here was de—
serving of downfall, because of a fatal character flaw or perhaps
a sin of omission. A.personally caused PI leading to downfall
ultimately implies the same unworthiness, although the nature of
the flaw would probably be ignorance of the facts, a flaw in logic
or mental derangement.

Since the PI itself has not been studied in previous experi—
ments, the feasibility of scoring D/E stories for personal and
heteronomous causality was determined by reviewing pilot data on
D/E and classifying the PI's into personally caused, heteronomous
and doubtfhl categories. The majority of PI's were clearly classi-
fiable. Examples of the autonomous and heteronomous PI's obtained
from this analysis appear in Appendix B. This list was used to
aid the raters in the main study in understanding the nature of
personal and heteronomous PI's, but more emphasis was placed on
rating the entire story than on blanket scoring of the PI's
according to the list. Raters assigned an AutonomouséHeteronomous

(AAH) score to each story in the main study, according to the scoring

2+2
system described in.Appendix B.

The downfall of the hero in a negative D/E story suggests that
males are dealing with a certain kind of deprivation in their pro-
tocols, more of a "trait deprivation," which will be defined as
the attribution of causality for the downfall to some inherent
flaw in the hero himself. Females seem.to be describing what ul-
timately proves to be "state deprivation," which is situational and
overcome either through personal or heteronomous causation. To
make a positive D/E sequence meaningful, the initial deprivation
experienced by the hero must be temporary and not due to any dlSpO—
sitional source. Hence, the story author maintains hope that the
future will improve because the hero possesses the inner capacity
to elevate himself or herself to enhancement, or because the hero
is worthy of the heteronomous event promised for inner goodness.

Attribution of trait versus state deprivation is crucial in
understanding one hypothesis for the sex difference in D/E scores.
Dispositional attribution most clearly enables the subject to
dissociate himself or herself from the actor, while state depriva-
tion does not imply a necessary distinction between self and other.
This is a restatement of the defensive attribution hypothesis
(walster, 1966), where the observer of another's fate can either
assure himself that the same will not happen to him because the
other had some personal trait which caused the bad luck or made him
worthy of the bad outcome, or the observer can convince himself that
similar good luck may be his any minute because he and the recipient

of the good luck are very much alike. Ultimately, then, both the

43
positive and negative D/E patterns reflect the story author's
desires for the future to be an improvement over the present,
although they are very different ways of expressing this feeling.

In order for attributional events to explain sex differences
in D/E scores, a sex difference must exist in the degree to which
the story author assumes similarity between himself and the story
hero. Since the story hero is a static figure in a picture, there
is very little objective evidence for assuming either similarity or
dissimilarity. The subject's attributions, then, are basically
projections of the self.

Evidence exists showing that females assume greater similarity,
i.e., project more, to the picture characters than males do, and
indirect evidence also exists that females either perceive less
trait deprivation than males do, or are less ready to admit it.
women, for example, have been reported to assume greater similarity
to others than males, as measured by the degree of parallelism
between "self" ratings and "other" ratings, to describe others more
favorably, and to differentiate less among significant others than
males do (Smith, 1963; Kohn and Fiedler, 1961; Shapiro and Tagiuri,
1959). In the Kohn and Fiedler study, males were more willing to
express strong negative affect towards others and saw themselves as
quite unlike others on the Construct Repertory Test.

In thematic studies, several projection indices suggest that
females do become more involved in the procedure of telling stories
(Murstein, 1963). Females produce consistently higher output in

terms of story length (Lindzey and Silverman, 1959; Sanford, 1943;

44
Lindzey and Goldberg, 1953), and obtain higher scores on indices
which depend indirectly on verbosity, such as the Transcendence
Index (TI) (weisskopf, 1950; Prola, 1972). Although verbal fluency
is undoubtedly a looming factor in TI scores, high scores also
reflect egocentric perception and the assumption that the situation
is the same for the other as for the self, and a greater consequent

willingness to attribute emotion and cognition to the other.

OVERVIEW.AND THE PILOT STUDY

,A total of three studies were conducted that relate to the
theoretical ideas that have been presented. First, a pilot study
was run, for three purposes: to give the author practice in
scoring TAT protocols for D/E, to replicate the sex difference
found in past research, and to explore the possibility that
pictures emphasizing male-female relationships might increase the
sex difference in D/E scores. This prediction was made because
opposite-sex relationships constitute one area where a person is
made very aware of his or her own sex. Using pictures in which
these relationships are salient may be one way to elicit "male"
and "female" ways of feeling and reacting. The results of the
pilot study will be discussed shortly.

The second study was a factor analytic study, conducted eight
months later. This study was concerned with the semantic meaning
of a large pool of pictures of potential use in the main D/E study.
The semantic differential ratings obtained in this experiment were
examined to help select the pictures used in the main study, and
to provide information about the meaning of pictures that have been
used in previous D/E work. The method for the factor analytic
study is included in the Method chapter, and the results are re—

ported in the first section of the Results chapter.

46

The main study directly tests a group of hypotheses concerning
the effects of instructions, sex role dissatisfaction and the sub—
ject's assumed similarity between himself and the picture hero on
D/E scores, RFC scores (Resurgence—Collapse), and the PI's. The
detailed description of the method for the main study appears in
the Method chapter. Briefly, subjects wrote stories to four
thematic pictures, rated the pictures on similarity to self, and
filled out a sex role questionnaire. Ratings of similarity Were
used because past attempts to control investment of self by men
and women in picture heroes by using same-sex figures have not
always been successful. Thus, subjects were asked to judge each
picture on similarity to self. Hypotheses concerning similarity
deal only with the stories written to the pictures judged as "most"
and "least" similar to the self.
Eh§_pilgt.§tugy. In the Spring of 1973, forty—seven undergraduate
volunteers at Michigan State University wrote TAT stories to a set
of six thematic pictures, including May's Pictures A (trapeze) and
B (bullfighter), and four other pictures chosen for this study.
These pictures are coded EAH in the Appendix. The pictures were
collated in several random orders, with story sheets interspersed
between them. Instructions followed the procedure outlined by
McClelland, Atkinson, Clark and Lowell (1953).

Protocols were scored for D/E and later rescored for R-C.
Mean D/E scores for males and females for each picture are given
in Table 2. The mean D/E scores from.May (1966, 1969) on Pictures

,A and.B are included for comparison.

47

 

 

mm. + 0:. + mm. + no.7 No.1 moé+ mm. + or. + mmagmm
mm.n- as.s- mm. - mm. - as. + mm.s+ as. + es. + moans
AmmeHoocosv AmmoaaooV m w m m m .<.
Mom
enough as: ngm

 

 

Rom. 68: so: eons Boos oassssdsoo fies
bosom soda tom 3.88 a} some

N wanna

48

The pilot study did not support May's earlier findings, either
over all pictures or for individual pictures. Only Pictures F and
H elicited even a trend in the predicted direction. Also, the
hypothesis that D/E sex differences would be increased by direct
reference to male-female relationships was not supported.

The mean age of the males in the pilot study was 21.9 years
and the mean female age was 20.0 years, both equal to May (1969).
The only factors which.made the pilot study procedure different
from.earlier work were the inclusion of additional pictures and
the instruction set. May instructed his subjects, in both the 1966
and 1969 study, to write stories which "are as dramatic and psy—
chologically meaningful" as possible.

The effect of dramatic instructions might well be to establish
an extremity response set in the subjects. If so, the D/E dif—
ferences which May found were apparent only because the subjects
were asked to invest effort in the creation of a special sequence
of events, and would not necessarily appear when such a response set
is not evoked. Thus, one logical hypothesis for the lack of support
in the pilot study for May's predictions is the change from dramatic
to control instructions.

The hypothesis that D/E differences are created in part by the
instruction set is supported by the fact that the major problem
with the pilot data was the large number of very bland, actionless
stories. Many of these protocols were difficult to score, just
because of the absence of a clear cut PI. Instructions to be

dramatic may serve to enhance D/E differences because they yield

49
a higher percentage of scorable stories with definite PI's and
outcomes rather than through the content of the instruction set
2££.§§°

The blandness of the protocols was also a major problem when
the stories were rescored for Resurgence-Collapse (R—C), according
to Fried (1972). .A very large number of stories obtained "0"
ratings, one hundred and eight out of two hundred and eighty-two.
Seventy-five stories were not scored for R-C at all because of a
lack of story ending or a very ambiguous ending. Thus, the pilot
data cannot provide much information about the relative value of
D/E and RFC scoring systems, except to illustrate most emphatically
that the task of scoring either D/E or RFC is most difficult when
the protocols lack dramatic action. This conclusion maximizes the
crucial nature of the instruction set.

Further evidence for the importance of the instruction set
comes from the fact that sex differences in EVE, under dramatic
instructions, are consistent with the finding that females tell
sadder stories than males. May (1966) had found that females told
significantly more pre-PI deprivation units than males, and.males
told.more pre-PI enhancement units than females. At that time, this
fact was used to argue against the hypothesis that D/E scores re-
flected general attitudes of optimism or pessimism. However, since
the PI usually occurs toward the end of a protocol, the emotional
tone sex differences are congruent with the pre-PI units difference.
This occurs since the major portion of the protocol is prior to the

PI, and this same portion contributes the most to an overall rating

50
of emotional tone.

The first hypothesis of the present study is based on the idea
that D/E sex differences and emotional tone differences are iden-
tical, and the post-PI reversal in emotion is the subject's attempt
to conform to a dramatic instruction set. If a subject initiates
a story with deprivation, enhancement is the only ending which
satisfies the instructions, and the same is true of a story which
begins on a note of enhancement. The hypothesis is made even more
plausible by the lack of support for D/E differences in the pilot
study, where stories were obtained under control instructions.

This hypothesis basically questions whether the appropriate unit
for study is the entire pattern or the emotional starting points of
males and females. If the hypothesis presented here is supported
by finding significant sex differences in D/E only under dramatic
instructions, then the sex difference to be studied is the initial
perception of the picture, and the rest of the D/E pattern is

artifact.

THE MAIN STUDY

Statement 2: the problem. The hypotheses of the main study address
three different issues. The mechanical explanation of D/E dif-
ferences focuses on the initial perception of the stimulus card
and the accompanying demand from the instruction set to write
dramatic stories. This hypothesis implies that the D/E variable

is best studied by separating it into components, pre-PI and
post-PI, and dealing with each as a relatively self-contained unit.

The second hypothesis assumes that D/E is a pattern which must
be studied as a whole. Here, differences in D/E scores are hypo-
thesized to reflect variations in the subject's identification with
the hero, and conversely, the degree to which the subject might
actively disengage himself or herself from the picture character.
Assumed similarity to the self was proposed as a measurable indicator
of this process.

The third issue concerns the subject's feelings about his or
her own sex role and the relationship between these feelings and
the type of events chosen to mediate stories. May (1966, 1969)
suggested that females learn to be passive and introspective because
they learn that suppressing themselves leads to reward. Further,
females who are unhappy with the female role resent the passivity

expected of them. This belief in passivity or activity should be

52
reflected in the nature of the PI, independent of D/E score, since
the story author should posit a sequence for the story hero that
reflects the way she views the world. Hence, a female who feels
comfortable with the passivity should write stories wherein the
hero is passive, and the events which occur are heteronomous, or
externally-caused.

The highly dissatisfied females, who by definition are re—
jecting the traditional female self-inhibition and are more emphatic
about personal activity, should mediate their stories with personally
caused PI's.
flypgtheses. The specific hypotheses of the present experiment can
be stated as follows:

1. The sexes will differ in their initial perceptions of the
stimulus. Females will write stories with significantly more de-
privation units and significantly fewer enhancement units prior to
the PI than will males. This prediction is based on emotional tone
differences found in earlier work, and on May's (1966) work report-
ing similar findings.

2. Initial perceptions of males and females will interact with
instructions to yield the D/E sex differences. Specifically, sig-
nificant sex differences in D/E will exist only when subjects are
given dramatic instructions and not with control instructions, where
subjects are asked just to write stories. This hypothesis is based
on the absence of sex differences in the pilot data, and the propo-
sition that dramatic instructions are a demand characteristic that

produces the reversal in emotion after the PI.

53
3. In addition to D/E sex differences under dramatic in-

structions, Resurgence—Collapse (RFC) will distinguish the sexes,
so that females will obtain significantly more 5 scores than males
and.males more Q_scores.

4. Dissatisfaction with sex role, as measured by a sex role
dissatisfaction questionnaire, will be positively related to ex-
tremeness in D/E score. Dissatisfied.males will relate the most
negative D/E scores, and dissatisfied females the most positive.
This prediction is based on May (1969), who found a linear rela—
tionship between dissatisfaction and extreme D/E scores. May argued
that the more extreme D/E scores reflected overcompensation of the
dissatisfied individuals.

5. Females receiving mean D/E scores above the median female
score should.produce a significantly greater number of personally
caused PI's than females with D/E scores below the median. The
latter group should produce more heteronomous PI's. This hypothesis
is based on the proposition that the positive D/E pattern can be
meaningfully subdivided to represent two groups of females on sex
role attitudes, and that the PI is a logical variable to reflect
these differences.

6. Both sexes will relate positive DVE stories for those
stimuli judged.most similar to the self. This prediction is based
on the assumption that the subject will project his or her own wishes
and desires onto the picture hero, when the stimulus is rated as
highly similar to the subject. iHence, the picture hero will be

ultimately successful and will overcome the present difficulties.

5h
According to this hypothesis, a positive or negative D/E score is
not sexprelated, except insofar as females assume greater similarity
between themselves and others than males do.

Males will relate less positive D/E stories, or negative D/E
stories, to pictures they judge as being least similar to themselves.
This prediction is based on the proposition that the mechanism un-
derlying negative D/E stories is dissociation of the self from the
picture hero. The evidence reviewed in an earlier section suggested
that males are more likely to dissociate themselves from others than
females are.

Females are expected to relate positive stories to all pictures.
Since females have a strong tendency to assume similarity between
themselves and others, the level of dissimilarity occurring for
females on the pictures they judge as least similar to themselves

may not be strong enough to produce negative D/E stories.

METHOD

Stimulus one study. ‘A preliminary study on the semantic structure
of some thematic pictures was conducted to determine the meaning
of pictures which have been used in past D/E studies and to help
select pictures for the main study. .A group of fourteen TAT cards
were included in the picture pool, along with May's Pictures A and B,
and those used in the pilot study (Pictures E, F, G and H). The
following criteria were used in deleting TAT pictures from.the
original set. The reader may wish to refer to verbal descriptions
of the TAT pictures, in Appendix A.
Criteria for Deleting Stimuli
1. Pictures without readily identifiable human
figures were deleted. This criterion elime

inated Cards 11, 12BG, 16 and 19.

2. Pictures designed for use with children were
deleted, including 13B and 136.

3. Cards which were especially unrealistic were
omitted, including 8EM and 18EM.

4. Several TAT pictures have male and female
counterparts. Only one of each pair was in-
cluded. Since the male cards have been used
in the past research, Cards 3GP, 6GP, 7GF, 9GP
12F, 17GF and 18(}F were eliminated. Card 8GP
was included in the pool because 8BM was elime
inated under Criterion 3.

A procedure similar to that employed by Goldfried and Zax (1965)

was used. Twenty-five males and twenty-five females rated each of

56

the twenty stimuli on ten adjective dimensions taken from Osgood,
Suci and Tannenbaum (1957). Fbur of the adjective dimensions were
evaluative, three represented the activity factor and three the
potency factor. The evaluation adjectives were: accepting-rejecting,
happy—sad, pleasant-unpleasant, and.hopeful-hopeless. The activity
dimensions were active-passive, impulsive-controlled and excitable-
calm, and the potency dimensions were dependent-independent, safe-
dangerous and lenient-severe.

Each subject had a booklet containing twenty sheets with a
randomized order of the ten adjective dimensions on each sheet, and
a folder containing the twenty pictures. All subjects viewed the
range of pictures before beginning the rating task, and the order
of pictures was shuffled after each subject. The subject's task
was to place the identification number of each picture at the top
of the rating sheet and then to make all ten ratings for each picture.
The ratings were made on a seven-point scale. A.copy of the rating
sheet can be found in Appendix C(P12).

The data to be analyzed consisted of ten ratings made by each
of fifty subjects on each of twenty pictures, totalling two hundred
ratings for each subject. Since the purpose of the preliminary study
was to obtain the factor structure for individual pictures, especially
for Pictures A and B, a separate factor analysis was done on the
ratings of each picture. A.principle axis solution was used, fol-
lowed by a varimax rotation of exactly three factors. Three factors
were rotated rather than setting an eigenvalue threshold because the

three factors of evaluation, potency and activity were expected to

57

appear in the ratings. The results of the factor analyses are
presented in the Results chapter.

Overview of the method of the main study. In order to test the

 

main hypotheses, a 2 x 3 x 2 factor experiment was conducted using
two levels of instructions, three experimental procedures and both
sexes in each instruction X.procedure combination. Half of the
subjects wrote D/E stories under instructions used by may (1966),
to be as "dramatic and psychologically meaningful as you can."

The rest of the subjects received control instructions, which were
identical to the dramatic instructions except that no statement
about the manner of writing stories was included.

Three different procedures insured that any effects on the
dependent variables of the order of writing stories and rating the
pictures on similarity to self could be assessed. In one procedure,
subjects wrote stories and then rated the pictures on similarity to
self. This was called the Story-Rate group. The second procedure,
Rate-Story, reversed these two phases of the study and subjects
rated the pictures on similarity to self before writing their stories.
The third procedure involved writing stories without rating, the
Story-Only condition. All subjects filled out the sex role dissat-
isfaction questionnaire at the end of their respective procedures.
It should be noted that, at the point of writing the stories, both
the Story-Rate and Story-Only subjects who were given dramatic in-
structions provided a replication of'May's earlier procedures.

Four stimuli were selected for use in the main study, according

to the following criteria. (1) Only pictures which.have previously

58
elicited D/E sex differences, to some degree, were considered.
This criterion insured a basic comparability between the present
study and past work. Whereas the factor analytic results of other
pictures can be used in the future to help determine the important
stimulus characteristics for eliciting D/E differences, this kind
of research question necessitates a separate study. (2) Stimuli
were selected.which represent the range of evaluative adjective
dimensions, and.which also show a range on impulsivity—control and
dependence-independence. This criterion was developed out of the
review of semantic ratings of TAT pictures presented in the stimulus
cue chapter. This review illustrated that the pleasantness of the
stimuli was not crucial to D/E, and that the power and activity
dimensions did seem relevant, but the pictures used in past studies
have varied along these dimensions. (3) .After pictures which
fit the above criteria were identified, several were omitted from
further consideration by their highly situational nature. Examples
of this kind of picture were TAT Cards 9EM (men sleeping in a field)
and 13MF (man standing and covering eyes, and woman in bed). Al-
though all the pictures are situational to some degree, an attempt
was made to select those for which the subjects could readily iden-
tify some personality traits in the rating task.

The four pictures that were selected from the pool were May's
Pictures A (trapeze) and B (matador), Picture H (couple gazing into
each other's eyes) and Card 33M (huddled figure). The first three
are included in Appendix A. Fbur pictures were chosen because several

experiments have shown that four pictures are optimal for measuring

59
different kinds of thematic variables, such as need for Achieve-
ment (nAch), or affiliative tendencies (Reitnan and Atkinson,
1958). Murstein (1963) summarized the issue of the optimal
number of cards by saying that a lengthy TAT in research studies
probably defeats its purpose by mitigating the subject differences
through factors such as fatigue.

01' course, D/E may not behave like nAch since D/E is not
scored by the mere addition of the number of references in a pro-
tocol, nor is D/E conceived of as being a drive. However, pre-
vious D/E studies have not demonstrated any optimal number of
cards. The number of stories written has ranged from two (May,
1971) to eight (Fried, 1972). Four pictures were used in this
study so that more than just May's two pictures could be used, for
comparison purposes, but more than four did not seem necessary.

To test the possible effect of number of stories on D/E score, a
separate study would have to be designed. Four counterbalanced
orders were used in this study to control for possible position
effects. The four orders were as follows, and were randomly assigned
within each experimental condition.

1. A, B, H, 313M

2. B, A, 3BM, H

3. H, 33%, A, B

1+. 313M, H, B, A
Two subjects in each of the twelve experimental cells wrote stories
in each order, and the two remaining subjects in each cell were
(assigned randomly to one of the orders. There were ten subjects in

each of the twelve cells, and three-to-four subjects were run in an

experimental session.

60

Subjects

Sixty males and sixty females enrolled in introductory
courses in psychology athichigan State University volunteered
for the experiment entitled, "Personality Expression." All sub-
jects received two experimental halfecredits for participation.
Apparatus

The experiment was conducted in a 9'x12' room equipped with
two tables and four chairs. Each subject was provided with a
set of instruction sheets and a folder containing the four pictures,
an order sheet listing one of the four counterbalanced orders, and
a stapled booklet of four story sheets. The set of instructions
was comprised of a series of pages which guided each subject
through his randomly assigned procedure in a step-by—step fashion.
The procedure section belOW'Will describe the Story-Rate group in
detail. The first seven instruction sheets in.Appendix C constitute
the instruction set for this group (with dramatic instructions).
The eighth sheet in the Appendix is an example of the story sheets
fer writing the protocols, and number nine is an order sheet. The
remaining pages in the Appendix were used for other experimental
cells to make the step-by-step procedure move smoothly. Fbr all
subjects, the first page (P1) of the instruction set was a general
introductory comment.
Procedure

.After reading P1, the Story-Rate subjects moved immediately to
P2, looked at all the stimulus materials, and arranged the pictures

according to the order sheet in the folder. Subjects then read P3,

61

which described the procedure for writing stories. The experi-
menter (E) reiterated the important aspects of the story—telling
task, such as writing a whole continuous story for each picture,
including a statement of the outcome, and when appropriate, writing
very dramatic stories. (In the control instruction groups, no
statement was made of the nature of the stories to be written.)
When all subjects indicated that they understood the story-writing
procedure, they looked at the picture on top for twenty seconds and
then turned it face-down on the table and began writing their story
on the first of the four stapled story sheets. The anatched the
time for the subjects, indicating when five of the six allotted
minutes had elapsed. This procedure continued until all four stories
were written. Subjects were asked to place their booklets back into
the folder and to move on to their next instruction sheet.

The next sheet (P4) was the first of three pages that comprised
the similarity to self rating task. Subjects were asked to take
the four pictures and study them, asking what personality character-
istics were being exhibited by the individuals in the pictures. The
subjects were not asked to write their impressions, but only to get
an idea firmly in mind for at least one of the people in all four
pictures. After doing this, subjects moved on to P5. They selected
the picture containing an individual who was "most" like them, and
placed the code letter for that picture on the first line at the bot-
tom of P5. From the remaining three pictures, they selected the one
they felt was "least" like them and.placed its code letter on the

last line. Then, they placed the code letter of the picture they

62

considered to be more like them of the remaining two pictures on
the second line, and the code of the last picture on the third line.

The order of the pictures from P5 was copied on P6. Then the
subjects were asked to estimate just how similar or dissimilar to
themselves these pictures really were. P6 contained four Likert
scales, ranging from "1" to "9." The extremes were labelled "very
unlike me" to "very much like me." Subjects were asked to rate each
picture on a separate scale.

.After P6 was completed, theԤ.asked all subjects to place
all the materials in the folder except for the last page of their
instruction set (the sex role questionnaire). ggexplained that the
purpose of the questionnaire was to obtain information about what
being male or female is like in this society. Subjects were asked
to answer the questions in whatever way best expressed.how they
really felt, and were asked not to answer any questions with just
one word but to elaborate on their feelings. Subjects were allowed
to complete the questionnaire at their own pace, and were encouraged
to ask if they were unsure at any time about the meaning of the
questions. .After the questionnaire was filled out, subjects placed
all materials inside the folder, and any questions were answered.

For subjects in the Rate-Story groups, P10 followed the general
introductory comment. P10 is identical to P4, in that both sheets
instruct the subject to study each picture and.make a judgment con-
cerning the personality traits being exhibited there, except that P110
also includes information about the contents of the folder. This

information had been given to Story-Rate subjects on the second

6}
instruction sheet, when they were asked to open the folder and
arrange the pictures in the order for writing stories. The dif—
ferent sheets were necessary to guarantee that all subjects had
the information they needed at the appropriate time. P11 was
inserted between the similarity to self judgment task and the
story-telling task. The Story-Only subjects moved directly to the
sex role dissatisfaction questionnaire after finishing the story-
writing task.

After the experimental session, the folders were opened and
the instruction pages were separated from the rest of the materials
and prepared for the next experimental group. All the sheets com—
pleted by a subject were given a common number. .After all subjects
were run, the story sheets were unstapled, randomly sorted and

distributed to the raters.

RESULTS

Egptor analysis 2; the picture ratings. Separate factor analyses
were performed on the ratings for each of the twenty stimuli and
three factors were rotated using the varimax method. The complete
set of rotated factor structures for the pictures used in the D/E
study are presented in Table 3. The factor structures of the
remaining sixteen pictures appears as Table 20 in.Appendix D. The
actual mean ratings obtained on each of the adjective dimensions
for each picture are also included as Table 21 in Appendix D. .A
rating of "4" in Table 21 indicates neutrality or irrelevance of
the adjective pair. .A rating below "h" indicates that the picture
was more often described by the adjective on the left-hand side of
the bipolar pair. ,A rating above "h" indicates the opposite. An
asterisk (*) marks the factor with the highest loading.

The most striking feature of the factor structures of Pictures
A and B (May's trapeze and bullfighter pictures) is the high simp
ilarity in the direction of their loadings. The only adjective
dimension which is reversed in direction is "dependence-independence."
Both stimuli were rated as happy, hopeful and pleasant on the evalua-
tive factor, and as dangerous, controlled, excitable and active.

Picture H differed from.Pictures A and B on the "safe-dangerous"

and "impulsive-controlled" continua. Picture H was rated as safe

65

 

 

 

 

mmme. mmme. ommm. mmme. momm. eemm. accesses> mo coapwodowa
mzmw.l owmo.l mooo.l wmwm.l *wnmo.l owmo. Eamoloandpwoxo
momo. *moao.u mooo.u*emem. ommo.u ammo. ooaaospsoouo>wcasosa apa>apo<
*meoo.n meom.u mmmo. aaeo.n omem.u *omeo. o>acacouo>apoc
*momm. meow. ammo. ommm.u *emmo. ammo. oso>ocnpsoasoa
have. *mmzm. meme. Nmmo. *moro. mmwo. odoaowddolommm aocopom
*>©m©.I «ONF.I mdmo.l*mwmm.l mmmz.l wmew. psoUGoQoUdHIPQoUCoQQU
mmmo. seem. *ommo. ammo.u aomm. *emee. psscaoaaosupesccoao
meeo.l mmwo.u *zmmw. mace. mmmm.l *wrmw. mmoaomonlapmomoz
soaposassm
meeo. were.u *mmom. memo. mmma. *oemm. townhouse
memo. momm. *ewoo. momm.u oeom. *mmem. weapoooosumsapooooo
HHH HH H HHH HH H mdoamcmeaa mpopoom
d o>apoo~t< asapsosoeeao
. . oapddaom
Haosapmx

 

 

storm m\o sou ooeooaom aassapm to coopootpm tosses

M mqmda

66

 

 

 

 

:emo. neon. emoe. mmme. neem. wwdm. ”moodasm> mo soapaomoam

ameo.n *oeoa. Nmmm.u mmeo.u omme.u eeoe.u sasouoasspaoxo
emeo.n oema. *emmm.u memo. ammo.u mare. ooaaotpsoonceaaaoosa apa>apo<

homo. *mmee. ammo. eoom.u *wmee.n ween. o>aoosouo>apos

ammo.n *mmoa.u omom. maem.u *oeoo. meee. owo>ooupsoaoofl
mmoo. *oemo.n mmem. ommo. mono. *oeam. coowcwesouousc aosopom

*Fmee. emeo. eomo.u*eoeo. mmmo. eooo. pooosooooeaupsoosoooo

aome. moem.u *ommm. emom. mome.u aemmm. psscsoaooonposcsoeo

wmww.l NNMO.I *Momm. *wmww. wxmm.l momm. mmmammofildfiMonn
Cowpesae>m

maem.n mmoe.u *momm. mmmm. ammo.u *momm. tornadoes

memm. emoe.n amemm. mmeo.u aomo. *mmmm. weapooocanmsapocoos
HHH HH H HHH HH H odoamsoeao oaMmeMmmao

2mm m ceasooooa H assessom.
Hassapm

 

 

A.o.psoov m mamas

67
and impulsive, and dependent (like Picture A). Including Picture H
in the D/E study provided a mild test of Fried's hypothesis that
danger is one of the most important characteristics of thematic
stimuli capable of eliciting D/E. This test was weakened by the
fact that the "impulsive-controlled" dimension also differentiated
Picture H.from Pictures A.and B. None of the other stimuli in the
picture pool were identical to.A and B except for the "safe-
dangerous" dimension.

Card 3BM differed from Pictures A.and B in two ways. First,
the evaluative ratings were uniformly in the opposite direction
from.those of.A and B, toward rejecting, sad, hopeless and unpleas-
ant ratings. Card 3BM was also described as impulsive rather than
controlled. The factor structure found for Card 3BM was totally
consistent with that reported by Goldfried and Zax (1965).

In sum, all four pictures were described as excitable and active,
and they grouped into different combinations on the other adjective
dimensions. Pictures A, B and.H.were rated.positively on evalua-
tion, while 38M was described as sad and hopeless. Pictures A, B
and 3BM were all considered dangerous and severe. Picture H was con-
sidered safe. Finally, Pictures H and 3BM were similar in that
both were described as impulsive, while Pictures.A and B were
controlled.

Results fggm;the main study. The three raters scored all protocols
independently. Correlations between raters on D/E scores and RFC
scores were based on random samples of one hundred stories each.

The percentage of perfect agreement between raters was determined

68
for.AutonomyéHeteronomy'(APH) scores. The results are presented in
Table 4. .After all protocols were scored by each rater, agreement
sessions were held to settle disagreements before the data analysis
was begun.
TABLE 4

Rater Reliability/Agreement
For.All Story Mhasures

 

 

 

Raters
Measures

R2 R3
D/E .88 .89 R1
RPC .88 .94
AAH .91 .95
D/E .83 R2
R-C .89
AAH .91

 

All three raters agreed that 95% of the stories were scorable.
The independent decision of a rater that a certain story was un-
scorable, however, was often not supported by the other raters.
Consequently, the consensus on unscorability of the remaining 5%
was only about 40%. Phi coefficients were calculated to determine
agreement on scorability and the correlations for each pair of
34, r

raters was as follows: =.53, r23=.48. .After the agree-

r12=' 13
ment sessions, a total of sixteen stories were considered unscorable
and the mean score of each of these subject's other three stories

was assigned to replace the missing value. The sixteen unscorable

stories comprised 3.3% of the total story pool. Four subjects were

$1.“

CU

69
replaced because two or more of their stories were unscorable.

In the results that follow, multivariate analyses of variance
were often.perfonmed, but the glvalues that are reported are the
univariate statistics after an overall multivariate Efvalue with
a probability of .05 or less was found. This is always true,
unless an §.priori prediction was made concerning a specific come
parison.

Story length. Females wrote significantly longer stories than

males for each of the four pictures. The mean number of words

for each picture is given in Table 5. When story length.was
TABLE 5

Mean Story Length (words)
For Each Picture

 

 

 

’Pictfires
Sex A* B 2H 3BM Overall Mean**
Males 106 110 105 109 107
Females 122 121 127 123 123

 

*Tests for sex differences on each picture are as followm:
Picture.A: F=10.08, 1 and 108 df, p(.001.
B: F=4.48, 1 and 108 df, p(.03.
H: F=21.32, 1 and 108 df, p(.0001.
3BM: F=7‘.31, 1 and 108 df, p(.OO7.

**
t=7.00, p(.001.

correlated with D/E score, however, no significance was found.
Rank order correlations were computed for each sex and each.picture,
and are presented in Table 6. Since both RFC and AAH scores depend

on qualitative aspects of the protocols, they are independent of

story length.

70
TABLE 6

Correlations Between Story
Length and D/E Score

 

 

 

Pictures
Sex .A B H. 3EM. Mean
Males +.12 +.03 -.13 +.03 +.01
Females +.ol+ +.18 +.ol1 -.13 +.03
Mean +.08 +.1o -.05 -.05

 

Picture effects. 'Within each treatment condition (instructions

X procedure X sex), subjects were randomly assigned to write

stories in one of the four counterbalanced orders. A sex.X order
multivariate analysis was performed on the D/E scores for each
picture and on the D/E scores summed across pictures. The separate
test on each picture would reveal whether any systematic differences
occurred in D/E score for a specific picture as a result of when
that story was written. The analysis of the summed D/E scores was

a test of the overall order effect. There were no significant
findings.

.A sex.X.pictures repeated.measures analysis of variance was
also done on the D/E scores to assess any picture effects irres-
pective of order. No significance was found.

Effects gf’procedur . Subjects were run in one of three procedures

so that the effects of the rating task and story-writing on each other
could be assessed. The three procedures were Story-Rate, Rate-Story
and Story-Only. IMultivariate analyses were performed on the D/E
and.RPC scores and on the.AéH classification of the PI, for the

effects of procedure. No significance was found. Also, the effect

71

of procedure on the D/E scores of the "most" and "least" similar
stories was tested, as well as any effect on the similarity
ratings themselves. Again, no significance was found. In the
results that follow, procedure is omitted from further con-
sideration.
Results for hypothesis 1, The first hypothesis predicted that
stories written by females would contain significantly more depri-
vation units prior to the pivotal incident (PI) than stories writ-
ten by males, and also fewer enhancement units prior to the PI.
Both hypotheses were supported. The mean number of pre-PI depri-
vation units for females was 2.91, and 1.86 for males (t=3.00,
p(2001), using a one-tailed.t¢test. The mean number of enhancement
units prior to the PI was 2.35 for females, and 3.56 for males
(t=3.40, p(.oo1). Thus, the males and females in this study did
differ in their emotional starting points when writing stories.
Results for hypothesis 2, The second prediction stated that a
significant sex difference in D/E scores would be found, such that
females would tell significantly more positive D/E stories than
males. This sex difference was hypothesized to occur in those ex-
perimental cells replicating the procedure used in former studies,
i.e., the Story-Rate and Story-Only groups with dramatic instructions.
Subjects in the Story-Rate and Story-Only groups receiving control
instructions were expected to ShOW'nO significant sex differences.

Table 7 presents the means for this specific comparison, with
n=24. An instructions X sex multivariate analysis was perfbrmed on

the EVE scores for each picture, and on the mean scores across

72
TABLE 7

Mean D/E Scores for Instructions and
Sex(Story-Rate and Story-Only Cells)

 

 

 

Pictures
Instructions Sex
A. B Ii 3BMI Mean
Males .40 .25 .50 .15 .55
IDramatic

Females 1.60 5.70 1.65 -1.60 2.67

 

Males .00 - .50 1.65 1.40 1.27
Control
Females 1.50 2.60 1.20 5.05 2.04

 

pictures. Two main effects for sex.were noted. First an analysis
of the mean D/E scores across pictures showed that the females
told significantly more positive D/E stories than males (F54.53,

1 and 76 df, p(.03). The second main effect for sex occurred for
Picture B (F=7.53, 1 and 76 df, p(.007). This difference was also
in the predicted direction. The mean D/E score for females was
+3.15 on Picture B, and the mean for males was -.12.

Although these results showed that the predicted sex difference
in D/E did occur, it did not happen for the reason presented in the
second hypothesis. There was no effect of instructions on D/E score.
In order to test for the possible effects of instructions in the
experiment as a whole, a multivariate analysis of variance was per-
formed on D/E,.A4H and RFC scores, using instructions as an inde-
pendent variable. The only effect of instructions was a main effect
on the D/E scores of stories written to Card 3BM (F=6.50, 1 and 108 df,

p(.01). The mean D/E score for stories written to 3BM under dramatic

73

instructions was -.55, and.nnder control instructions, the mean
score was +2.03. Thus, dramatic instructions produced a negative
D/E pattern while control instructions yielded a positive pattern.
This result did not interact with sex, nor was it apparent on any
picture other than 3BM. Finally, instructions did not interact
with either sex or procedure on any of the dependent measures.

Two other analyses were performed as part of the second hypo-
thesis, both testing for sex differences in D/E scores, but with-
out instruction level as a factor. First, an analysis for sex
differences was done using only those subjects in cells that
replicated earlier studies, i.e., those cells comprising the top
half of Table 7. The analysis was again performed on each picture
and the mean across pictures. A.significant sex difference was
found for Picture H only (1753.76, 1 and 38 df, p(.05). The test on
the means across pictures was not significant, undoubtedly because
of the reversal in direction that occurred for females on Card.3BM.
This reversal, however, was not significant.

.A separate test for sex differences was performed on the D/E
scores of Pictures A and B combined, since these two pictures to-
gether have been used in most of the past research. Together,
Pictures A and B did differentiate the sexes in the predicted dir-
ection (t=2.09, p(.025).

An analysis of D/E scores using all twelve experimental cells
combined, showed a main effect for sex for the mean across pictures
(F=4.51, 1 and 118 df, p(.03) and for Picture B (bullfighter) indi-

vidually (F55.42, 1 and 118 df, p(.02). The data showing these

74
effects are given in Table 8. .All the pictures except H showed
trends in the predicted direction, although only Picture B was
significant by itself.
rJ:ABLE8

Mean Male and Female D/E Scores
For All Cells and the Mean Across Pictures

 

 

 

Pictures
Sex
A. B H 3BM, Mean
Males + .68 + .30 + .88 + .36 + .55
Females +1.71 +2.65 + .61 +1.11 +1.52

 

Results for hypothesis 3, The third hypothesis predicted that the
Resurgence-Collapse (R-C) dimension developed by Fried (1972) would
differentiate the sexes. The female mean R—C scores were expected
to be more positive than those obtained by males (i.e., in the Re—
surgence direction). The tests were performed on the scores of all
stories, and the hypothesis was supported both for Picture A.(FE4.88,
1 and 118 df, p(.02), and for Picture B (F=8.60, 1 and 118 df, p(.004).
The mean scores on R—C for all pictures are given in Table 9, based
on n=60.

TABLE 9

.Mean R-C Scores For All Pictures

 

 

 

Pictures
sex A B H _, 3BM
Males -018 -015 ‘002 -054

Females +.05 +.24 -.08 —.52

 

75
Results for hypothesis 4, Dissatisfaction with sex role was ex-
pected to be positively related to extremeness in D/E score, in
the appropriate direction for each sex. Dissatisfied females
would write the most positive D/E stories and dissatisfied males
the most negative.

The sex role questionnaire yielded the distribution of scores
given in Table 10. As might be expected from a sample of university
students, the female subjects were, as a group, more dissatisfied
than the males (F'=50.61, 1 and 118 df, p(.OOO1).

TABLE 10

Distribution of Sex Role Scores

 

 

 

Sex
Nhles Females
Mean 2.57 4-55
s. d. 1.28 1.66

Range 0.0-5.5 1.0-9.0

 

To test the fourth hypothesis, high and low sex role dissat-
isfaction groups were formed for each sex. Subjects classified
into each group had sex role dissatisfaction scores above and below
the median score for their sex. The median score for the male
sample was 2.5, and the median for the females was 4.5. The average
D/E score of each subject was used to calculate the means given in
Table 11. The number of subjects in each group is given in paren—
theses. A.sex X sex role dissatisfaction analysis of variance was

not significant.

76
TABLE 11

Mean D/E Scores of Sex
Role Dissatisfaction Groups

 

 

Dissatisfaction Scores

Sex Above Mdn. Below Mdn.

 

Males + .57 (28) + .71 (25)

Females +1.04 (25) +1.81 (28)

 

Note.—Seven females and seven males with
dissatisfaction scores on the median
values were omitted.

The large differences in dissatisfaction scores, especially
among females, meant that a single division into high and low
groups created highly heterogeneous categories. In order to make
the sex rOle groups more uniform, subjects were re-divided into
three levels of dissatisfaction: low, moderate and high. This pro-
cedure produced the distribution shown in Table 12. In creating
these groups, the author attempted to compromise between equal
ranges of scores and equal numbers of subjects in each level. The
resulting intervals are of unequal size, and hence, the comparability
of the sexes at any level is unknown. The three levels can only be
interpreted comfortably on a within-sex basis without further
evidence for their comparability. However, in order to determine
whether the three-way classification distorted a relationship be-
tween dissatisfaction and D/E scores in the sample as a whole, the
correlation between these two variables was calculated, and found
to be +.006. Even when sex is partialled out of this correlation,

the result is -.13. These correlations provided evidence for

77
TABLE 12

Threedway Classification of
Sex Role Dissatisfaction Scores

 

 

sex.Ro1e Scores

 

 

Sex
LOW’ 0.0 - 2.0 25
Males IMbderate 2.5 - 3.0 20
High 3-5 - 5-5 15
LOW’ 1.0 - 3.0 12
Females Moderate 3.5 - 4.5 23
High 5.0 - 9.0 25

 

concluding that the three—way classification does not distort any
relationship in the data and is thus acceptable as a method for
breaking down the sample into more homogeneous cells.

The mean D/E scores for the six dissatisfaction groups are
given in Table 13. Nete that "high" refers to high dissatisfaction.

TABLE 15

Mean DVE Scores for Sex Role Groups

 

 

 

 

Sex Picture

.A B H 3BM Mean

LOW' + T84 +1.80 + .40 + {80— + .96

Males Mbderate + .55 -2.40 + .60 + .75 - .18
‘High + .86 +1.40 +2.06 - .86 +1.73

LOW' +1.50 +3.75 + .50 +4.41 +2.54

Females Moderate +1.82 +3.34 +1.73 - .78 +1.53
High +1.72 +1.44 - .36 +1.28 +1.02

 

78

The first analysis of the data in Table 13 was a sex X.sex
role dissatisfaction multivariate analysis on the D/E scores for
each picture and on the mean scores across pictures (last column).
No significance was obtained. Then, the D/E scores of the different
sex role dissatisfaction levels suggested that the scores might be
following certain patterns. For example, the means given in the
last column of Table 13, for females, seemed to be decreasing
linearly with increases in dissatisfaction, and the same means for
males appeared to resemble a quadratic curve. Although patterns
could be discerned in some of the means for individual pictures,
they were not consistent across pictures or sex. Tests for linear
and quadratic effects were perfbrmed.on the data in Table 13, using
sex and sex role dissatisfaction as independent factors. Any results
from this analysis are strongly tempered by the fact that the tests
are totally'pggtlhgg, The tests were performed because the author
felt that finding strong linear and/or quadratic patterns would be
important for future research considerations.

The D/E scores of the three dissatisfaction levels of males on
Picture B fit a quadratic curve (Fe7.75, 1 and 114 df, p(.006), and
a quadratic curve was also found for female scores on Card 334
(F55.84, 1 and 114 df, p(.01). For both sexes, the pattern was one
in which the moderately dissatisfied subjects produced negative D/E
scores, while the other two dissatisfaction groups wrote positive
stories. In addition to these two specific findings, the overall
mean D/E scores for females were marginally linear (F22.96, 1 and

114 df, p(.08). For females, then, D/E scores decreased in

79
positivity with increased dissatisfaction.

As stated, then, the fourth hypothesis was not supported. The
operational definitions of "satisfied" and "dissatisfied" subjects
were considered unsatisfactory, however, because they did not re—
flect the large range of sex role dissatisfaction scores that was
obtained. ‘When dissatisfaction levels were redefined, significant
patterns for dissatisfaction levels appeared on two pictures, one
for each sex. In both patterns, the moderately dissatisfied sub-
jects related quite negative D/E stories, while the low and high
groups obtained positive D/E scores. There was a borderline linear
pattern for the overall mean D/E scores for females, in which the
most positive D/E stories were told by the least dissatisfied fe-
males. This last relationship is the exact opposite to that re-
ported by May (1969).

Results for hypothesis 5, Females who obtained mean D/E scores

above the median female score were expected to produce a significantly
greater number of personally caused PI's than females with.mean D/E
scores below the median. The latter group was expected to mediate
their stories with heteronomous PI's.

The frequency of autonomous and heteronomous PI's in stories
with D/E scores above and b610W'the median for each sex is given in
Table 14. The median DVE score for both sexes was +1.0. There was
a strong preference for autonomous PI's in stories with D/E scores
above the median, but both.kinds of PI appeared equally likely in
stories with D/E scores below the median. The Chi-square was not

significant.

80
TABLE14

PI Classification for D/E Groups

 

 

 

 

D/E Scores
Sex
Above Mdn . Below Mdn.
A H A. H
Males 88 15 48 62
Females 91 27 49 52

 

Note._Frequencies exclude seven unclassifiable
stories and all stories with D/E scores on the
median score for each sex.

Hypothesis 5 was based on the assumption that the females
who wrote the most positive D/E stories were also those females
who were most dissatisfied with the female role. These females
were expected to mediate their stories with autonomy, because
autonomy verifies their belief in personal effectiveness. In this
study, the dissatisfied females did not relate significantly more
positive stories than the relatively more satisfied.females (Hypo-
thesis 4). Therefore, the fifth hypothesis was restated to directly
express the relationship expected between dissatisfaction with sex
role and PI classification. Those females who were classified as
highly dissatisfied, according to the three-level grouping, were
expected to relate more autonomous PI's than the females who were
less dissatisfied.

The frequency of autonomous and heteronomous PI's in each sex
role group is given in Table 15. Since the number of subjects in
each sex role group was different, the proportions represented by

the frequencies appear in parentheses. A Chi-square was performed

81
on the largest difference, i.e., the frequencies in the low and
moderately dissatisfied male groups, which was not significant.
TABLE 15

Proportion of Type of PI
of Sex X Sex Role Groups

 

 

Sex Role PI Type
Groups A. H

 

Sex

 

LOW" 73 (.73) 26 (.27)
Males Moderate 45 (.54) 57 (.46)
High 59 (.66) 20 (.54)
Low 52 (.67) 16 (.33)

 

Females IModerate 60 (.68) 29 (.32)

High 61 (.61) 39 (.39)

Netey-Frequencies exclude two male stories
and four female stories with unclassifiable
pivotal incidents.

 

Finally, the frequency of autonomous and heteronomous PI's
occurring in stories told to each picture was examined. As can
be seen from Table 16, both sexes related more autonomous PI's to
all the pictures except for Card 3BM. The reversal from autonomous
TABLE16

Frequency of PI Classification

 

 

Pictures
Sex A B H 3BM
A H A H A H A H

 

 

 

Males 47 15 58 22 51 8 20 40
Females 44 15 42 17 44 15 22 58

 

(Note.-Table excludes six unclassifiable PI's.

82
to heteronomous PI's in writing stories to Card 3BM was not sig—
nificant.
Results for hypothesis 6, The sixth hypothesis predicted that the
D/E pattern would be dependent on the story author's assumption
of similarity between himself or herself and the picture hero.
IMales were expected to write stories to those pictures they rated
as being most like themselves which produce significantly more
positive EVE scores than stories written to pictures they rated as
being very unlike themselves. Although females were not specifically
included in this hypothesis, since assumed similarity was not ex-
pected to affect their scores, both sexes were included in the
analysis.
The mean D/E scores for stories written to the stimuli se-
lected as "most" and "least" similar to the self are given in
Table 17. There was no effect of procedure (Story-Rate or Rate-Story)
on the selection of pictures, and all four pictures were chosen, at
some time, for both the "most" and "least" positions. Males told
stories with D/E scores well within the "male" range on D/E regard-
less of similarity to self. Therefore, the hypothesis that males
TABLE 17

Mean D/E Score of Stories Told To
Pictures Judged.Mbst and Least Similar to Self

 

 

Similarity
Most Least

Sex

 

Males + .52 + .51

Females +1.62 + .72

 

83
write less positive D/E stories than females because they assume
"dissimilarity" between themselves and the story hero was not
supported.

.A further analysis was conducted on the "most" and "least"
ratings because the effect of similarity judgments on D/E score
was potentially quite different, depending on the sex role dis-
satisfaction level of the subject. For example, a story which
a dissatisfied female writes to a picture she judges as being very
similar to herself may be quite different in D/E than a story
written to the "least" similar picture, or a story written to a
"most" similar picture by a female who is less dissatisfied with
the traditional role. For this reason, a sex X sex role dissatis-
faction X similarity analysis of variance was performed on the
D/E scores. A.significant threeaway interaction was found
(F:3.17, 2 and 148 df, p(.04). The means are presented in Table 18.

TABLE 18

Mean D/E Scores of Sex Role Groups
on.Mbst and Least Similar to Self Stories

 

 

 

 

 

Sex Sex Role Similarity
Groups Most Least
Low +1.46 +1.00
Males Moderate -1.46 +1.60
High +2.10 -1.30
Low +1.70 + .90
Females Moderate +2.40 - .33
High + .80 +1.55

 

84
Tests of simple effects revealed that those subjects who
were classified as moderately dissatisfied showed a sex X
similarity interaction (F=4.74, 1 and 56 df, p(.05). Table 19
contains the means for this specific comparison.
TABLE 19

Nban.D/E Scores for Similarity Levels
For Moderately Dissatisfied Subjects

 

 

 

 

Similarity
sex Most Least
Males -1.46 +1.60
Females +2.40 - .55

 

These results showed that the degree of similarity to self had

a strong effect on the D/E scores within the moderately dissatis-
fied groups. Both sexes told stories to the "most" similar picture
which obtained D/E scores extreme in the direction appropriate for
their own sex. They also told stories producing the opposite
pattern score for those pictures they judged as being "least"

like themselves. This sex X similarity interaction was not sig-
nificant for either the low or high dissatisfaction group.

The male subjects also showed a marginal sex role dissatis-
faction x similarity interaction (F=2.91, 2 and 74 df, p(.06). The
means for this finding appear in the top half of Table 18. Eftests
were run on the difference in D/E score between "most" and "least"
similar stories within the moderately and highly dissatisfied groups,
and on the difference in D/E between moderately and highly dissatis-

fied groups within the "most" and "least" similar categories. All

85
of these comparisons fell short of significance. One should note,
however, the complete reversals in D/E score that occurred with
differences in similarity judgments for both moderately and highly
dissatisfied subjects. The moderately dissatisfied subjects told
stories which obtained D/E scores in the direction predicted for their
own sex on those pictures they judged as "most" similar to them—
selves, and they told stories to the "least" similar pictures which
obtained scores expected of the opposite sex. These findings were
exactly reversed for the highly dissatisfied individuals.

The sex X similarity interaction among moderately dissatis-
fied subjects is very similar to the findings of May (1969). May
found that dissatisfied subjects told stories significantly more
extreme in the direction predicted for their own sex. The data
from this study, however, does not totally replicate May's finding
because it is based on those two of the four stories which were
written to the "most" and "least" similar to self pictures. If
one can assume that May's subjects identified with the picture
heroes in their stories (and hence, their stories would be com-
parable to those written to the "most" similar pictures in this
study), then May's subjects were very much like the low and mod-

erately dissatisfied subjects in this experiment.

DISCUSSION

Effect 2: instructions. The D/E sex difference cannot be explained
by differences in initial perception combined.with the reversals in
emotion produced by a dramatic instruction set. Analysis of the
initial portions of TAT protocols did show that females began

their stories on a more negative tone than males, but instructions
to tell a dramatic story did not produce the reversal in emotion
documented by May. In fact, the sex difference in D/E, over all
four pictures, which was not significant using only the dramatic
instruction subjects, was magnified by the control instructions

and was significant. Since the sex difference was not significant
when only the control instruction groups were used, the significant
overall sex difference from the combined groups was probably the
result of a larger sample size. Also, the fact that no differences
in D/E were found in the control instruction groups of the main
study means that the pilot study and the main experiment were con-
sistent with each other.

One conclusion to be drawn from comparing D/E scores under
control and dramatic instructions is that the males behaved.much
the same as the females, although they showed less variation. Males
told.positive stories with control instructions, and the effect of

dramatic instructions was to dampen this positivity. The sex

87
difference reported in this study is comparable to that found
in previous work, since it resulted from extremeness of female
responses rather than from a female positive pattern versus a male
negative pattern. The males in this study exhibited such positive
D/E scores under control instructions (XB2.55) that they were well
within the range of female scores in earlier work. The strength
of the female pattern in this experiment is congruent with the
implications from the developmental work cited in the empirical
evidence chapter showing that the female response pattern is more
stable and extreme than that of the male.
Resurgence—Collapse. Despite the fact that the mean D/E score
for males was positive, males were more negative in Resurgence-
Collapse (RFC) than females, who obtained positive scores. This
was found on Pictures A and B only. Males ended their stories with
significantly more downfall episodes than females.

Negative Rec scores are possible even though the same stories
produce positive D/E scores because the D/E scoring system.depends
heavily on the location of deprivation and enhancement units within
the protocol and not just on the qualitative nature of the ending,
as does RFC. Of course, R—C is not a pattern difference, and it
could reflect sex differences in preferences for certain kinds of
endings, such as a female preference for "happy endings." This
hypothesis was discounted in May's early work as an explanation
for D/E differences. However, the fact that females produced signi-
ficantly'more deprivation units prior to the PI than males, and

males produced.more enhancement units prior to the PI than females,

88

indicated that more happened.in the protocols than just a mere
difference in endings.

The principle drawback of the Rec scoring system.is its in-
ability to differentiate the majority of stories. Fried (1972)
did not report the proportion of stories obtaining B and Q_scores
as opposed to "0" scores, but in both the pilot data and the pres-
ent study, we found that the majority of stories had to be assigned
to the heterogeneous zero category. In the main study, a total of
131 out of 240 stories written by females and 133 of 240 stories
written by males were assigned the "0" rating.
Stimulus effectiveness. .Although there were no picture or order
effects, the matador (Picture B) elicited the largest sex difference
when any differences were found. If one could order the stimuli
used in this study for degree of eliciting the D/E difference,
Picture B would be first, then Picture H,.A and 3BM. Picture H
was the only stimulus that produced significant sex differences
by itself under dramatic instructions. Picture H, of course, was
the most similar to the subjects both in age of the figures, and
in the relevance of the scene it depicted for their everyday lives.
Although Picture A (trapeze) consistently elicited.mean differences
in the predicted direction, it was not significantly discriminating
by itself in any of the comparisons. This seems unusual, given
past research, since Picture A has been used.most often and has been
the principle stimulus for finding significant sex differences.

.A factor analysis provided no clues for determining why Pictures

B and.H would differentiate the sexes more clearly than the other two

89
stimuli. They were not alike in any way that distinguished them
from the other pictures. Although Picture H was slightly more
popular than the others in the most—similar-to-me ratings,
Picture B was chosen as often as Picture A.or 3BM.

Card 3BM showed a trend in the hypothesized direction in D/E
score when all experimental cells were combined, but the females
in the replication groups exhibited a total reversal in D/E pat-
tern to Card 3BM from their scores on the other pictures. Their
mean score on Card 3BM revealed a quite negative sequence, while
female stories to the same picture were very positive in D/E when
the subjects were given no specific directive to write dramatic
stories. Males showed the same trend toward greater negativity
in stories to 3BM under dramatic instructions. (Together, the more
negative scores of both sexes produced the significant main effect
for instructions on stories written to 3BM.)

The reversal in D/E pattern on Card 3BM may have occurred be-
cause the embtion depicted in the picture was so intense and so
"dramatic"; that is, the picture illustrated that something drama-
tic had already happened, and that the dramatic event had led to
negative consequences. Since females usually show more consistency
between their emotional starting points and the objective ratings
of the pictures, they were more likely than males to respond directly
to the dejected state shown in the stimulus. Under dramatic in-
structions, responding directly to the stimulus means focusing on
the immediate past in an attempt to describe the drama surrounding

the huddled figure. The fact that females told quite positive

90
stories under control instructions implied that, in the absence of
specific instructions, the females emphasized the present state of
the figure and the ways in which the person coped with and overcame
the situation.

This explanation of the D/E reversal in stories to Card 3BM
suggests that the effect of dramatic instructions was not to pro-
duce a change of ending so much as a change in the time frame of
the story. This time change is probably not a necessary result of
dramatic instructions, but of the emotion in Card 3BMthich compelled
the subject to devise an explanation. The reason why dramatic in-
structions had no effect on stories written to the other pictures
is unclear. Perhaps producing a story with an intensity level so
different from that existing in the picture is very difficult. The
stories written to 3BM wnder control instructions may have been no
less dramatic than those written under dramatic instructions, even
though the PI occurred in a different chronological place.

Card 3BM was also different from the rest of the stimuli because
subjects produced a majority of heteronomous PI's in their stories to
3BM, and a majority of autonomous PI's to the rest of the pictures.
Although enough subjects related autonomous PI's in stories to 3BM
and heteronomous PI's in stories for the remaining pictures to
seriously qualify any conclusion that a picture effect existed here,
the high heteronomy on Card 3BM seems quite consistent with the
factor analytic results. Besides being rated negatively on the eval-
uative dimensions, the only difference between Pictures A and B and

3BM was the impulsivity-control dimension. The former cards were

91
"controlled," while 3BM was rated as "impulsive." Impulsivity
alone did not produce heteronomy, since Picture H was also rated
as "impulsive," and H elicited a clear majority of autonomous PI's.
But, Picture H was also rated as "safe" and as "pleasant," while
3BM was described as "dangerous." Impulsivity had quite different
connotations in story content when in context of safety and happi-
ness than in context of danger and unhappiness.

The kinds of themes told to Pictures 3BM and H support the
hypothesis that impulsivity has different implications for story
content depending on its context. In stories written to 3BM, the
impulsivity was almost invariably attributed to the figure's imme-
diate past. The person acted "out of control," from rage or
ignorance, for example. Picture H, however, elicited very little
emphasis on the past and.much.more on the future of the couple.

The impulsivity in Picture H was present— or future-oriented and
usually occurred from.the love that the individuals felt for each
other and not from an uncontrollable outburst of emotion. Returning
momentarily to Pictures A and B, subjects responded with very action-
and achievement-oriented stories, which usually involved some degree
of personal control. Heteronomous themes to Pictures A and.B
usually revolved around the hero's ignorance or submission to family
tradition.

AutonomyéHeteronomy. The results from the.AaH measure showed that the
majority of themes were autonomous. There were two factors which
probably influenced that fact more strongly than sex role identifica-

tion level. First, the sample was drawn from a university population,

92
where autonomy is probably more prevalent than in the general
population. Repeating the PI analysis with a wider sample of
people might produce the larger number of heteronomous PI's needed
to test the hypothesis more fairly. Secondly, the pictures used
may have had an unexpectedly strong influence on the nature of the
PI. This means that beliefs in different levels of causality, as
reflected in the PI, are easily masked and should be measured in
a different way.
Similarityjtg_§§lf, The method of measuring similarity appears to
have worked quite well. None of the stories written to "most" or
"least" similar pictures had to be eliminated because the ratings
indicated that subjects really did not view the stimuli they chose
for either category as being "like" or "unlike" themselves. Only
two subjects chose a rating as low as "6" on the 9-point scale
fer the "most" like me rating, and the highest rating for the "least"
like me picture was "3." IMost ratings were 8-9 for the "most" like
me judgment, and 1-3 for "least" like me.

Sex role satisfaction and similarity. There was no significant sex

 

difference in D/E scores when stories written to "most" and "least"
similar to self pictures were compared. However, similarity did
have a strong effect on D/E scores when the sexes were subdivided
according to members' satisfaction with the sex roles society has
defined for them. The most clear effect of similarity occurred for
the moderately dissatisfied subjects. Males in this group told
negative stories to the "most" similar picture, and positive stories

to the "least" similar picture, and the moderately dissatisfied

93
females did the opposite. This finding could be interpreted as
strain felt by both sexes to incorporate the appropriate role
model. Both.males and females who are moderately dissatisfied
are in a state of tension because they do not feel comfortable in
their roles but they also do not reject them.entirely, Hence, they
exert effort to conform as an attempt to increase belongingness.
This conformity interpretation would also explain the change in
direction in D/E score that occurred for each sex in writing
stories to the "least" like me picture.

The questionnaire data on which the sex role groupings were
based provide support for the contention that moderately dissatis-
fied subjects feel strain. The males in this group said they were
glad they were males, that they enjoyed the advantages of males,
but that they also felt that males were very limited in some res-
pects. Their most frequent complaint was the social dictum against
the open expression of emotion by males. Many males in the mode
erately dissatisfied group resented the expectation that they
succeed.materially.

Among females, the moderately dissatisfied subjects' ques-
tionnaire responses left the impression, sometimes stated directly,
that these females felt confused. They expressed desires to have
some kind of career or independent time, but they also wanted
marriage and the traditional advantages. They were confused about
how to direct their lives to obtain the best of both worlds. These
females expressed some resentment over tradition, but also felt

strain to conform.because they did.not reject traditional values

94
outright. They only wanted to adapt the norms to their personal
ambitions.

Those subjects who were highly dissatisfied with their sex
roles wrote stories which produced D/E scores expected of the
opposite sex for those pictures they judged as being "most" simp
ilar to themselves. This makes intuitive sense if one assumes
that the highly dissatisfied subjects have rejected the sex role
that accompanies their biological roles. This conclusion seems
justified by the intensity of the statements on the sex role
questionnaire that obtained the highest ratings, and by the fact
that several of these intense statements had to be present if the
total score was to be included in the high dissatisfaction group.
However, the conclusion that the highly dissatisfied subjects pre-
ferred the role of the opposite sex is not justified from the data,
but only that they did feel strong rebellion against the role asso-
ciated with their sex.

Those subjects who expressed no dissatisfaction with their
roles told stories which were quite uniform in emotional pattern
fer both "most" and "least" similar pictures. This D/E pattern
can occur when the fantasy is relatively balanced on both sides of
the PI. Rather than describe these individuals as "bland and flaccid,"
as May (1969) did, it is more descriptive to refer to them as people
who expressed no unhappiness about being their particular sex, and
who seemed quite comfortable living within their defined roles.
These subjects may have told similar D/E stories to "most" and "least"

similar pictures because their lack of tension about sex roles has

95
not resulted in emotional boundaries between themselves and the
opposite sex.

In order to assess the discrepancies between this study and
that reported by May (1969), two facts must be considered. First,
May reported a direct linear relationship between dissatisfaction
with sex role and extremeness in D/E score in the direction for
each sex. This finding was based on all protocols. In this study,
an analysis of all the stories showed that the only relationship
between dissatisfaction and D/E scores occurred for females, and it
was both weak and in the opposite direction from.that described by
May. The mean D/E scores for some individual pictures were sig-
nificantly quadratic in form, but this finding did not occur for
most of the pictures and should not be considered representative.

Secondly, the linear relationship between dissatisfaction and
extremeness in D/E score found in this study was limited to the low
and moderately dissatisfied groups, and also it appeared only when
stories written to the "most" similar pictures were included. Thus,
May was not supported when the procedure was the same as in his 1969
study, and was supported when the new variable of similarity to self
was included in the analysis. In sum, when the subject's identifica-
tion with the picture hero was held constant at a high level, the
low and moderately dissatisfied individuals in this study produced
D/E patterns similar to the satisfied and dissatisfied people in
May's experiment.

The "most" and "least" similar judgments demonstrated that the

degree of assumed similarity between the subject and.picture hero was

96

an important determinant of the direction of D/E score, especially
if the subject was at all dissatisfied with his or her own sex role.
Both moderately and highly dissatisfied subjects related stories to
pictures they felt were unlike themselves that obtained D/E scores
expected of the opposite sex. The moderately dissatisfied subjects,
especially the males, told stories to pictures they judged as "most"
similar to themselves that fit the male pattern. Although the highly
dissatisfied.males told positive D/E stories to pictures they judged
as "most" similar to themselves, the reason may not be because they
were straining toward a female identification. They may merely
have felt less resentment toward a female role than toward the male
role. There is no evidence from this study that the highly dissat-
isfied subjects felt a phenomenological identity with the opposite
sex.

In context of interpreting the behavior of subjects classified
into different levels of sex role satisfaction, one should note
that the questionnaire grouped individuals into categories according
to the number and intensity of statements of dissatisfaction. iHence,
subjects who obtained low scores, and even zero scores, cannot be
said to be more "satisfied," necessarily, than people who obtained
a higher score. People with very low scores may be defined only as
those individuals who responded to the questionnaire without men-
tioning any source of dissatisfaction. These individuals are ine
teresting in a relative way. The fact that no females completed the
questionnaire without stating ways in which they were dissatisfied

is of psychological import when one also recognizes that several

97
males did expound on the items without any such reference. Howe
ever, the fact remains that no conclusions can be drawn here about
the D/E behavior of highly satisfied people, except insofar as one
is willing to assume that the absence of statements of dissatis-
faction mirrors that psychological state.

Since the range of scores on the questionnaire was smaller
for males than for females, the defining points for the intervals
of the three dissatisfaction levels differed for each sex. .As a
result, the classifications are only comfortably interpreted on a
within-sex basis. However, the similarity in D/E behavior of the
moderately dissatisfied.males and females (in reverse directions)
provided strong support for their comparability, as did the parallel
behavior between the two highly dissatisfied groups. More evidence
from.some other independent source would be even more convincing.

It is possible that the greater number of dissatisfied females
than males may have partially resulted from the tendency of females
to disclose more about themselves than males do (Cozby, 1973; Jourard,
1964). This tendency may have resulted in higher dissatisfaction
scores for females because their longer answers would increase the
probability that a scorable statement would be made. However, when
males and females do not differ in the length of their statements,
females still disclose more intimate information about themselves
than males do (Pederson and Breglio, 1968). The quality of the
information given by males and females would.produce the differences
found in this study, since females are prone to relate more about

their anxieties and.worries than males, while the sexes do not differ

98
in the number of pleasant feelings they report (Levinger and Senn,
1967). On the basis of past research, then, we should expect a
certain proportion of the variance in the questionnaire data to
mirror sex differences in self disclosure.l Hopefully, however, a
greater proportion of the variation in the questionnaire responses
does reflect differences in dissatisfaction. The point concerning
self disclosure provides more evidence in favor of a within-sex
interpretation of the dissatisfaction scores.
Generalization. The problem of generalizing to broader populations
from.this study involves two issues. First, the representativeness
of the subject sample for the overall college population must be
considered. The portion of the study which would be most damaged by
a biased sample is the sex role dissatisfaction data. Since no
norms exist for "dissatisfaction," the questionnaire scores can be
interpreted only according to their position within the distribution
obtained in the study. This problem could be alleviated by admin-
istering a standard.masculinity-femininity‘(MEF) test twice to the
subjects, both with instructions to answer the questions honestly
and to answer them in a way that describes their ideal selves. The
discrepancy score would be a measure of dissatisfaction.

The second issue for generalizability concerns the influence of
the total university environment on the distribution of sex role
dissatisfaction scores, as compared to the general population. The
university sample probably included.more females who were openly
dissatisfied with the traditional norms of dependency and passivity
than a noncollege sample the same age. Also college men are more

feminine in their orientation than males not in college (Dahlstrom

99
and welsh, 1960). Secondly, there were probably more dissatisfied
people who were willing to state their feelings on the questionnaire
than was the case in May's 1969 study. Since a great deal of con-
sciousness raising has occurred for both males and females in.the
last four years, there is really no way to determine just how
comparable the groups used here were to those defined in 1969. The
effect of liberation movements would be to provide a psychological
"haven," especially for females who are highly dissatisfied, and
to help them feel capable of living in a world where deviation from
the female pattern would not evoke as much pressure to conform.
Expression of viewpoints running counter to tradition might also be
accepted.more now than in 1969, which would alter the criterion
for being called "dissatisfied." Consequently, the highly dissatis-
fied subjects in this study were probably a very different group
of people than May could encounter, and.most importantly, they were
people who felt free to admit high dissatisfaction, and free to
create a fantasy pattern very much like that expected of the other
sex when they felt that similarity between themselves and the sub-
ject person of their fantasy was high.

The final problem for generalizability is the issue of volunteer
status of subjects. This issue is difficult to assess because little
is known about the exact nature of the differences between.volune
teers and nonvolunteers. However, there have been no inconsistenp
cies in D/E studies for which a "volunteering" bias might be solicited
as a partial explanation, May's original study in 1966 utilized one

hundred and four nonvolunteers, while the 1969 study was conducted

100

with individuals from the general population who responded to a
newspaper ad describing "a study in imagination." These two
experiments represent the extremes of volunteer status, yet the
data obtained in both studies were consistent. In fact, not only
was the sex difference replicated in the 1969 study, but the males
produced an even more extreme negative D/E mean score than the
males in other experiments using college students. This finding
supported the results often obtained on MAP tests that males in
college populations are more feminine than.males in the general
population.
Reception-counteraction. The results of this study help evaluate
Fried's hypothesis that D/E differences are produced by magnifying
the empathic ability of females and.males' inhibition of spontane
eous emotional expression. Fried suggested that the pictures which
elicit D/E involve danger and depression, which the female subject
can readily accept and express in the initial parts of a story.
This direct emotional expression is called the "receptive" phase.
Males, however, cannot express the negative emotion, and they
"counteract" the depressive mood with positive emotion. Only
through continued exposure to the stimulus do the males come to
ultimately accept the negative emotion. Fried implied that pictures
which are not dangerous and depressing would not elicit D/E dif-
ferences.

The present study provided clear evidence against the "danger
and depression" thesis. Pictures B and H elicited D/E most clearly

and.neither was rated as sad or depressing. Actually, the factor

101

analysis showed that both pictures loaded positively on the
adjective dimensions of pleasantness, happiness and acceptance.
Also, Picture H was described as quite "safe," suggesting that
the danger dimension is also unnecessary for D/E differences.

Fried's receptive-counteractive hypothesis to explain D/E
sex differences only works when the pictures are depressing. If
the stimuli are pleasant, the female pattern of moving from rela-
tively more negative to more positive emotion cannot be explained
since there is no reason for her to begin on a negative tone. The
male pattern of increased negativity later in the story sequence
would also be without theoretical foundation.
Future ggpsiderations Englspeculations. Since the D/E sex dif-
ference is not a commonly known stereotype, the sex X sex role
dissatisfaction X similarity to self interaction cannot easily be
explained on the basis of social desirability. The fact that the
moderately and highly dissatisfied subjects of both sexes relate
stories to the "least" similar pictures that obtained Q/E scores
in the opposite direction from those of the "most" like me pictures
cannot be explained by any of the standard experimental demand
characteristics. This finding can only be interpreted by saying
that the D/E variable is tapping a phenomenological self-identifica-
tion of’males and females, which occurs on an emotional level. The
most exciting suggestion from.this experiment was that subjects
reflected the sex difference in D/E through their intra-individual
distinctions between stories written to "most" and "least" similar

pictures. This means that the subjects were cognizant, on some

102
level, of the emotional world of the opposite sex, and were aware of
the differences between their own and the other's emotional reality.
Substantively, this finding represents a firm step beyond previous
work, in which only inter-individual differences had been demon-
strated.

This experiment suggests several avenues for more work. The
sex role questionnaire needs to be refined and reliability needs to
be established. The relationship of scores on the dissatisfaction
questionnaire to scores on other instruments dealing with adherence
to sex roles needs to be determined. Also, the raw data from this
study should be rescored for statements of satisfaction. As exp
plained earlier, the extent to which the scores on dissatisfaction
are reciprocal to scores on satisfaction is unknown at this time.
Perhaps, in light of the research cited above that reported a lack
of sex differences in related positive affect, the distributions
of satisfaction scores might be very similar for both sexes.

Finally, the semantic differential results from the stimulus
cue study identified some picture characteristics which were not
crucial in eliciting D/E, such as danger and depression. The main
experiment was too limited to provide much evidence on the necessary
elements of the pictures. There was a considerable amount of unused
information from.the stimulus cue study that could be used to de-
sign a more systematic set of studies to scale the pictures along

relevant dimensions.

BIBLIOGRAPHY

105

BIBLIOGRAPHY

Allport, G. Personality. A_ psychological interpretation. New
York: Holt, 1937.

Atkinson, J. W. Towards experimental analysis of human motivation
in terms of motives, expectancies and incentives. In J. W.
Atkinson (Ed.), Motives in fantas , action and society.
Princeton, N. J.: Van Nostrand, 1958. Pp.—288-305.

Bijou, S. W., and Kenny, D. T. The ambiguity of TAT cards.
Journal o_f Consulting Psychology, 1951, 15, 203-209.

Brown, D. G. Masculinity-femininity development in children.
Journal g Consulting Psychology, 1957, _2_1, 197-202.

Clark, R. A. The projective measurement of experimentally-induced
levels of sexual motivation. Journal <_>_f_ Mgrimental Ps cholo ,
1952. 53.. 391-399-

Combs, A. W. The use of personal experience in thematic apper-
ception test story plots. Journal 9f Clinical Psycholo , 1946,
3: 357-363-

Cozby, P. Self-disclosure: A literature review. Psychological
M: 1973: 229 73-91-

Cramer, P., and Bryson, J. The development of sex-related fantasy
patterns. Developmental Psychology, 1973, 8, 131-134.

Dahlstrom, W. G., and Welsh, G. S. A_n_MV!PI handbook. Minneapolis,
Minn.: University of Minnesota Press, 1960.

DeCharms, R., Morrison, H. W., Reitman, W. R., and McClelland, D. C.
Behavioral correlates of directly and indirectly measured
achievement motivation. In D. C. McClelland (Ed.), Studies in
motivation. New York: Appleton-Century—Crofts, 1955. Pp. 414-

23.

Deutsch, H. The s cholo 9}; women. Vol. 1. New York: Grune and
Stratton, 19 .

Dollin, A. P., and Sakoda, J. M. The effect of order of presenta-
tion on perception of TAT pictures. Journal 5;; Consultigg
Psychology, 1962, 23, 540—544.

104

Epstein, S. The influence of drive and conflict upon apperception.
Progress Report, NIMH Grant No. M-1293, 1961. Cited in B.
Murstein, Theogy gang research i_n_ projective techniques. New
York: John Wiley and Sons, Inc. , 1963.

Eron, L. Responses of women to the thematic apperception test.
Journal o_f_ Consulting Ps cholo , 1953, _11, 269—282.

Eron, L., Terry, D., and Callahan, R. The use of rating scales for
emotional tone of TAT stories. Journal _g_f_ Consulting Psychology,
1950. H. 473-478.

Freud, 8. Some psychical consequences of the anatomical distinction
between the sexes. (Standard Edy), Vol. 19. London: Hogarth
Press, 1961.

Fried, C. Icarianism, masochism and sex differences in fantasy.
Journal 9: Personality Assessment, 1971, 35, 38-55.

Garfield, S. L. , and Eron, L. Interpreting mood and activity in
TAT stories. Journal 9; Abnormal and Social Ps cholo , 1948,

31. 338-345.

Goldfried, M., and Zax, M. The stimulus value of the TAT. Journal

9_f_ Projective Techniques and Personality Aisessment, 1965, 22,
46-57.

Gough, H. Identifying psychological femininity. Educational and
Psychological Measurement, 1952, 12, 427-439.

Hamilton, E. Myth—Olga. New York: Little, Brown and Co. , Inc. , 1942.

Heider, F. Social perception and phenomenal causality. Psychological
Review, 1944, 51, 558-574.

Heider, F. The psychology g_f_ interpersonal relations. New York:
John Wiley and Sons, Inc. , 1958.

Jourard, S. The transgarent self. Princeton, N. J .: Van Nostrand,

1964.

Kagan, J. Stylistic variables in fantasy behavior: The ascription
of affect states to social stimuli. In J. Kagan and G. S.
Lesser, (Eds.), Contemmragy issues i_n_ thematic apgrceptive
methods. Springfield, 111.: Charles C. Thomas, Publisher,
1961. Pp. 196-220.

Kagan, J. The acquisition and significance of sex-typing and sex
role identity. In M. Hofflnan and L. Hoffman (Eds.), Review 93
child develgiment research. Vol. 1. New York: Russell Sage,
1964.

105

Kagan, J ., Hosken, B., and Watson, S. The child's symbolic con-
ceptualization of the parents. Child Development, 1961, 32,
625-656.

Kelley, H. H. Attribution in social interaction. In E. E. Jones,
D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins and B.
Weiner (Eds.), Attribution: Perceiving the causes 9_f_‘_ behavior.
Morristown, N. J.: General learning Press, 1972.

Kohlberg, L. A cognitive developmental analysis of children's sex
role concepts and attitudes. In E. E. Maccoby (Ed.), The;
development 9_f_ sex differences. Stanford, Calif. : Stanford
University Press, 19662 Pp. 82-173.

Kohn, A. R., and Fiedler, F. E. Age and sex differences in the
perception of persons. Sociometry, 1961, 2_4_, 157-164.

Lebo, D. Immediate affective reaction to TAT cards. Journal 93
Clinical Psychology, 1955, _1_l , 297-299.

Lerner, M. J. Evaluation of performance as a function of performer's
reward and attractiveness. Journal 9: Personality and Social

Psychology, 1965, 1, 355-360.

Levinger, G., and Senn, D. J. Disclosure of feelings in marriage.
Merrill-Palmer Quarterly, 1967, 13, 237-249.

Lindzey, G. and Goldberg, M. Motivational differences between males
and females as measured by the thematic apperception test.
Journal 9_f_‘ Personality, 1953, g, 101-117.

Lindzey, G., and Heinemann, S. H. TAT: Individual and group ad-
ministration. Journal o_f_ Personality, 1955, _2_4, 34—55.

Lindzey, G., and Silverman, M. Thematic apperception test: Tech-
niques of group administration, sex differences and the role
of verbal productivity. Journal 9: Personality, 19599 2.1!

311-323.

Lubetsky, J. Assimilative projection as measured by trait attri-
bution. Doctoral dissertation, Northwestern University,
Ann Arbor, Mich.: University Microfilms, 1960, No. 60-4774.

May, R. Sex differences in fantasy patterns. Journal of Projective
Techniques and Personality Assessment, 1966, 32, 57?— .

May, R. Deprivation-enhancement fantasy patterns in men and women.
Journal of Projective Techniques and Personality Assessment,

1969. 553-464—469.

106

May, R, Further studies on deprivation-enhancement. Unpublished
Progress Report. Amherst College, 1973.

May, R, A method for studying the development of gender identity.
Developmental Ps cholo , 1971, 5, 484-487.

McClelland, D. C., Atkinson, J. W., Clark, R. A., and Lowell, E. L.
The achievement motive. New York: Appleton—Century-Crofts,

1953-

McIntyre, C. J. Sex, age and iconicity as factors in projective
film tests. Journal 2£_Consulting Psychology, 1954, 18, 475-477.

Mischel, W. .A social-learning view of sex differences in behavior.
In E. E. Maccoby (Ed.), The development gf_sex differences.
Stanford, Calif.: Stanford University Press, 1966. Pp. 56-81.

Meney, J. Differentiation of gender identity and gender role.
Psychiatric Annals, 1971, 1, 44-59.

JMurray, H, Explorations 12 personality. New Yerk: Oxford Uni-
versity Press, 193 .

Murray, H, Thematic apperception test manual. Cambridge, Mass.:
Harvard University Press, 1943.

Murstein, B, Theory and research.1n projective techniques. New
Yerk: John Wiley and Sons, Inc., 1963.

iMurstein, B. 'Handbook 2:.projective techniques. New Yerk: Basic
Books, Inc. , 1965.

Newbigging, P. L. Influence of a stimulus variable on stories told
to certain TAT pictures. Canadian Journal 2; Psychology, 1955,

9, 195.206.

Osgood, C. E., Suci, G. J., and Tannenbaum, P. H, The measurement
9; meaning. Urbana, Ill.: The University of Illinois Press,
1957-

Pederson, D. M., and Breglio, V. J. Personality correlates of actual
self-disclosure. Psychological Reports, 1968, gg, 495-501.

Prola, M. .A review of the transcendence index. Journal g§_Personality
Assessment, 1972, 6, 8—12.

Rapaport, D. Principles underlying projective techniques. Character
and.Personality, 1942, 19, 213-219.

107

Reitman, W. R., and Atkinson, J. W. Some methodological problems
in the use of thematic apperceptive measures of human motives.
In J. W. Atkinson (Ed.), Motives _in fantasy, action and
society. Princeton, N. J.: Van Nostrand, 19%. Pp. 664—684.

Rotter, J., Chance, J. E., and Phares, E. J. Applications 91";
social learning theory 9; personality. New York: Holt,
Rinehart and Winston, Inc., 1972.

Sanford, R. N. Physique, personality and scholarship. Mono ra hs
for the Society for Research E Child Development, 1943, _8,
No. 1.

Sarason, B. R., and Sarason, I. C. The effect of type of adminis-
tration and sex of subject on emotional tone and outcome ratings
of TAT stories. Journal o_f Projective Techniques, 1958, 22,
333-337-

Saunders, J. A theory of sex role development in the female child.
Unpublished Master's thesis. California State College at Long
Beach, 1971. Cited in R. May, Further studies on deprivation-
enhancement. Unpublished Progress Report. Amherst College, 1973.

Shapiro, D., and Tagiuri, R. Sex differences in inferring personality
traits. Journal 5g Psychology, 1959, 511, 127-156.

Silverstein, A. B. Identification with same sex and opposite sex
figures in thematic apperception. Journal _o_f Projective Tech-

nigues, 1959, 23, 73-75-

Smith, R. M. Sentence completion differences between intellectually
superior boys and girls. Journal 31: Projective Techniques and
Personality Assessment, 1963, _2_7, 472-480.

Tomkins, S. Affect, imagery and consciousness. Vol. 1. New York:
Springer Publishing Company, Inc. , 1963.

Veroff, J. Thematic apperception in a nationwide sample survey.
In J. Kagan and G. Lesser (Eds.), Contemporary issues E _thg-
matic gpperceptive methods. Springfield, 111.: Charles C.
Thomas, Publisher, 1961. Pp. 83-110.

Veroff, J. , Wilcox, S. , and Atkinson, J. W. The achievement motive
in high school and college-age women. Journal 9_f_ Abnormal and
Social Psychology, 1953, 48, 108-119.

Walster, E. Assignment of responsibility for an accident. Journal
o_f_ Personality and Social Psychology, 1966, 3, 73-79.

108

Weiner, B., Frieze, I., Kukla, A., Reed, L., Rest, S., and Rosen-
baum, S. Perceiving the causes of success and failure. In
E. E. Jones, D. Kanouse, H. H. Kelley, R. Nisbett, S. Valins,
and B. Weiner (Eds.), Attribution: Perceiving the causes 9;
behavior. Morristown, N. J .: General learning Press, 1972.

Weisskopf, E. A transcendence index as a proposed measure in the
TAT. Journal 91!; Psychology, 1950, _22, 379-390.

Winter, S. K. Characteristics of fantasy while nursing. Journal
9_f_’ Personality, 1969, 31, 58-72.

APPENDICES

Appendix.A
THEMATIC STIMULI
Picture.A
Picture B
Picture E
Picture F
Picture G
Picture H

verbal Description of TAT Stimuli

Picture A

 

 

‘1

3...: I. .2... . -. . . .. . . . ... ... . .
l1 H . i put“? .:1..1.“m .1. . .Iulahl..kh1‘.l «1.. LL, .1..th . .qup. .15.. .

Vt.

«A... . L,
we:

' . ”4.“. v a
73‘s

.‘4,

. M

...,‘,_‘..

.7 if}

A?!
I

IL. y . '0 .1 :3 l "‘ '3»; . :. .
sea a f.“ .. . , eaves... ares-4"

.v' ‘ ‘
“"7"“: ,-
3. '..{ .

. 2" 7 . .

 

E
e
m
it
c
i
P

 

  

I»

. .. ........:.. .5. .a .1.
.uliurr1lqlfintru'umnzf‘im .u .- ‘E w

I. .. . . . .... . q . I...
{fitnrurh‘ tin. . Cr: f .t I C»

 

Picture F

 

G
e
m
t
C
.l
P

 

 

 

Card

3BM

3GP

6BM

60F

7GF
813114
8GF
9BM
9GF

10

11

12M

115

verbal Descriptions of
Standard TAT Pictures

Description
A boy contemplating a violin laying in front of him.

.A girl holding books with a farm scene in the back-
ground.

.A huddled figure next to a couch with an unidentified
object on the floor.

.A girl in‘a doorway, covering her face with her hand.
A woman holding a man's arm as he turns away from.her.
A woman standing at an opened door looking into a room.

An elderly woman is standing at a window with her
back to a young man.

A woman is sitting and looking over her shoulder at
a man. The man is looking at her.

An elderly man is looking toward a young man close
to him,

.A little girl sitting with a woman on a couch. The
girl holds a doll and is looking away from the woman.

.A young man in the foreground with a surgical operation
taking place behind him, There is a rifle on the side.

.A young woman sitting in a chair gazing off into space.
Men sleeping in a field.

Two women running along the seashore.

A man and a woman in an embrace.

.A narrow stone path surrounded by rocks and steep
cliffs. In the distance, figures are rushing over a
bridge. The head and neck of a dragon appear on the
side of the cliff.

.A man reaching toward the forehead of a boy who is
lying on a bed. The boy appears to be sleeping.

116

verbal descriptions of the standard
TAT Cards, cont'd.

Card

12BG

12F

15B
13G

13MF

14
15

16

17BM
17GF

18BM

18GF

19

20

Description

 

.A forest scene with a rowboat on the ground next
to a stream.

A young person in the foreground looking off into
space. An old woman appears in the background.

A small boy sitting in the doorway of a wooden house.
A small girl climbing a long, winding staircase.

A man standing and covering his face with his arm.
A woman is lying in bed next to him.

A silhouette of a man standing in an open window.

A darkly-clad gaunt figure standing in the midst of
a field of tombstones.

The blank card
A young muscular man climbing a rope.

A girl looking over a bridge. Below her, a group of
men appear to be unloading something from a boat.

A man being held by three hands touching his shoulders
and arms.

A.woman with her arms around the throat of another
person. Both are standing at the bottom of a stair-

case.

A picture of clouds surrounding a snow-covered
dwelling.

A.man standing under a single street lamp.

Appendix B

THE SCORING SYSTEMS

117

Appendix B

The Scoring Systems

This chapter includes detailed descriptions of the procedure
for scoring Deprivation-Enhancement (DVE), Resurgence-Collapse
(R-C), the Autonomy-Heteronomy (AAH) of the pivotal incident, and
the scoring method for the sex role dissatisfaction questionnaire.
Scoring for Deprivation-Enhancement. The initial task.in scoring
any protocol for D/E is identification of the pivotal incident (PI).
The PI is the turning point in the action, the central act or feel-
ing that mediates between what went before and the final outcome.
This author's experience in scoring showed that the term "PI"
is acceptable as a general name for the "time immediately prior to
the outcome," since this broader definition includes the PI's of
some stories which do not have a dramatic, critical "incident" as
such, These stories are scorable because they do contain a chron-
ological development and an ending, but the subject's manner of
writing the story occasionally does not provide a fitting state-
ment that qualifies for a narrowly-defined PI. Hence, the PI refers
to an overt act or event whenever possible, but on occasion it refers
to a strong statement of feeling or desire or, rarely, the interim
before the outcome. Feelings and desires are acceptable as PI's
in May's system.as long as they directly relate to the outcome.
.Action statements, however, always take precedence over feelings in
choosing a PI. Stories which defy identification of a PI are not

scored, and usually comprise 5—10% of the total story pool.

118

After the PI has been identified, the story is separated into
pre-PI and post-PI sections and the number of references to de-
privation in the pre-PI and the number of enhancement units in the
post-PI sections are summed and given a positive score. The number
of references to enhancement in the pre—PI section is added to the
number of deprivation units in the post-PI section and given a
negative score. The difference between the positive and negative
scores is the final D/E score for that protocol. A positive score,
therefore, indicates that more deprivation units were found in the
pre-PI A_N2 more enhancement units were found in the post—PI sections
of the story, while a final negative D/E score indicates that more
pre-PI enhancement units A_1\I__D more post-PI deprivation units were
counted.

Deprivation units can be a variety of story references. State-
ments qualifying as deprivation units describe psychological or
physical discomforts, needs, poverty, professional obscurity (being
a newcomer), harm, exertion, failure or falling out of control.
Negative psychological states are also scorable, such as most nega-
tive emotions, pressure from external sources, ignorance, self-
sacrifice, and generalized desires when they mean "hope for" or
"yearn for" rather than "expect."

Enhancement is scored when the story author refers to physical
satisfaction if a desire for this had been previously established.
Enhancement references also include flying and other references to

height, positive emotions, positive shifts in tension level,

119

anticipation, receiving help or affection, success, fame, retalia-
tion, successful resistance, insight and realizations (even if the
realization is of something bad or undesirable). Growth is scored
as enhancement when the implication is good rather than growing
fat or ugly.

Separate units in a story are given separate scores if the
references have distinctly independent meanings. A.phrase such as,
"He is poor, tired and sad..." would receive three separate depriva-
tion scores because each of the conditions is independent of the
others. However, a statement that "She was shunned and humiliated
by her loVer..." would receive only one deprivation score, since
"shunned and humiliated" are close in meaning and the implication is
that they are simultaneous with one occurrence of rejection.

.A few examples of protocols help make the scoring procedure
more clear. The following examples were taken from data collected
by the author. The reasons for each score are listed in order,
under the appropriate heading.

Sggple Protocols
"Aurelio Rodrigues, a beginning fighter (D, +1), has 'ust fogght
his first bullfight (E, -1). It was very tense (D, +1), but
exciting (E, -1). Aurelio came from a cor family (D, +1) but he
had always dreamed of being a bullfighter D, +1). 'However, bull-
fighting has always been an occupation reserved for the elite and
prestigious. Thus,.Aurelio's dream was next to ' ossible (D, +1),
but he was a being of great fortitude and determination (E, -1).
Though he had 19 overcome many obstacles (D, +1), he succeeded (PI),

and today he has a right to be proud, as he is (E, +1), for his
great accomplishment (E, +1)."

 

Final score: +5

Deprivation Units: Enhancement Units:
1. professional obscurity 1. ’success
2. tension 2. positive emotion

120

Deprivation Units: Enhancement Units:
3. poverty 3. ability/resistance
4. strong yearning 4. positive emotion
5. desire is directly denied 5. success
6. difficult battle

"The man is the leading trapeze artist 1n;1h§_world (E, -1) and
the girl is more or less just learning the trade (D, +1). He is
performing for a large circus (E, —1), and is well-known for being
very daring (E, —1): After practicing for g long time (D, +1),
they become very good (E, —1) and they also fall 1n_love (E, -1).
The man doesn't want the girl §9_perform and.he wants to marry her
(E, -1). She agrees but wants to perform once just to prove her-
self. The setting is the night before the wedding and her pre-
miere performance. The man is thinking of how beautiful the girl
is (E, -1) and about the wedding the following day (E, -1). He
doesn't concentrate (PI) and so he fails 12_catch‘hg£_(D, -1) from
a triple sommersault (E, +1) and she falls (D, —1) to her death

(D9 ‘17:"

 

Final score: -8

Deprivation Units: Enhancement Units:

1. professional obscurity 1. accomplishment

2. physical exertion 2. attention, publicity

3. failure 3. skill and daring

4. falling 4. ability/precision

5. reference to death 5. love
6. concern for well-being
7. physical excellence
8. positive anticipation
9. elaborate trick

Scoring 2; Resurgence-Collapse. Fried's (1972) RFC scoring system
assigns a single score to each story, based on the outcome. Col-
lapse ratings may be strong (-1.5) or weak (-1.0). Strong collapse
refers to extreme and irreversible negative reaction to crisis, such
as death, insanity or mutilation. weaker collapse scores are
assigned when the outcome is marked by unrelieved negative affect.
A.mild resurgence rating (+1.0) is assigned to a realistic recovery
from a negative experience, while a strong resurgence score (+1.5)

is warranted only when the resurgence is extreme and spectacular.

121
.A strong resurgence score would be assigned to a story in which the
hero, who is frustrated here on earth, leaves to become ruler of
another planet.

R-C scores of zero are reserved for cases where no real action
occurs, for ambiguous outcomes, for passive acceptance of condi-
tions, and for such low-key action that no crisis can be said to
have occurred.

D/E and RFC scoring systems are similar but they are not re-
dundant. D/E is much.more sensitive to the events in the story
than R-C, especially in the zero range of the latter measure. On
the other hand, the only way in which intensity of the positive and
negative references can be reflected in D/E score is through quan-
tity of references, and even then, the references.must be located
in the right places to avoid cancellation. R-C is basically an
intensity measure and is therefore independent of story length.

In most cases, the direction of D/E and R-C ratings are con-
sistent with each other. Some specific kinds of stories, however,
are instructive because they illustrate the large discrepancy that
can occur between the two scoring systems. One particular story,
occasionally told to Card 3BM, is an example. The individual
is described as overcome by some misery. He decides to take his
own life (PI). The story ends with a description of relief. This
sequence is likely to receive a mildly positive score, but the R-C
rating is —1.5.

Autonomy-Heteronomy Scoring. The examples of autonomously caused and

heteronomously caused.pivotal incidents collected from the pilot

122
data are listed below. The criterion for classifying a pivotal
incident (PI) as autonomous or heteronomous was basically one of
whether the protagonist in the story obviously was the source of
the event, i.e., was the necessary condition for occurrence of
the event, regardless of intention. The hero's effort in obtain-
ing the goal state or outcome, or of avoiding it, had to be clear.
In the case of heteronomy, the PI had to be clearly beyond the
protagonist's direct influence or control. Occasionally, the
same PI could be either autonomous or heteronomous, depending on
the author's implications in writing the story. Such a PI might
be "time." Ekamples of how time might be classified as both auto-
nomous and heteronomous are given below, In scoring AAH, autonomous
PI's were given a +1, and heteronomous PI's were given a +2.
Autonomous Categories

Social competence

explaining one's feelings

active effort to solve marital problems

being attractive to opposite sex (when intent is stated)

doing what others want (as social manipulation. Actor's
free choice must be clearly present.)

control over interpersonal relationships

"tuning in" on each other

Task effort
working hard
skill, eg. in the bullring
overcoming great obstacles
persistence
making decisions
time (when context implies that the actors will "work it out")
willingness to compete

Self control
moving into a new house (thereby gaining independence)
maintaining one's head through tragedy
nonconforming acts
maintaining moral standards

123
Autonomous incidents, cont'd.

Self control, cont'd.
voluntary role playing
positive thinking
self confidence
keeping secrets

Heteronomous Categories

'Victim of circumstances
tragedy
poverty
overwhelming physical attraction (implication is one of
being "possessed," of uncontrollable passion)
immaturity (implication is that person cannot control world)
sudden appearance of new lover
accidental pregnancy
loss of job (for "no" reason)
God's bestowal of gifts

Impelled by external forces
the "show" must go on
family tradition
pride (forces one to act)
confusion
time (implication is that time "will tell me if I love you")

Blindness
ignorance of important facts
blindness:
to one's own stupid way of life
to interpersonal relations and feelings
role playing (emphasis on going through the motions without
emotion)

Impelled by social forces
inability to say "no"
living for another's approval
being "told," criticized or given attention

Falling
out of control

124

Scoring 9£_the sex role dissatisfaction questionnaire. .A copy of
the questionnaire appears in Appendix C, but the questions and
examples of answers and the scores they received are presented
below. The questionnaire was scored so that statements indicating
satisfaction with traditional norms received a score of "0." The
higher the score, the greater the number of statements expressing
dissatisfaction with traditional role expectations. Also, only
statements showing that their authors were desirous of the tradi-
tional role of the other sex were scored. Specifically, statements
referring to the activity-passivity dichotomy in sex role norms, to
independence-dependence or to concerns with personal effectiveness
were scored because these issues focus on the same kinds of com-
plaints made by subjects classified as "dissatisfied" by May (1969).
Answers which were irrelevant to sex role norms were not scored.

Six questions appeared on the questionnaire and five were
scored. The fifth question, "What do you think society's expecta-
tions are of you as a male (as a female)?" was asked as a reference
point in scoring the final question, "Do you think that living up to
those expectations is difficult?" Obtaining the subjects' answers
to the former question helped to guard against scoring the second
question incorrectly in case the subjects were not referring to the
traditional expectations.

Whereas most scorable questions received either a "0," "1" or
a "2" rating, the first question received either a "0" or a "1."
The answers were invariably short and most expressed satisfaction

with being a male or female. If any comment was included indicating

125
negative feeling about being a man or a woman, the answer received
a "1" rating.
Qpestion 1: "How happy are you about being a man or a woman?"
Answer Score
I'm.happy about being male because my traits
fit those of a male, dominant, 215 lbs., etc.(Male) 0
very content. Wouldn't want to be a male. (Female) 0
Being a woman, sometimes I think you get the raw end
of the deal... in a relationship with a guy, you
have to go along letting him.make the first moves. 1
(Female)
The second question asked the subject to describe growing up as
a boy or a girl. Descriptions of childhood experiences usually con-
sidered traditional for the opposite sex received either a "1" or
a "2," depending on the intensity of the statement, or in some cases,
multiple references to opposite sex activities. .A "0" rating was
assigned to statements expressing satisfaction or participation in
same sex activities.
gagstion.g; "What was it like growing up as a boy or a girl?"
Answer Score
Exciting, adventurous, football, irresponsible..(Male) 0
As a girl, you always had to help out with the house-
work, where as a boy, you didn't. I would get more
attention from.my father. (Female) 1
I was jealous, because Dad would let my little brother
do things, but not us girls just 'cause he was
the boy. (Female) 2
The third question asked subjects to describe the advantages of
being a member of their own sex. .Answers were scored according to

the sex role norms for adults in this society. References by a male

to being satisfied with the greater activity, independence and

126

assertiveness expected of him were given "0" ratings. Similarly,
a "0" rating was assigned to a female's answer if she said the
best thing about being a female is being passive, and being taken
care of. Scores of "1" and "2" would be assigned to a male's res-
ponse to Question 3 if he stated that the best thing about being
male is not having to be active or independent. Whereas these
scores were never used for male answers, females occasionally re-
marked that the best thing about being female is not having to be
passive and dependent. Hence, the female data necessitated the
"1" and "2" ratings. Answers which were irrelevant to activity-
passivity and/or independence-dependence were not scored.
Qgestion 3; "What do you see as being the advantages of

being a male or a female?" (Answer only with
respect to your own sex.)

Answer Score
A.male has more freedom to do what he wants. (Male) 0
I can do more athletic things than a girl. (Male) 0
To be able to do things for someone I like or love.

To accomplish something in life. (Male) 0
I don't have to try to be best at everything. Women

are supposed to be inferior. (Female) 0
Bringing up children and teaching them, (Female) 0
Having guys pick up the tab. (Female) 0
.A female has a lot of options. She can have a

career or be a housewife. (Female) 1
The best thing about being a female is proving

to a guy you're just as good as he is. (Female) 2

The fourth question asked subjects to describe what they think
would be the best thing about being a member of the opposite sex.
If males referred to the absence of pressure on females to succeed,
or to other traditional advantages of females, they received scores

of "1" or "2," depending on the intensity or number of references.

127
Males who said that females didn't have any advantages, or who

mentioned things irrelevant to the activity and independence
dimensions received "0" scores.

Females received scores of "1" or "2" on Question 4 if they
mentioned.male freedom or success potential as being the best thing
about being male or some related aspect of the male norm, such as
being protective, or strong. Females received "0" ratings if their
answers were irrelevant to these dimensions, or if they could not
think of any advantages of males.

Question 5; "What do you think would be best about being a
member of the opposite sex?"

Answer Score

(Example of irrelevant answer) WOmen are able to

get along well with each other. (Nble) O
WOmen are less restricted by society in terms of
emotional expression. (Male) 1

The only advantage that the opposite sex has in
American society is that they are pampered and
constantly put on a pedestal. Men have to
prove themselves. (Male) 2
The major advantage of being a girl would be that
you wouldn't have to work as much as a man.

The man spends most of his money on women. (Male) 2
(Irrelevant answer) .Men aren't as gossipy, and
get along with each other better. (Female) 0

Maybe as far as job opportunities go, I'd like to
be a man. The ability to travel alone without

society saying it's deviant. (Female) 1
Getting to ask a member of the other sex out on
a date. (Female) 1

That you could go wild, mostly. Playing in sports

and being more aggressive. Just having the

freedom. (Female) 2
Almost all subjects described the traditional norms in response

to Question 5, "What do you think society's expectations are of you

as a male or a female?" This question was not scored, but served as

128

an anchor for the last question, "Do you think that living up to
these expectations is difficult?" For those subjects who did
describe the traditional norms, the last question was scored as

"O," "1" or "2" for the degree of strain expressed in the answer.
Specifically, scores of "0" were assigned for answers which implied
that the expectations were not difficult and were well within reach.
Scores of "1" were given for answers which reflected a belief that
the expectations were difficult, but obtainable. Scores of "2" were
assigned for answers indicating rejection of the norm, or an in—
ability to meet sex role criteria.

Those subjects who did not express the traditional norms in
response to Question 5 invariably stated their belief that no ex?
pectations exist. These subjects fell into two groups. One group
answered the last question as if they had described the traditional
norms under Question 5, or they elaborated on the traditional norms
under Question 6. The second group responded to Question 6 by
reiterating that no expectations exist. Scores of "2" were assigned
to all of these answers for Question 6.

Question 6; "Do you think living up to those expectations
is difficult?"

Answer Score
No, it's easy and natural. (Male) 0

Sometimes I feel like I will never get ahead,
but then I look at how lucky I've been, and

things have always worked out for me. (male) 1
Oh, yea, they're so unrealistic to say the least.
Can't do it. (Nble) 2

I think that it isn't too hard to live up to
society's expectations. I like to be a girl.
I'm looking foreward to raising a family. It
isn't hard taking care of a home if you enjoy
doing these things. (Female) 0

129
Answers to Question 6, cont'd.
Answer Score
Yes, and especially if you'd like to have fUn and
not get married right away. (Female) 1

Difficult? Yes, because I couldn't stand it. 2
(Female)

Appendix C

INSTRUCTION SHEETS

130

Appendix C

Instruction Sheets
(P1 )
PERSONALITY’EXPRESSION

This study is concerned with individual differences in
personality and in personality expression. we are asking your
cooperation in examining the relationship between some of your
personal characteristics (as you see them), and the way you or-
ganize and express the meaning of a complex stimulus situation.
Other studies have shown that people organize experience in many
different ways which are meaningful to them, and to which they
can "relate." We wish to look for patterns in these operations
and their connections with other personality variables.

The first portion of this study enables you to attach.meaning
to ambiguous situations. This "organization" is not something you
will have to work at, but will occur rather easily when you look
at the stimuli. After this part of the study, the experimenter will
explain to you.what specific kinds of information we need, and you
will be provided with paper for your answers. We wish to assure you
that all information is greatly appreciated and is totally anonymous.
When you leave, your materials will be assigned a random.number
completely unconnected with your name. Thank you for your

cooperation.

131
(P2)

The folder in front of you contains four pictures, in no
specific order. Each picture has an identification letter/number
on the back. When you open the folder, look at all the pictures,
and place them.in the order given on the separate sheet inside
your folder. When the pictures are in order, turn them.face down
on the table in front of you. (The picture corresponding to the

first number on the order sheet should be on the bottom.)

132
(P3)

For this portion of the study, you are to write a story for
each picture. Please write stories that are as dramatic and psy—
chologically meaningful as you can. You.will be given six.minutes
for each story and the experimenter will watch the time for you.
First, you will be asked to pick up the picture on top and look at
it for twenty seconds. When this time is up, you will be asked to
place the picture face—down on the table, not on top of your other
pictures. Ybu may begin writing your story immediately on the first
of four sheets of paper provided in the folder. There are questions
on the sheets to help guide you in writing the story, but please do
not merely answer the questions rather than write a story. It is
very important that you write a continuous, whole story. Make sure,
however, to include answers to the questions at some point in the
story. BE SURE to include a.glga§_statement of what happens (or
did.happen, or will happen) in your story that leads to the outcome.
Concentrate on writing as dramatic a story as you can.

When you begin writing your story, try to spend the entire six
minutes allotted to that picture on the story for that picture. Do
not move on to the next picture before the time period is up, and do
not go back to any previously written story for any reason, Finally,
please do not pick up the picture again and look at it after the
twenty seconds for viewing it are over. The experimenter will tell
you.when the six minutes has elapsed. The same procedure will be

followed for each picture.

133
(P4)

For this portion of the study, take all four pictures and
look at them. Order is not important. .As you look at them, ask
yourself what the person (or people) in the picture is (are) like.
In pictures containing more than one person, you may wish to select
only one, but you do not have to. Specifically, ask yourself to
identify what traits or personal characteristics the person
possesses that makes or enables him.(her) to behave as he(she) is.
Ybu.might feel, for example, that the person is being honest, naive,
perhaps aggressive, bold or careless. There is a very large number
of possible characteristics you.might choose. Basically, trustnygug
feelings about the person in the picture to select the most obvious
characteristics shown there. 'Think about at least one person in all

four pictures. Make sure you don't skip any of the pictures.

1 311
(P5)

Now, select the picture that includes a person who best repre-
sents some characteristic(s) that you believe you yourself possess.
Your selection here should not necessarily be the picture that shows
your "best" traits, but simply the picture that comes closest to
showing a trait of yours. Please do not state what that trait is.
Rather, place the identification number/letter of the picture on the
FIRST line below. BE CAREFUL not to let the activity of the person
in the picture confuse you. Do not choose a picture on the basis of
whether you do the same specific activity, or ever would. (Many
people with very similar traits do not do similar things!) Choose
the picture according to how well the personal characteristics of
the person shown there match yours.

Look at the remaining three pictures, and select the picture
containing the person who you feel exhibits traits which are LEAST
like you. Place the number of that card on the LAST line. From.the
remaining two cards, select the picture including the person who is
more like you, and place that number on the second line from the
top. Fill in the number of the last card on the third line.

New, look to see where you have placed cards A and H. Next to
them, indicate which person in the picture led you to place the card

in that position.

 

 

 

 

135
(P6)

Copy the picture numbers from the preceding page on the four
lines below. IMake sure to copy them in the same order. Next to
each line, you will find a scale ranging from "1" to "9." Number
"1" means "very unlike me" and number "9" means "very much like
me." Taking each of your four choices, estimate where that picture

would be on the scale by circling the appropriate number.

Picture No,
1 . 1 2 3 1+ 5 6 7 8 9
very very
unlike much
me . like me
2. 1 2 3 1+ 5 6 7 8 9
3. 1 2 3 b. 5 6 7 8 9

136
(P7)

Please be as honest as you.can. Ybur cooperation is greatly

appreciated.

How happy are you about being a man or a woman?

What was it like growing up as a boy or as a girl? (Ybu can just
list adjectives if you wish.)

What do you see now as being the advantages of being a male or
a female? (Answer only with respect to your own sex.)

What do you think would be best about being a member of the opposite
sex?

What do you think society's expectations are of you as a male or as
a female?

Do you think.living up to those expectations is difficult?

137
(P8)

The questions below are guidelines. Write a whole story.

Who are the people? What has led up to the situation in the picture?

What is happening now? What are they feeling and thinking?

What happens next? What is the outcome? Remember: include a
clear statement of what happens that leads to the outcome.

138
(P9)

Place the pictures in the following order.

3m

139
(P10)

The folder in front of you contains four pictures, in no
specific order. Each.picture has an identification letter/number
on the back. For this portion of the study, take all four pictures
and look at them. Ask.yourse1f what the person (people) in the
picture is (are) like. In pictures containing more than one person,
you.may wish to select only one, but you do not have to. Specifically,
ask yourself to identify what traits or personal characteristics the
person possesses that makes or enables him (her) to behave as he
(she) is. You.might feel, for example, that the person is being
honest, naive, perhaps aggressive, bold or careless. There is a
very large number of possible characteristics that you might choose.
Basically, trust your feelings about the person in the picture to
select the most obvious characteristics shown there. Think about
at least one person in all four pictures. Make sure you don't

skip any of the pictures.

1110
(P11)

New, look at all the pictures again. Place them in the order
given on the separate sheet inside your folder. When the pictures
are in order, turn them face down on the table in front of you.
(The picture corresponding to the first number on the order sheet

should be on the bottom.)

1M1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(P12)
PICTURE NUMBER
accepting ; ; ; ; ; ; rejecting
1 2 3 h 5 6 7
active ; ; ; ; ; ; passive
1 2 3 E 5 5 7
dependent ; ; g ; g ; independent
1 2 3 *‘E 5 Z? 7
happy ; ; ; ; 1 ; sad
1 2 3 5 5 6 7
hopeful ; 3 ; ; g ; hopeless
1 2 3 R 5 6 7
impulsive ; ; g g ; ; controlled
1 2 3 h 5 6 7
pleasant ; ; ; ; ; ; unpleasant
1 2 3 4 5 6 7
safe ; ; ; ; ; ; dangerous
1 2 3 E 5 6* 7
severe ; ; ; g g ; lenient
1 2 3 h 5 6 7
calm. ; ; g ; g ; excitable
1 2 3 E 5 6 7
STOP!!! Before turning the page, make sure you have

placed the picture number on the line above.

Appendix D
FACTOR ANALYSIS DATA

Table 20. Factor structure of stimuli from stimulus
cue study: Remaining pictures

Table 21. Mean ratings of pictures on semantic
differential

142

 

 

 

 

momm. mem. when. hmwr. mama. omnn. .moqdfins> no :owpnonoam
*emom.u meme. monr.u memm.u*mmmm. Nmmm. .eamoumanapaoxm
*maea.n mmoo. maao.u *omso.u memo. sero. e.aaoaveooum>amasgaa apa>flpo¢
mmao.u *ooma. mama. amom.u*nmmm. mnmm. m>ammaonm>apoa
mmmo. *mmao.n mama. *smeo. Psor.u moan. wtm>mmupqmflcma
some. mwmm.l *Fvoo. #:omm. mmwm.l pram. odoamwodolmmom homopom
*mmmn. omoo. onnm. mm? . £32. . mtb . Sweemomefinpemeamome
omen. maor.u *Fmaa. mmam. memo. *soem. pcammmaoqsupemmamao
mmem. amen. *wome. ommr.u asnr.u *aooo. mmmawooeuapooooe
scavenge
omra. amao.n *mmom. somm. some. *Fsmm. emanaooee
emoe.n maom. *mmao. ommm. memo. *ommn. weapomamequapomoom
HHH HH H HHH HH H accomqmaao meoposa
N e o>apomno<. HwflpomaoMMfiQ
. . oaPGdSom
.1manaapm

 

 

sespm was msflsaflpm scam sassflpm mo weeposppm nopoaa

ON HAMdH

1A3

weer. mamm.

mmom. ores.

warm. omws.

«moqowao>_mo defipaomoam

 

mmee.u smam.u

*mmmm.u mmmo.1

mamo.s *mmam.u

aHmoumHnspaoxm

 

 

 

mmmo. maem. *aemm.u eeoo.u meeo. *Hmam.u emHHoepeooum>amHsasH apH>Hpo<
mmro.u momo.u *mmra.u mmwo.u aomo.u *sasm.u w>Hmmmoum>Hpom
move. sews. *Mwwm. oosm.1 ammo. *memw. opm>mm1pcmflooa
memo.1 *mmmm. mmom. Nwas.u meso. *mmem. msopmwcmeumomm socmpom
*Psmm. mmer.u mmPH. omma.u*mon.u mmmo. guacamomeanpcmecmome
mmom. *mmma. mmme. mmmm.u*.mmm. mmos. pnsmamHoesupnamamHo
*mmme. memo. om:m.u mmrr.u*mmmm. ooso.1 mmmHmooeuHsowooe
cowpodao>m
emoo.u *moem. ommm.- mamm.u*ammm. meme. samuaooma
*merm. memo. Hmmo. *mmma.n come. earm.1 meromnmeanHpomooa
HHH 2““ H HHH “H H Mmmwmemwo HawmmmmmmmHo
HHzaHpm

 

 

A.U.PSOOV ON mumdfl

14A

 

 

 

 

mmoF. room. seme. aeoe. «mom. mmmm. .moeaHpm>.oo ooapgogogm
ommm.n*omHo.u momm.n mmma. *mmma. mamm.u _aHooumHnopHoxm
*Hooe.u moor. smHH. Hmm:.u *mmmo. mmee.u omHHoopooouo>HmHsoaH apH>Hpo<
memo. *oamm.n momo. msmo.u *rama. mmar. m>Hmmooum>Hpoo
mom..- mess. *smmm. Fmom.a *Hmom.u meam. mom>mmupomHowH
owsm. Norm. *mmoo. mmso.u omam.u *mamo. msoomwcsoumoom aoeopoo
mmmo. *sram.u Psoo.u *aamm.u Hmeo. moss. pomoooomooHupomoemomo
aaom.u moao. *omem. mmma.u onH.u *mmom. posmmmHoqsupowmomHo
memo.u mmoo.n *asom. momr.1 mmmo.u *Hmmo. mmmHmooeuHsomooo
cosposHo>m
oewo.u mmmo.1 *oaom. >_m3.1 mama. *mmmo. oomnaoooe
euro. Fomm. *ooom. *memo.u Howm.u omma. moHpomomaumoHpooooo
HHH HH H HHH HH H mooaoooaHo Hmdmmmwwwwdo
mom EN. o>flpoono4 mangom.
HHserm

 

 

A.o.poooV om mamas

1A5

 

 

 

 

Sow. 2.9. moom. omme. mmot 8+3. ”manage. no ooHfiooonm
mHHm.u*HmHo. moma.1 *mamm.u zoom. omma.u eaoouoHnopHoxo
ermm. *omrm. Heme. *mmmo.u memo. aeoa.u omHHonpoooumpHmHsoaH apa>Hpq¢
meme.u*mmra. ono. omoo. *aamo. moor. o>Hmmooum>Hpoo
emsm. *oomm.n mmom. ooeo. *omma.u Hmae. wom>mmapooHooH
*saam. mmar.n mmme. pose. mamr.a *smrm. msoommomoummom soompom
*msma. momH.a momo. *smom.a somr.n oomo. pomoewomooanomoomomo
.mmr. mmso.1 *smmm. memo. mmHH.1 *mmom. pesosmHoqsupoomomHo
mmsm. pomp. *aoma. mmme.u mmPo. *omma. mmonooaanomooo
:OHpmsHmsm
mooo. ooam.1 *mmmm. memo.n some. *smem. oomuaoooo
*mmmo. waoo. mmrs. mmoo.u maer.n *ommm. moHpomooouonpomoom
HHH HH H HHH HH H mooameosao mnopomm
or 2mm m>apomoo«. HaHpomommoHo
OHpouamo
HHosHem,

 

 

7.01503 ON a

146

moor. momm. ommm. ewes. emom. samm.

" condenser mo GOHPHOQPHA

 

mmmm.a *ommo.u Hmmm.u mmmm. *oowo.u mHFH.1

floolmansfioxm

 

 

 

mmae.u *aooa.a mmem.u *momo. momo. mamo.1 omHHoopoooum>HmHsoaH opH>Hpoa
*msmm.n Hmom.u memo. Name. *mmra.1 moao.n m>Hmmmonm>Hpoo
mmmo. mmee. *Fooo. woos. woos. *wmmo. mom>mmupechmH
HmHH. *maoa. ammF. ommm. ammo. *seoa. msooowomoumoom socmpom
mmow. ammo.u *mmom.u *momm.u ommm. ammo.u pomoomomooanpomoomomo
mmmm. oomm. *amao. mmwo. mama. *mmmm. poamomHoosupoommoHo
mama. oamo.u *mmaa. amsa.n mmam. *aamo. mmmHmooouHsomooa
ooaposHmsm
ammo. mmmm. *mmem. aomo.u meme.u *Hsma. oomnaooae
*mmra. moor. omar. emmr. *mmma. Famm. qupomnmnumoapomooo
HHH HH H HHH HH H mooHoemaHo Haammmwwmmao
E F 2N r “65906.34. ".6333
HHoemom

 

 

A.o.peoov om mamas

1117

mooa. mmoo.

mmmm. moaa. mmmo. mama.

328.39, .Ho scoohoaosm

 

amma. aoaa.-

mmao.._ aomo.u mama. oowo.u

aHooumHoopooxo

 

 

 

*aoom. oamo.u oomo... momo..*momo. maoe.- omHHoopoooum>HmHsoeH apo>opo<
ammo. *oaaa.u moma. aomo..*omam. omoo.- m>oomooum>opoo
*mmmm. ommm. oooa. ommm.- moam.u *ommm. mnm>omupomocmH
amoa.u memo. *ooam. maoo.n mamm.u *mamo. msoumwomonmoam monopom
*amom.1 omma. ammo.u *mooa.1 mmmo. moma.a pomooommoooupomoooomo
aaae. moao.n *omoa. ammo. mooo.u *moow. poommmHoos1poomsmHo
mmoo.a osam.u *mmmm. oooa. amoa.u *amom. mmonooorHsomooo
ooopmsHoam
mmoo. ommo.u *mooa. oamo.1 mmaa.u *mamw. oomuaooae
mmao. mmmo.: *oama. ooom. ammo.u *asaa. weapomomouwaopomooo
HHH HH H HHH HH H mcoHoomsan Hoammmwamao
Ea a a _. 9590034 wfipqmamm
HHsemom

 

 

A.o.poooV oo mamas

148

 

 

 

 

mama. amma. :amm. aama. sooo. ommm. ”moooHnoa mo ooHpooooom
mama.u *omaa.u omoa.u omao.u *msmo.u omao. _aHmouoHoopHoxm
*mmam.1 ooom.1 moaa. *maom.: moma. aomo. omHHoopooono>omHsoso apo>opoa
*wsom.u mamo.n aaoo. *mmmm.u Famo.u mmmo. m>Hmmoonm>Hpoo
amoo. *amoo. oaso.n *maam. mesa. mmom. moo>omupoooooH
*nmm©.1 moaa. mmmo. a:©N.I :a©:. *Nwam. moosomcdolmmom monopom
mmoo.u amam.1 *momo. mmao. *mmao.u aoaa. pomoemomoooupomoomooo
maoe.n oooo.- *mmma. moaa. ammo.u *ammm. pommomHooonpoamooHo
momo.u mooo.u *mmam. mmoo.1 omoo.s *asma. mmmHmoooanomooo
cooposHo>m
maaa.u amoa.u *mamm. omoa.u moam.u *smaa. ommnaoooo
aomm.1 mmmo. *aomo. mooo.u maae. *oomm. ooHpomomouonpomooo
HHH HH H HHH HH H mooamomaoo mompoom
m oo m>opomno<. HaHpo omooHo
ooPQdEom
HHssopm

 

 

A.U.PQoov 0N mAde

aoma. smao. mamm. sooa. mmao. omoa. ”mooano> mo cOHpeooonm

 

 

149

msmo.u *moma. moma.u ommo. *mmaa.u mmao.- sHaoumHnmpooxm

amao. *oaso. ammo. oamo. *ommm.u ooao. omHHoppoooum>HmHsoeH mpH>opo<
maom. *aaam. ooao.1 ammm.u *aama.u mmaa. m>ommooumsopom

aaoo. *omme.u Paar. *aaom.u oaso. asoo.u mom>omupoooomH
*mmam. awom.u ewes. ossa.u omma.u *mooa. msoommosouooom aoompom

oaoo.u esao.u *omam. oomm. aama.u *mmoo. pomoooomooo1pemoomomo

 

 

asao. ammo. *oesm. mamo. ooaa. *oosa. pouoomHoosupoomamHo
*mmao. amoo. oooa. oamo.u moaa.1 *mama. mmmHmoooquomooe
_ coopmsHaam
amoa. oaoo.u *maom. oaam. moao. *ooma. oamuaoomo
ooma. mmmo.: *mmam. amoa. ammo.u *omoo. moHpomompumoopomooo
HHH HH H HHH HH H moOHmooaHo moopoom
o a m>apomooa. HoopcmomooHo
. . owpcdeom
HHssHpm

 

A.o.pcooV oo mamas

150
TABLE 21

Mean Ratings of Pictures
on Semantic Differential

 

 

 

 

Pictures

Adjective Dimensions 1 2 38M 4 6BM
accepting—rejecting 3.72 3.44 5.08 4.44 4.54
happy-sad 5.18 4.58 6.36 4.80 5.70
hopeful-hopeless 4.84 3.38 6.12 4.12 4.60
pleasant-unpleasant 4.86 4.02 5.94 4.98 5.44
dependent-independent 3.92 4.36 3.90 3.62 3.88
safe-dangerous 3.38 3.30 5.04 4.74 4.50
lenient-severe 4.16 4.22 5.30 5.12 4.82
active-passive 5.06 3.40 4.00 2.52 3.76
impulsive-controlled 4.66 4.62 2.64 2.46 4.42
excitable-calm 5.14 5.02 3.38 2.66 3.72
7BM 8gp 9BM. 10 12M

accepting—rejecting 3.76 2.52 3.48 2.00 3.88
happy-sad 4.58 3.24 3.76 3.28 5.20
hopeful-hopeless 3.98 2.66 4.42 2.84 4.14
pleasant-unpleasant 4.48 2.78 3.52 2.92 5.12
dependent-independent 3.96 4.56 4.14 2.08 3.36
safe-dangerous 3.92 2.68 3.22 2-62 4.78
lenient-severe 4.56 3.70 3.02 3.26 4.66
active-passive 4.24 5.58 5.82 4.06 3.96
impulsive-controlled 4.50 4.80 4.50 3-62 4.20
excitable-calm 4.22 5.92 5.64 4.88 4.54

 

 

1

51

TABLE 21 (cont'd.)

 

 

 

 

Pictures

.Adjective Dimensions 13MF 14 17BM 20 .A

accepting-rejecting 4.88 3.82 3.36 4.44 2.96
happy-sad 5.98 4.32 3.42 5.40 3.22
hopeful-hopeless 5.82 3.44 2.98 5.20 2.96
pleasant-unpleasant 5.74 3.80 3.52 5.24 3.28
dependent-independent 3.46 4.94 4.62 5.06 2.32
safe-dangerous 4.78 4.00 4.30 5.10 5.80
lenient-severe 5.10 4.52 4.48 4.52 4.72
active-passive 3.26 4.66 1.80 4.82 1.48
impulsive-controlled 2.72 3.92 4.32 4.26 5.08
excitable-calm 3.00 4.98 2.98 4.68 2.66

B E F G H

accepting-rejecting 2.92 2.58 1.74 2.56 1.94
happy-sad 3.00 3.54 1.68 2.50 1.94
hopeful-hopeless 2.92 3.08 1.78 2.42 1.70
pleasant-unpleasant 3.86 3.26 1.56 2.14 2.02
dependent-independent 5.04 3.24 2.66 3.54 2.58
safe-dangerous 5.30 3.06 1.98 3.16 2.40
lenient—severe 5.02 3.92 3.54 3.68 3.66
active-passive 2.20 3.62 4.20 3.58 3.66
impulsive-controlled 4.46 4.34 4.36 3.48 3.12
excitable-calm 2.92 4.56 5.02 3.76 3.96