I

1|

1

122
200

”THS

 

ANTXtETY,‘ W'GMATISM AND S'l-IADING
(Y.YF) RESPONSES ON THE
RQRSCHACH qFEST

Timests (“or five Dog?“ 0:"? M. A.
HICHEGEN STATES WEE’ERSETY
Janice I. Potash
1964-

TH ESIS

LIBRARY

Michigan State
University

 

ABSTRACT

ANXIETY, DOGMATISM AND
SHADING (Y-YF) RESPONSES
ON THE RORSCHACH TEST

by Janice I. Potash

Shading or Y-YF responses to the Rorschach Test are gener-
ally considered indices of anxiety within the context of the
Rorschach protocol. This study was designed to test the valid-
ity of interpreting Y-YF responses as indicators of anxiety.

It was hypothesized that individuals high in anxiety would
give more Y-YF responses to the Rorschach than would individuals
low in anxiety.

Sixty-two freshman males were group tested on the Taylor
Manifest Anxiety Scale and the Psychology Questionnaire, or
Dogmatism Scale. The latter instrument was used as an additional
measure, as the dogmatic individual's anxiety seemed to be theo-
retically similar to the anxiety Klopfer refers to in his inter-
pretive hypothesis of Rorschach shading responses. This anxiety
is of a "diffuse and free floating nature, reflecting a frustration
of affectional satisfaction." It was thus hypothesized that those
individuals high in dogmatism would also be highly anxious. This
relationship was confirmed by a significant chi-square and a
significant correlation between the measures of dogmatism and

manifest anxiety.

Janice I. Potash

Individual Rorschachs were administered to 47 of the students
who had participated in group testing. It was found that those
individuals high in either or both Taylor anxiety and dogmatism
did not give significantly more Y-YF responses to the Rorschach.

The study also postulated that freshman males high in
dogmatism and anxiety would have parents scoring high in
dogmatism. The parents of the subjects were administered the
Psychology Questionnaire, by mail, and 50 pairs responses with
completed tests. No significant relationship was found between
parental dogmatism and subjects’ dogmatism, or between parental
dogmatism and subjects' manifest anxiety.

The results of this study then Show that pre-adult males
are neither anxious nor dogmatic because of parental dogmatism.
Those individuals highly dogmatic are also highly anxious, but
they do not indicate these states by giving proportionately
more Y-YF responses to the Rorschach. Therefore, it is believed
that Y-YF or shading responses to the Rorschach are not indices

of anxiety within the Rorschach protocol.

Approved‘r‘zm Chairman
Date W 7.”.6‘)‘

ANXIETY, DOGMATISM AND
SHADING (Y—YF) RESPONSES

ON THE RORSCHACH TEST

By

Janice I. Potash

A THESIS

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

MASTER OF ARTS

Department of Psychology

1964

ACKNOWLEDGMENTS

I want to express my deepest gratitude to my husband,
Herbert, for his continual and inexhaustible support during
this research project. His suggestions and ideas were of
immeasurable assistance in the preparation of this paper.

I am very grateful to Dr. C. L. Winder for finding time to
be chairman of my thesis committee and for the many helpful
suggestions he offered concerning the study.

I wish to thank Dr. R. E. McMichael and Dr. G. Marian
Kinget for offering comments and advice on the research and
for serving on the thesis committee.

I am also endebted to Dr. Charles Hanley for statistical
consultation, and to Dr. Hans Toch for his aid during the

initial phases of the research.

ii

TABLE OF CONTENTS

LI ST OF TABLES O O O O O O O O O 0 O O O O O 0

Chapter

I.

II.
III.
IV.

V.

BIBLIOGRAPHY

APPENDIX

HISTORICAL AND THEORETICAL BACKGROUND
Hypotheses
PROCEDURE , , , , , , ,
RESULTS

DISCUSSION

SUMMARY 0 o o o o o o o o o o o o o 0

iii

Page

iv

16
17
2O
25
28
30

32

LIST OF TABLES

Relationship between subjects'
and dogmatism scores . . . . .

Relationship between subjects'
and anxiety scores . . . . . .

Relationship between subjects'
and anxiety-dogmatism scores

Relationship between subjects'
anxiety scores . . . . . . . .

Relationship between subjects'
dogmatism scores . . . . . . .

Relationship between subjects'
dogmatism scores . . . . . . .

iv

shading responses

shading responses

shading responses

dogmatism and

and parents‘

anxiety and parents'

Page

21

22

22

23

24

24

Historical and Theoretical Background

According to most clinicians, the shading response to the
Rorschach Test can be interpreted as an index of anxiety. This
is in line with the hypothesis of Rorschach psychologists,
specifically KlOpfer, who state that such responses (K and KP)
"indicate anxiety of a diffuse and free floating nature,
reflecting a frustration of affectional satisfaction." (7, p.
268) However widely accepted such an interpretation is, the
literature reveals a paucity of research directly aimed at
validating this hypothesis either per se or in the context of
an entire Rorschach protocol. In a 1940 summary of the liter-
ature on the shading response, Hertz (6) stated that although
meager, the evidence seemed to support the relationship between
the K or Ch categories and anxiety, depression, fear, and
uneasiness. The purpose of this paper is both to review the
relevant literature, post 1940, and to report a research study
designed to investigate the above aspect of the Rorschach.

The major empirical contribution "designed to investigate
the validity of the interpretation of anxiety from the shading
response on the Rorschach" is a study by Waller. (14) She
hypothesized that there would be a positive correlation of
shading responses on the Rorschach with other clinical measure-
ments of anxiety, in a psychiatric population; and also that
use of the Baughman paired comparison inquiry would lead to a
greater number of shading responses than obtained with the
regular inquiry, as well as enhance any existing relationship

1

2

between shading responses and other anxiety measures. The
subjects were 60 literate psychiatric in and out patients who
were not psychotic and had no organic impairment. The
procedure involved an interview with each subject from which
two ratings were made: one of overt clinical signs of anxiety
(handwringing, etc.), and another involving a seven-point
scale of overall impression of anxiety. In addition the
subjects were given the Vocabulary and Word Fluency scales
from Thurstone's P.M.A., the Welsh Anxiety and Taylor Manifest
Anxiety scales.

In the Rorschach testing, 30 of these subjects were admin-
istered the free association and regular inquiry; the other 30,
the free association and the Baughman paired comparison inquiry.
The latter procedure involves use of each particular Rorschach
card during the inquiry plus a series of similar cards in which
various determinants of the Rorschach card are omitted. The
individual is asked if his percept of the Rorschach card is.
still present in the Baughman card; and thus the exact determ-
inant is presumably specified.

It was determined that the subjects were comparable in
P.M.A. scores, education, and total number of responses to the
Rorschach. In scoring the shading responses (Y,V,T), the reli-
ability between judges ranged from .95 to .98, without regard
for the particular kind of shading. For analysis the shading
responses were weighted and totalled, with the exception of

texture responses which were analyzed separately. The one

3

hypothesis confirmed of the prOposed three was that the use of
the Baughman paired comparison inquiry led to a greater number
of shading responses than that obtained with the regular
inquiry procedure (significant beyond .01 level). No signif-
icant relations were found between any of the objective anxiety
ratings or test scores with the shading responses on the
Rorschach. However, a relationship between texture responses
and overall impression of anxiety was noted, which lends some
support to the aforementioned first hypothesis.

In conclusion the author questioned the use of anxiety as
a unitary concept. This observation may point to a possible
explanation of the negative results for the first hypotheses;
the anxiety reflected by shading responses on the Rorschach
may not, in fact, be exactly the same kind of overt anxiety as
that measured by either the Taylor or Welsh scales or the
behavioral ratings.

Several criticisms can be made of Waller's study. One
could question the rationale for using the Taylor and Welsh
scales with a psychiatric population. However valid such a
paper and pencil test is, the caution exercised in making infer-
ences from the scores of a "normal" population should be inten-
sified when interpreting psychiatric patients' scores. In
addition, the subjective evaluation of an individual’s anxiety,
based on a single interview, leaves much to be desired in the
way of accuracy for research purposes. However, the major flaw

in methodology appears to concern the treatment of Y, V, and T

4

responses as equal indices of anxiety. The investigator
attempted to discover the validity of interpretations of
anxiety from shading responses on the Rorschach. The theoret-
ical hypotheses for this interpretation do not refer to all
shading responses (Y,V,T) but specifically to the K or Y
determinants.

Eichler designed a study for the purpose of investigating
the effects of stress-produced anxiety upon 15 Rorschach
factors, commonly declared to be signs of anxiety. Among the
15 hypotheses the one concerning shading responses predicted
an increased number of weighted shading responses (FSh =.5,

ShF = 1., Sh = 1.5) as a function of stress-produced anxiety.(4)

Sixty college students matched on age, college year, cumu-
lative grade point average and on certain factors on the Behn-
Rorschach test, were equally divided into experimental-stress
and control groups. Experimental stress was produced by the
administration of electric shock of increasing strength, the
examiner's implications of stronger shock to come, and the use
of certain equipment for effect. Both groups were given contin-
uous subtraction problems immediately before the Rorschach.

The experimental group were shocked during the subtraction but
not during the Rorschach. After the Rorschach both groups
completed self-ratings on an anxiety-tension scale.

The results showed that the experimental group made signif-
icantly more mathematical errors and were significantly more

anxious or tense on the scale. In addition, four of the

5

Rorschach indices reflected the conditions as predicted, and
among these four was an increase in weighted shading responses.
These positive results were considered as evidence for the role
of weighted shading responses et al. as anxiety indicators on
the Rorschach.

Allerhand (2) attempted to discover if there was any rela-
tionship between the Chiaroscuro (shading) responses to the
Rorschach and the manner and degree of expression of anxiety in
experimental problem-solving situations. Fifty college students
were each subjected to two problem-solving situations, similar
except that one involved no conflict and the other, an
avoidance-avoidance conflict. The anxiety criterion was based
on a series of behavioral movements, rated as suggestive of
anxiety by clinical psychologists and psychiatrists. The
Rorschach protocols and manifest anxiety inventory scores were
obtained before the problem-solving situation; there was a
lapse of three weeks between the two sessions.

A decided increase in the number of "anxiety" responses
was emitted by the subjects in the conflict situation compared
to their reactions in the nonconflict situation; therefore, it
was concluded that the conflict situation was in fact eliciting
an increase in anxiety.

Of the 66 correlation coefficients between the shading
variables and anxiety criteria variables, 13 were significant
(which is greater than the number expected by chance). The

texture determinants were most consistent and effective in

6

indicating overt anxiety, and also showed a significant
relation with factors which tapped the subject's mediate and
immediate reaction to a conflict situation. These determinants
were correlated with the total of the individual's anxiety
signs during the nonconflict and conflict situations and in

the test periods after each problem-solving situation. The
factors Ch and Y were also significant at the five per-cent
level on the prediction of anxiety potential.

These last two studies carried out to a fine degree what
they purported to do and restricted experimentation and results
to a specific situation with well defined criteria. The three
previously discussed studies, therefore, indicate no
relationship between clinical measures of anxiety and shading
responses on the Rorschach, but do provide two cases of a
positive relation between experimentally induced anxiety and
shading responses on the Rorschach.

The two following articles deal with the Rorschach and
anxiety, using children as subjects. Sarason, et al., (11)
investigated the Rorschach behavior and performance of high
and low anxious children matched for grade (second through
fifth), sex, and intelligence scores. The measure of anxiety
was obtained by the use of Test Anxiety and General Anxiety
scales, constructed by the investigators for this research
project. A relatively brief and blind clinical analysis by
Sarason of the Rorschach protocols for each of the 32 high and

32 low anxious children, resulted in significant discrimination

7

with the p value being .03. However, Sarason did not specify
which aspect of the Rorschach he used to make the discriminations.
The 64 records were then arranged into the 32 matched pairs; 18

of the 32 were correctly categorized, the result being signif-
icant at only the .30 level. The highly anxious children,

among other things, responded less to any aspect of color
(chromatic, achromatic, or shading) on the cards than did the

low anxious subjects.

Levitt (8) approached the same area by a different route,
in an attempt to assess children's use of the Rorschach
variables which are commonly regarded as anxiety indices in
adults. (These indices are low on W, P, M, sum C, and total
number of responses; high on Hd, Dd, A, and sum Y). The
subjects were 39 children's clinic patients, mean age of 10.70,
and 155 normal children, mean age of 10.75. Of nine signif-
icant differences (related to the analysis of average frequency
or percent), only the frequency of the shading responses was in
the predicted direction, high on sum Y. With the group of
clinic patients no Rorschach variable was found related to the
scores on the Children's Manifest Anxiety Scale. The inves-
tigators determined that neither the effects of the differential
sex ratios within the two groups nor the variabilities of age
between the two samples significantly influenced the results.
It was concluded that, thus far, only the frequency of the
shading responses on the Rorschach can be considered an anxiety

indicator in children.

8

Levitt's study makes a valuable contribution in showing
that a difference exists between Rorschach anxiety indicators
for children and for adults. The results found with the
children‘s clinic patients' Rorschach protocols and anxiety
scores raises a relevant question - the sagacity of using
objective anxiety scales as anxiety measures of clinic patients.
In a comparison of Levitt's research with Sarason's, however,
the most obvious facet is that the studies present contradictory
results. It seems reasonable to attempt to explain this contra-
diction in terms of the different mean ages of the two studies'
groups, the different measures of anxiety used, and especially
the different methods of analyzing the data.

These five articles do not seem to be particularly strong
pieces of evidence for either accepting or rejecting the
anxiety interpretation of the Rorschach shading responses.

Among the weak points of the studies, the major difficulty
seems to lie in delineating a specific, independently measured
type of anxiety. It is the writer's belief that such a measure
of anxiety would contribute a great deal toward validation of
the major hypothesis.

Before presenting theoretical background for such a
research project, it would seem wise to clarify the writer's
views on any research which attempts to validate a Rorschach
interpretive hypothesis. In making interpretations from a
Rorschach protocol, one never makes an inference from one

response without also considering the response in the context

9

of the entire protocol. In other words, there is no one-to-
one relationship between any Rorschach variable and an aspect .
of the personality. Perhaps in view of this, a research study
to validate the anxiety interpretation of Rorschach shading
responses seems contradictory and of no value. However, it is
the writer's contention that the interpretive hypothesis bears
testing not as the specific determinant of anxiety but as a
part of a theoretical framework of Rorschach interpretation.

In interpreting the Rorschach responses in their context,
each determinant has associated with it a tentative interpretive
hypothesis. Although we cannot validate an interpretive hypoth-
esis as a specific and universal indicator simply because it
does not purport to denote a one-to-one relationship, we can
assess the validity of the hypothesis‘ position within the
entire interpretive framework. The research here will not
attempt to establish a one-to-one relationship between the K
or Y responses and anxiety, but will try to indicate the
strength of the relationship. It would seem that with a group
of individuals, independently designated as high in a specific
type of anxiety stemming from frustrated dependency needs, one
could expect more Rorschach shading (K, KF or Y, YE) responses
than from another group low in the particular anxiety. With
these points in mind the next step is setting ourselves to the
task of specifying a type of anxiety that is not only present
in the individual's overt behavior but also has specified roots

in the individual's character formation and development.

10
The work done by Adorno, et al., in The Authoritarian

Personality not only has been enlightening in itself but also
has pioneered a very important and stimulating research area.
The major hypothesis of Adorno's study was that "the political,
economic and social convictions of an individual often form a
broad and coherent pattern as if bound together by a mentality
or spirit which is an expression of deep-lying trends in his
personality." (1) One of the most crucial results of the study
was the finding that the individual shows strong consistency in
his approach and outlook in a great variety of life situations.
"Thus a basically hierarchical authoritarian, exploitive parent-
child relationship is apt to carry over into a power-oriented,
exploitively dependent attitude toward one's sex partner, one's
God, conventionality, rigidity, repressive denial; and the
ensuing breakthrough of one's weakness, fear and dependency,
and the culmination in a political philosophy and social out-
look which contains only a desperate clinging to what appears
to be a strong and disdainful rejection of whatever is relegated
to the bottom, are all offshoots of this authoritarian up-
bringing." (1) A broad, general statement such as this appears
to imply two important points: one, that the parent-child
relationship has profound effects upon the formation of the
child's personality, needs, and ideology; and two, that the
child of an authoritarian upbringing may be a stronghold of
authoritarianism in his relations with other people.

Some attempt has been made to assess the attitudes and

ll

behavior of parents whose children score high on authoritari-
anism - ethnocentrism. The rationale of a study by Lyle and
Levitt (9) lies in a statement made by Frenkel-Brunswik (5):
parents of high ethnocentric-authoritarian children tend to
"use more harsh and rigid forms of discipline" than parents of
children less high on ethnocentrism-authoritarianism. The

Lyle and Levitt study investigated the interrelationships

among 1: ethnocentrism - authoritarianism in children, 2: the
child's willingness to be punitive, and 3: the punitiveness of
the parents as the child saw it. The subjects were fifth grade
children composed of two groups of 58 and 157 individuals. The
measures used for each of the above variables, respectively, were
the CADS (a children's form of the California F Scale, which was
used as an indirect measure of authoritarianism), PST (a
multiple choice situations test), and an incomplete sentence
test. The data supported all three hypotheses: parental puni-

tiveness had a positive correlation (Pearson r — .34, signif-
icant at .01) with the child's punitiveness; ethnocentrism -
authoritarianism in children had a positive correlation with
parental punitiveness as the child saw it and also with the
child's willingness to be punitive. This last finding was
qualified to situations where the child could express puni-
tiveness without fear of retaliation.

It is interesting to speculate as to the possible author-

itarian rating of the parents of the high authoritarian

children. Since no such measure was made, these points will

12

have to be inferred from the previous theory. It seems
probable, however, that these parents would themselves score
high on authoritarianism for two reasons; first, although "puni-
tiveness" is not spelled out in Adorno’s description of the
authoritarian parent-child relationship, it would logically
seem that such behavior would produce the rigid, emotionally
cold relationships described. And second, although authori-
tarianism in an individual possibly can develop from sources
other than an authoritarian up-bringing, authoritarianism in
fifth grade children is most likely a result of the parent-
child relationship, since the child has not yet had much oppor-
tunity to develop independently of his parents' attitudes and
behavior.

Singer and Feshbach (12) carried out a study which gives
us further insight into the ramifications of the authoritarian
personality. A sample of 147 male college students was given
a questionnaire composed of the Taylor Manifest Anxiety Scale,
the California F Scale, and the Child-Waterhouse Reactions to
Frustration Scale, in an attempt to discover the interrela-
tionships among these three indices. A significant positive
correlation (Pearson) was found between the F Scale and Taylor
Scale, the correlation being accounted for primarily by the
relation between the high scores on each of the two scales.
The subjects with middle and low F Scale scores were almost
identical on the Taylor Scale.

Again one may speculate on possible interpretations and

13

hypotheses from the study. Here is a positive relationship
between authoritarianism and manifest anxiety, which suggests
that perhaps high anxiety is a function of the authoritarian
character. This seems more plausible when viewed in the light
of the non-existent relationship between the middle and low
scores of the F Scale and Taylor Scale. Perhaps another aspect
of an authoritarian character structure is, then, a great deal
of anxiety. Such an inference would seem logical both from the
results of this study and also from the general description
made by Adorno of the "weakness, fear, and dependency..." in an
authoritarian personality.

As stated previously the study of the "Authoritarian
Personality" by Adorno, et al., has proved a valuable and
fruitful area of research. However, there have been some severe
criticisms of the study's measuring devices and implications.
First, the California F Scale elicits a response set; individ-
uals may score high on the scale simply because they tend to
agree with any statement put to them. And second, the investi-
gators are theorizing about one kind of authoritarianism, not
authoritarianism per se, but predisposition toward fascism.

Such criticisms have been levelled, in the main part, by
Rokeach (10) who advocated a departure from the study of right
authoritarianism to the study of general authoritarianism. In
making this transition Rokeach focused on the study of belief
systems as a whole, rather than on belief content alone, and

evolved his theory of the organization of belief-disbelief

14
systems. Briefly, this theory states that all the expressed

and inferred beliefs of an individual, that he accepts at a
given time as true of the world in which he lives, are organized
into an interrelated system. The dimensions of this organi-
zation are belief-disbelief, central-peripheral, and time-
perspective. The system is open or closed with respect to its
organization along the continua of these three dimensions.
Thus the individual accepts or rejects peOple, ideas, and
authority, according to how well they fit into his existing
belief-disbelief system. The instrument used to measure the
individual's position in terms of Openness or closedness of
belief system is the Dogmatism Scale (a scoring form applied
to the Psychology Questionnaire).

Rokeach's theory has stimulated a great deal of thought
and research; however, for the research in this paper, only
the study by Rokeach and Kemp (10, ch. 19) is relevant. They
suggested that "to the extent that a belief-disbelief system
is closed, it represents a cognitive network of defenses against
anxiety. This leads (us) to the simple hypothesis that those
with relatively closed systems should manifest more anxiety
than those with relatively Open systems." An additional hypoth-
esis was "that the differences in openness-closedness and
anxiety should be traceable to differences in childhood experi-
ences.” (10, pp. 347-348)

Various groups in the United States and England were

tested using the Dogmatism Scale and an anxiety scale composed

15
of 30 items from the MMPI, many of which overlap with the items

on Taylor's Scale. Information about childhood experiences was
obtained by getting the subjects' answers to several questions
concerning their attitudes "toward father and mother, anxiety
symptoms they had manifested in childhood, and the nature of
the identifications they had formed in childhood." (10, p.337)

A correlation ranging from .36 to .64 was found between
anxiety and dogmatism scores, and a factor analysis of the data
showed that dogmatism and anxiety were similar factorially.
The data on childhood experiences was consistent, for the main
part, with Frenkel-Brunswik's idea that "the inability to
express emotional ambivalence toward parents predisposes one
to form an authoritarian outlook on life." (10, p. 359)

From the various theories and studies cited, it can be
postulated that the parent-child relationship of the dogmatic
individual is likely to be or have been a threatening one in
which the parents (most likely also dogmatic) were emotionally
cold and punitive toward the child; and the child did not
express his ambivalent feelings toward them. From Rokeach's
theory it was seen that the dogmatic individual (exemplifying
closed-mindedness) is the one who clings tenaciously to his
belief system which he has develOped to protect himself from
the world, which he perceives as highly threatening. He most
often immediately rejects the intrusion of beliefs contradic-
tory to his own, and rejects or accepts other individuals

according to their agreement or disagreement with all authority

16

which he views as absolute. The dogma of this individual acts
similarly to the network of psychoanalytic defense mechanisms;
and he is perceived as threatened, defensive, and, most impor-
tantly, extremely anxious.

Such anxiety seems to be highly similar to the anxiety
Klopfer refers to in his interpretive hypothesis of Rorschach
shading (K and KF) responses, which is "of a diffuse and free
floating nature, reflecting a frustration of affectional satis-
faction." (7) In view of this it would seem reasonable to
make the following hypotheses:

1. A dogmatic individual will give more K or Y shading
responses to the Rorschach than will a non-dogmatic individual.

2. A dogmatic individual will score higher on a manifest
anxiety scale than will a non-dogmatic individual.

3. An individual high in anxiety will give more K or Y
shading responses to the Rorschach than will an individual low
in anxiety.

4. A dogmatic individual will have parents score higher on
dogmatism than will a non-dogmatic individual's parents.

5. A high-anxiety individual will have parents score
higher on dogmatism than will a low-anxiety individual's parents.

6. A high-anxiety, dogmatic individual will give more K or
Y shading reSponses to the Rorschach than will a low-anxiety,

non-dogmatic individual.

Procedure

To test these hypotheses a sample of 62 college freshman
males from introductory psychology classes were used as
subjects. Verbal announcements were posted asking for research
volunteers; in return for participating, the subjects were given
extra credit toward their psychology class grades.

The subjects were individually given the Rorschach Test,
and in several small groups, the Taylor Manifest Anxiety Scale
and the Psychology Questionnaire. The last instrument is
composed of 76 statements with which the subjects indicates
degree of agreement or disagreement along a six point scale.
The test was scored for dogmatism, with the intention of
measuring the individual differences in Openness or closedness
of belief systems.

In addition to the above measurements, the investigator
asked for the co-operation of the subjects' parents. Each
subject was given a c0py of the Psychology Questionnaire and
answer sheets to send to his parents. The parents were asked,
by letter, to complete the test and return the materials to
the investigator by mail. The letter read as follows:

Dear Parents:

I am a graduate student in Psychology at Michigan
State University, presently collecting data for my
thesis research. Your son is one of the freshman
men who has volunteered to serve as a subject for
this research.

Enclosed is a c0py of a Psychology Questionnaire,
which your son has already completed. I would

greatly appreciate it if you would also complete the

17

18

questionnaire, as I am interested in the correlation
between son's and parents' opinions. If you are
willing to participate, the enclosed directions and
answer sheet are self-explanatory. All of the testing
is anonymous since I know you and your son by number
only. I would be grateful if you would return this
folder as soon as possible; I have enclosed a stamp
and return address label.
Thank you very much for your time and co-Operation.
Sincerely,
As stated in the letter to the parents, all testing was done
anonymously with the subjects identified only by number and
each set of parents given numbers corresponding to that Of
their son.

The rationale for the use of the Taylor Manifest Anxiety
Scale may be questioned in light of the results reported by
Waller.(l4) However, it is believed by the writer that Waller's
data proved non-significant because she combined all Rorschach
shading responses (Y, T, V) for a total score and also used a
psychiatric sample. In the present study, the results were
expected to Show not only a positive correlation between
dogmatism and manifest anxiety, such as reported by Rokeach
and Kemp (10, ch. 19), but also a greater number of Rorschach
Y and YF responses from the individuals high on dogmatism than
from the non-dogmatic individuals.

The Rorschach was administered by the investigator and was
scored according to the Beck system, using Y and YF to

correspond to Klopfer's K and KF responses. The protocols were

scored by the investigator and independently by another graduate

18

19
student in clinical psychology. A total of 94 per cent

agreement was obtained on the Y-YF scoring of the 47 protocols.

Results

Individual Rorschach tests were administered to 47 of the
62 students who participated in group testing. Of the
remaining 15 students, 12 did not appear for the scheduled
appointments, and three gave very lengthy protocols which could
not be completed. 0f the 62 sets of parents asked to partic-
ipate in the research project, 50 pairs (80.6%) responded. In
addition, eight unpaired parents (two fathers, six mothers)
filled out the questionnaire. Thus, a total of 87 per cent of
the parents participated in the study.

Scores on the Taylor Manifest Anxiety Scale ranged from
three to 36 with a median Of 14.5. The median for the sample
is slightly higher than the median of Taylor's standardization
sample, which has a median of 13. (13, p. 285) Subjects
scoring 14 and below were classed as the "low anxiety" group,
and those with scores of 15 and above were called the "high
anxiety" group. College students' scores on the Dogmatism
Scale ranged from 117 to 219; the median score was 155.5 with
those below classed as "low dogmatic" and those above as "high
dogmatic" individuals. The frequency distribution of the
parents’ dogmatism scores ranged from 94 to 208 with a median
of 158.5. Only those cases in which both parents responded
were used in determining the median. The scores below the
median were called "low dogmatic" and those above, "high

dogmatic." The dogmatism scores in this study were comparable

2O

21
to those obtained by Rokeach.*

The data for the six hypotheses were analyzed by chi-
squares. The results show no significant relationship between
shading responses to the Rorschach and any of the other measures
used in this study. Neither the high dogmatic nor the high
anxiety group gave more Y-YF responses to the Rorschach than
did the low dogmatic or the low anxiety group. (See Tables 1
and 2) Also, the number of Rorschach shading responses from
individuals high in both dogmatism and anxiety did not differ
significantly from the number of shading responses from indi—
viduals low in both anxiety and dogmatism. (See Table 3)
Hypotheses one, three, and six were thus not supported.

Table 1. Relationship between subjects' shading responses

and dogmatism scores.

Subjects' shading responses

 

 

Y No Y
Subjects' High 14 ll
dogmatism
scores
Low 10 12

 

 

 

 

chi-square = < 1

* personal communication

22

Table 2. Relationship between subjects' shading responses
and anxiety scores.

Subjects' shading responses

 

 

Y No Y
Subjects' High 13 11
anxiety
scores
Low 11 12

 

 

 

 

chi-square = <1

Table 3. Relationship between subjects' shading responses
and anxiety - dogmatism scores.

Subjects' shading responses

 

 

 

 

 

Y No Y
Both 8 8
high
Subjects’ -
anxiety - Mixed ll 6
dogmatism
scores
Both 5 9
low

 

 

chi-square = 2.59, significant
between .50 and .75 levels

A relationship at the .025 level of significance was
found between dogmatism and manifest anxiety, confirming the
second hypothesis that dogmatism and anxiety are related
concepts. (See Table 4) In addition, the data on anxiety and
dogmatism were analyzed by the correlation coefficient. The
relationship between subjects' anxiety and dogmatism was
significant, with an r value of .725. Subjects high in

dogmatism tended to be high in anxiety, and those low in

23

dogmatism were also low in anxiety.

Table 4. Relationship between subjects‘ dogmatism and
anxiety scores.

Subjects’ dogmatism scores

 

 

High Low
Subjects' High 20 ll
anxiety
scores
Low 11 20

 

 

 

 

chi-square = 5.04, significant
at .975 level
Parental dogmatism was not significantly related to any

measure in the study. No significant relationship was found
between dogmatism scores of the parents and of the college
students. (See Table 5) Also, there was no significant
relation between the parents' dogmatism scores and the students'
manifest anxiety scores. (See Table 6) No significant corre-
lation was found between fathers' and subjects' dogmatism
scores or between fathers' dogmatism and subjects' manifest
anxiety scores. While both these correlations were non-signif-
icant, the r values were negative (and in the direction Opposite
that predicted). In addition, the correlations between the
mothers' dogmatism scores and both subjects’ anxiety and
dogmatism scores were non-significant. These results give no

support for hypotheses four and five.

24

Table 5. Relationship between subjects' and parents'
dogmatism scores.

Subjects' dogmatism scores

 

 

 

 

 

 

High Low
Both 8 7
high
Parents'
dogmatism Mixed 9 11
scores
Both 6 9
low

 

chi-square = < 1

Table 6. Relationship between subjects' anxiety and
parents' dogmatism scores.

Subjects' anxiety scores

 

 

 

High Low
T
Both 9 6
high
Parents'
dogmatism Mixed 9 11
scores
Both 7 8
low

 

 

 

 

chi-square = 1.7, significant
between .50 and .75 levels

The study therefore shows that in freshman males, manifest
anxiety and dogmatism are related concepts. However, the results
do not indicate that shading responses to the Rorschach are
related to either dogmatism or to anxiety as measured by the

Psychology Questionnaire and the Taylor Manifest Anxiety Scale.

Discussion

In examining the non-significant relationships between the
Rorschach shading responses and the anxiety measures, it is
apparent that this study offers no evidence to conclude that
shading or Y-YF responses to the Rorschach indicate anxiety.
Thus, the validity of interpreting Rorschach shading responses
as anxiety indices remains in question.

The results demonstrate a significant relationship between
the scores of the Taylor Manifest Anxiety Scale and the Psychol-
ogy Questionnaire, or Dogmatism Scale, Both instruments are
either measuring the same variable or measuring two different
variables which have a high correlation. We may refer to these
as anxiety measures or as anxiety and dogmatism. Therefore, it
is possible to infer that an anxious individual becomes
dogmatic for defensive reasons, or that "Taylor anxiety" and
dogmatism are both types of anxiety of similar levels and inten-
sity. According to the results of this study, such anxiety is
not reflected in shading responses to the Rorschach.

It is necessary to qualify acceptance of the above
conclusions, as the general concept of anxiety is still a rather
hazy one. We have specified a variable, measured it by two
different instruments having a high correlation, and called the
variable anxiety. It is possible that anxiety is a multidi-
mensional concept, and that shading responses to the Rorschach
represent a dimension of anxiety different from the others used
in this study. However, if, in fact, shading responses reflect

25

26

some kind of anxiety, one would expect a minimal but signif-
icant relationship between the different dimensions of anxiety.

The investigation of relationships between parent and
child performance on these two scales provides some interesting
data on the develOpment of attitudes and emotional states in
children. It was hypothesized that dogmatic parents would
produce dogmatic, and therefore anxious, children. This is not
the case, at least with regard to freshman males. We are left
with the conclusion that no simple correspondence exists between
the attitudes of parents and children. A child does not seem to
become dogmatic and/or anxious because his parents are dogmatic.
Considering the age of the subjects in this study, it is
possible that they are at a stage of pre-adult rebellion. Thus,
one would expect them to respond contrary to if not independ-
ently of their parents' attitudes and behavior. The presence
of anxiety and dogmatism in these individuals must then be a
function of factors other than dogmatic parents.

The drawbacks of the Taylor Manifest Anxiety Scale have
been previously mentioned and discussed. The scale presents
the problems of any paper-and-pencil test, viz., subject's
awareness and truthfulness, test reliability and validity.
However, the Taylor seems to be the best of its kind for use by
clinicians. The rationale for use of the Psychology
Questionnaire as an additional measure of anxiety is best
supported by the significant correlation of the dogmatism scores

with the Taylor scores. Again we have the limitations set by

27

the use of an objective test. In addition, since the parents'
dogmatism scores do not have a positive correlation with any of
the other data, it is possible that their scores are not valid,
i.e., the parents may not have independently made responses to
the test. While the subjects themselves may have been a repre-
sentative sample of freshman males, the results of the study
perhaps may not be permitted generalization to non-freshman
populations.

However, the procedure and logic of this research do not
seem to have faults sufficient enough to warrant rejection of
the results. In spite of the evidence being contrary to
clinical expectations, we are left with the conclusion that
shading responses to the Rorschach are not indicators of

anxiety.

Summary

Shading or Y-YF responses to the Rorschach Test are gener-
ally considered indices Of anxiety within the context of the
Rorschach protocol. This study was designed to test the valid-
ity of interpreting Y-YF reSponses as indicators of anxiety.

It was hypothesized that individuals high in anxiety would
give more Y-YF responses to the Rorschach than would individ-
uals low in anxiety.

Sixty-two freshman males were group tested on the Taylor
Manifest Anxiety Scale and the Psychology Questionnaire, or
Dogmatism Scale. The latter instrument was used as an
additional measure of anxiety, as it was further hypothesized
that those individuals high in dogmatism would also be highly
anxious. This relationship was confirmed by a significant
chi-square and a significant correlation between the measures
of dogmatism and manifest anxiety.

Individual Rorschachs were administered to 47 of the
students who had participated in group testing. It was found
that those individuals high in either or both Taylor anxiety
and dogmatism did not give significantly more Y-YF responses to
the Rorschach.

The study also postulated that freshman males high in
dogmatism and anxiety would have parents scoring high in
dogmatism. The parents of the subjects were administered the
Psychology Questionnaire, by mail, and 50 pairs responded with
completed tests. NO significant relationship was found between

28

29

parental dogmatism and subjects‘ dogmatism, nor between parental
dogmatism and subjects' manifest anxiety.

The results of this study then show that pre-adult males
are neither anxious nor dogmatic because of parental dogmatism.
Those individuals highly dogmatic are also highly anxious, but
they do not indicate these states by giving prOportionately
more Y-YF responses to the Rorschach. Therefore, it is believed
that Y-YF or shading responses to the Rorschach are not indices

of anxiety within the Rorschach protocol.

10.

ll.

12.

13.

BIBLIOGRAPHY

Adorno, T.W., Frenkel-Brunswik, Else, Levinson, D.J., and

Sanford, R.N., The Authoritarian Personality. Harper, New
York, 1950, 1-27, 971-976.

Allerhand, Melvin E., Chiaroscuro determinant of Rorschach
Test as indicator of manifest anxiety. JI Projl Tech.,
1954, 18, 407-413.

Beck, Samuel J., Rorschach's Test I, Basic Processes.
Grune and Stratton, New York, 1950.

Eichler, Robert M., Experimental stress and alleged Rorschach
indices of anxiety. J.,Abn- Soc. Psychol., 1951, 46, 344-
355.

Frenkel-Brunswik, Else, A study of prejudice in children.

Human Relations, 1948, 1, 295-306.

Hertz, Marguerite R., Shading response in the Rorschach Ink
Blot Test; review of scoring and interpretation. J, Gen,
Psychol., 1940, 23, 123-167.

Klopfer, Bruno, DevelODments in the Rorschach Technique:

Vol, 1 Technique and Theory. World Book Co., Yonkers-on-
Hudson, New York, 1954.

Levitt, Eugene E., Alleged Rorschach anxiety indices in
children. J, Proj, Tech., 1957, 21, 261-264.

Lyle, William H., Jr., and Levitt, Eugene E., Punitiveness,
authoritarianism, and parental discipline of grade school
children. b o P c 01., 1955, 51, 42-46.

Rokeach, Milton, The Open and Closed Mind. Basic Books,
Inc., New York, 1960.

Sarason, Seymour 8., Davidson, Kenneth, Lighthall,
Frederick, and Waite, Richard, Rorschach behavior and
performance of high and low anxious children. Child Devel-
opment, 1958, 2 , 277-285.

Singer, Robert D., and Feshbach, Seymour, Some relationships
between manifest anxiety, authoritarian tendencies, and
modes of reaction to frustration. J, AbnI Soc, Psychol.,
1959 (Nov), 53, 404-408.

Taylor, Janet, A personality scale of manifest anxiety.

J, Abn, Soc, Psychol., 1953, 38, 2235-290.

30

14.

31

Waller, Patricia F., The relationship between the Rorschach
shading response and other indices of anxiety. J, Proj,
Tech., 1960, 24, 211-216.

APPENDIX

Subjects' Rorschach responses, dogmatism and manifest anxiety

scores 0

Subject
Number

10.
11.
12.
l3.
l4.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
'41.

Taylor
Manifest
Anxiety
Score

31
l5
l8
8
3
l9
5
3O
21
18
9
16
9

~26

7
36

5
3O
11

6
11
27
18
15
10
17
25
14
22
22
12
36
32
2O
36
22
18
12

8
21
3O

Dogmatism
Score

164
125
167
155
129
153
145
153
168
155
142
145
131
167
154
149
159
196
160
144
179
145
155
155
141
161
179
131
151
168
137
158
160
160
186
184
156
142
152
219
180

32

Parents'
Dogmatism Score
F M
115 124
175 185
198 158
134 165
147 146
177 183
122 152
126 174
156 179
178 114
159 164
163 158
153 122
200 94
146 175
186 197
170 183
120 169
155 148
130 143
155 139
156 -
135 143
- 144
189 198
120 117
175 167
140 171
153 174
100 123
188 208
129 137
161 131
- 200
145 140
191 186
102 178
159 178

Number of
Shading (Y-YF)
Responses to
Rorschach

Il-‘ll—‘Ml—‘O

IMbI

I0 I MMOl-JOOMIO I OOIHOt—‘OOOOHOOOO-b-Ol

Subject
Number

42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.

Taylor
Manifest
Anxiety
Score

35
12

9
18
10
11

7
3O
11
18

7
14
13
14

4
12
16
14

4
14

3

33

Dogmatism
Score

175
117
157
145
179
157
142
195
126
160
158
167
159
151
145
146
159
162
151
189
151

Parents'

Dogmatism Score
M

F

165

172
156
159
152
177
133

160
148
177
193
138
149
166
134
171

197
147

121 A

192
164
139
170
173

96
158

162
158
183
192
137
173
128
172
131

Number Of
Shading (Y-YF)
Responses to
Rorschach

l O

OOOI—‘Oi—‘OOM I (OHMOOHOOI

ROOM USE ONLY

 

 

 

F-

MICHIGAN STATE UNIVERSITY LIBRARIES

ll Hllllillll I

II

3 1293 03177 9170