Ill

IIIHMHIIIIHI

1|

1’

H

 

333 Ill
(DA -P M

“4 Hz: ’3. A. {2. a ’24 Elli??? 4 HKT'LSH S???

m' g-
i. 1185.

C'IGAH 3T6": "

Hisi‘w .. as

Irrfl
6‘51!

if
V

‘3

‘Lmn

3

:4 fear Hm Dagsas 4}? 5414 A4

¢H§4ERS§TY

,. Eerraié’i

THESK;

LIBRARY

Michigan State
University

 

ABSTRACT

A D.A.P.: A VALIDATION STUDY

by Nicholas J. Borrelli

The present study attempted to validate the clinical usefulness
of the D.A.P. Test. It was hypothesized that the agreement between
measures of personality characteristics as obtained through self-
description and the D.A.P. is a function of psychopathology. Both
measures of personality characteristics were obtained through the use
of the Corsini Standard Adjective Q Sort. The first measure was ob-
tained by asking §_to sort according to how he saw himself; the second
through judges' sorts based upon analysis of §:§_drawings.

A comparison of the average self-judge correlations of 20
orthopedic patients (alledgedly more accurate group) to those of 20
ulcer patients (alledgedly less accurate group) failed to reveal a
difference significant at the .05 level. It was concluded that no
support cou1d be offered for the clinical usefulness of the D.A.P. in
revealing §:§_personality characteristics.

Three difficulties with the present study were discussed, the
first dealing with choice of §§_and the second and third with limita-

tions of the Q-sort technique. In light of these difficulties some

Nicholas J. Borrelli
suggestions were made for future research which might be involved in

testing hypotheses similar to that of the present study.

Approved (lfikzy4//<?{fl[{:1 Date /%Qég?/2 Z

Mafir Wrofessoy

THE D.A.P.: A VALIDATION STUDY

BY
9,
.IA‘”

Nicholas J.”Dorrelli

A THESIS

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

MASTER OF ARTS
Department of Psychology

1963

p 0' U",
11;}. /.‘; "J
I
A

4 l4—

Rx

ACKNOWLEDGEMENTS

The author wishes to express his appreciation to his Committee
Chairman, Dr. Joseph Reyher, for his invaluable assistance in the
planning of this research and the preparation of the manuscript.

A special thanks too to Mssrs. Frank Kirby and Leonard Handler

who served as judges in the study.

ii

TABLE OF CONTENTS

Page

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . i 1
METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Subjects . . . . . . . . . . . . . . . . . . . . . . . . . 2
Materials . . . . . . . . . . . . . . . . . . . . . . . . . 2
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . 3
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

REFERENCES 0 O o O o I 0 0 0 O O O O O O O O O O O O O O O O O O I 25

iii

LIST OF APPENDICES

Appendix Page
A. Background Research . . . . . . . . . . . . . . . . . . . . 8
B. The Standard Adjective Q-Sort . . . . . . . . . . . . . . . 20

C. Table I. Means and S.D.'s for Ulcer and Orthopedic
Groups on Age, Education and I.Q. . . . . . . . . . . . . 23

D. Table II. Inter-judge Comparisons on Average
Q-Sort Ratings for Si . . . . . . . . . . . . . . . . . . 24

iv

Machover maintains that important aspects of personality are
revealed in human figure drawings (1949). Numerous studies have ex-
plored avenues suggested by Machover's theoretical position with par-
ticular attention being focused on her contentions regarding the rela-
tionships between qualitative features of drawings and specific per-
sonality characteristics of §§: The results of these studies have
rarely substantiated her claims (Swenson, 1957). While the techniques
employed in these studies have varied, researchers seem to have over-
looked the possibility of using Sig introspective reports in their
investigations. This approach, however, has been used productively
with other projective devices (wylie, 1961). Friedman (1955), for
example, attempted to evaluate the accuracy of §:§_phenomenologically
reported self-concept. The rationale used was that the greater the
discrepancy between §:§_self-estimate and §:§_projected self qualities
(i.e. characteristics of the heroes in §l§_T.A.T. stories), the less
accurate was §:§_self-appraisal.

The present study proposed to test Machover's claim regarding
the clinical usefulness of the D.A.P. by using a technique similar to
that employed in the Friedman study. If the above claim is valid,
then where §_is inaccurate in his self appraisal, no necessary simi-
larity would exist between §:§_self appraisal and a measure of §l§_
personality characteristics as obtained through analysis of his drawings.
Where §_is more accurate in his self appraisal, a greater similarity
would exist between these two measures of personality characteristics.
A comparison of accurate and inaccurate groups in self appraisal was

1

2
expected to reveal differences in terms of the degree of correspondence
between measures of personality characteristics as obtained through

self-report and through analysis of S's drawings.
Method

Subjects.--According to psychoanalytic theory, peptic ulcers
may be understood as the end result of repressed and chronically
frustrated oral-receptive, demanding attitudes (Fenichel, 1945). On
this basis it was assumed that peptic ulcer patients would be suffering
from.more severe and more inaccesible conflicts than orthopedic, but
otherwise normal, patients. The former group was thereby thought to
be less accurate in their self-appraisal than the latter group.

In accordance with the above rationale, the sample selected con-
sisted of 40 hospitalized male patients: 20 peptic ulcer patients
and 20 orthopedic patients. The former group consisted entirely of
patients with the diagnosis of peptic ulcers. The latter group con-
tained primarily patients with broken limbs, although the sample in-
cluded two patients with back complaints. Both groups consisted of
patients hospitalized at V.A.H. Dearborn, and were equated as to age,
education and I.Q. (This information is presented in tabular form in

Appendix C).

Materials.--The materials used consisted of the Standard Adjective
Q Sort (SAQS) technique (See Appendix B), an 8-1/2 by 11-1/2 unlined
pad and a number two pencil. The latter materials were provided for
use on the D.A.P. test. All testing took place in §:§_ward physiciads

office with only §_and §_present.

3

Procedure.--In the first phase of the investigation Si§_coopera-
tion was gained with the explanation that S_wished to assess the value
of an instrument with which S himself would have an opportunity to be-
come acquainted later. S was then given the D.A.P. with instructions
according to Machover (1949). Upon completion of this task, S was
introduced to the Q sort procedure with the instructions provided by
Corsini (1956) and asked to sort according to how he saw himself.

In the second phase of the study each Si§_figure drawings were
analyzed by two independent judges. On the basis of these individual
analyses, and without knowledge of S:§_self-sort, the same 50 words
were sorted by the judges in the same manner in which S had previously
sorted these words for his self description. The final measure of SL3
personality characteristics from drawing analyses was based upon the

sortings of both judges.
Results

The data which were subjected to statistical analysis consisted
of three rectangular distributions of 50 words for each S, These Q-
sort distributions represented SL5 self descriptive sorting and the
two judges' sortings based upon their analyses of Sl§_drawings.
Pearsonian correlations (Corsini, 1956) were computed to determine the
degree of relationship between S:§_self sort and the two judge-sort
distributions for each S, The reliability estimate obtained by com-
puting the average inter-judge correlation was .54. No significant
differences were found between judges in terms of their average ratings
for the experimental groups. (Refer to Appendix D for a tabular pre-

sentation of this data.)

4
In order to test whether group differences with regard to these
correlations were significant, the self-judge correlations were trans-
formed into 2 scores and the mean z score for the self-judge correla-
tions of the two groups compared (Walker and Lev, 1953). The obtained
mean 2 score of -.03 for the ulcer group as compared to the mean 2
score -.10 for the orthopedic group failed to reveal a difference

significant at the .05 level.

Discussion

 

The hypothesis that the agreement between §L§ self-description
of personality and judges' description based upon drawing analysis
would be greater for the orthopedic group (more accurate S§_than for
the ulcer group (less accurate §§) was not substantiated. In view of
this result, no support can be offered for the clinical usefulness of
the D.A.P. in revealing S:§_personality characteristics. However, at
least three difficulties with the present study must be mentioned as
they bear upon the study's negative conclusion.

First, contrasting the orthopedic group to the ulcer group on
the assumption that the former group was more accurate in self-appraisal
than the latter, was perhaps not the wisest choice of comparison groups
possible. In retrospect, it seems likely that the orthopedic's physical
disability may well have activated conflicts about body image and body
functioning which were as severe and hence as inaccessible as those of
the ulcer group. Moreover, the dependent position of both types of R3
in the hospital situation would more likely make them become less
sincere about describing themselves than usual. Hence, it is con-

ceivable that the two groups would not differ as substantially on

5
accuracy of self-appraisal as originally assumed.

Second, in addition to the questionable reliability of the
S.A.Q.S. for use on the D.A.P., the task of Q-sorting itself imposed a
somewhat artificial character upon the judges' representation of Sig“
personality structure. It required that the judges fragment the data
obtained through drawing analysis to fit the descriptive adjectives
of the Q-sort. Such compartmentalization is neither the way personality
is organized, nor is it the manner in which clinicians usually organize
information obtained from the D.A.P. Generally, clinicians seem to
organize this data into integrated personality descriptions or global
judgments about R12 personality, e.g. neurotic versus psychotic adjust-
ment etc. This type of an approach to personality description, how-
ever, differed markedly from the judgemental task required in the present
study wherein judges were asked to fragment meaningfully interrelated
D.A.P. data, which data otherwise is usually integrated by clinicians.
Conceivably, such a departure from the usual and probably better
approach, may have contributed considerably to both the meaningfulness
and the accuracy of the judges' task in the present study.

Finally, mention should be made of one particular characteristic
of the S.A.Q.S. as it bears upon the obtained inter-judge agreement.
The S.A.Q.S. words tend to be phenotypic descriptions of personality
features. In the event that one judge used them in their phenotypical
sense, whereas the remaining judge tended to think about the genotypic
implications of the words, this difference in conceptualization may
well have contributed to low inter-judge reliability.

Future research aimed at testing hypotheses similar to that of

the present study might do well to keep the above difficulties in view.

6
Where possible, clinical groups which differ more markedly with regard
to severity of psychological impairment should be compared. Such a
measure would make it more probable that the groups would differ in
accuracy of self appraisal. Then too, where a judgemental task is
required, it would seem advisable to take an approach more congruent
with the way clinicians organize their impressions about personality.
One possible means to overcome both aforementioned difficulties is to
compare interview and D.A.P. impressions from the same S§_with the same
clinician rendering both impressions according to some prearranged
scheme or format. The design of the research would be set up in such
a way that the clinician would not know which interview and D.A.P.
protocals belonged to the same S, According to Reyher (1961), this

procedure substantially improves reliability and validity coefficients.

Summary

The present study attempted to validate the clinical usefulness
of the D.A.P. Test. It was hypothesized that the agreement between
measures of personality characteristics as obtained through self-
description and the D.A.P. is a function of psychopathology. Both
measures of personality characteristics were obtained through the use
of the Corsini Standard Adjective Q Sort. The first measure was ob-
tained by asking S_to sort according to how he saw himself; the second
through judges' sorts based upon analysis of Sl§_drawings.

A comparison of the average self-judge correlations of 20
orthopedic patients (alledgedly more accurate group) to those of 20
ulcer patients (alledgedly less accurate group) failed to reveal a

difference significant at the .05 level. It was concluded that no

r . . - . .

 

7
support could be offered for the clinical usefulness of the D.A.P. in
revealing S:§_personality characteristics.

Three difficulties with the present study were discussed, the
first dealing with choice of S§_and the second and third with limita-
tions of the Q-sort technique. In light of these difficulties some
suggestions were made for future research which might be involved in

testing hypotheses similar to that of the present study.

APPENDIX A

Background Research

 

In the 14 years since the publication of Machover's monograph,
many studies have aimed at validating Machover's interpretive hypothesis
about drawing features in the D.A.P. Swenson (1957), after an ex-
haustive review of the literature until 1956 concluded, however, that
while there was modest confirmation of some of Machover's hypotheses
regarding group trends, there was little evidence for the value of the
instrument for individual diagnosis. He did concede, however, that the
D.A.P. had some usefulness as a gross indicator of adjustment.

More recently, many other studies have been undertaken whose
intent has been to evaluate the clinical usefulness of this instrument.

The bulk of these are reviewed below.
Research Related to the Body-Image Hypothesis

The basic assumption underlying D.A.P. interpretation is that
when asked to draw a picture of a person, S will respond by drawing a
picture which symbollically represents himself. This is called the
body image hypothesis.

Kamano (1960) sought to investigate the body image hypothesis
by determining whether human figure drawings represent the drawer's
perception of himself, an idealized figure or an unfavorable one by
using the semantic differential technique. To do this 45 hospitalized

8

9
schizophrenic females drew a same-sexed figure. They then rated the
drawing together with the concepts ideal-self, actual-self and least-
1iked-self which were presented in counterbalanced order on a set of
15 bipolar scales. The results showed that S§_tended to draw a figure
which represented themselves much more than an ideal or unfavorable
one. This was interpreted as supporting the hypothesis that human
figure drawings do represent the drawer's perception of himself.

Lakin (1956) reasoned that on the basis of observations and
theory regarding human figure drawings and their relations with self-
concept and body image, certain formal characteristics of drawings
would reflect aspects of change in self-concept and body image as a
result of age. The drawing characteristics studied were area, height
and centeredness. Comparing the drawings of 24 institutionalized
elderly persons with those of 25 third-graders confirmed the hypothesis.
The aged S§_produced more constricted, shortened and less adequately
centered figures than the third graders. The finding was viewed as
supporting the validity of the view that the central variables of self-
concept and body image are reflected in human figure drawings.

McGuire and Moss (1962) asked a group of 50 S§_containing normals,
deaf S§_and clinicians, to match the cartoons and captions of William
Steig. The importance of this task, the authors point out, was that
Machover had singled out these drawings as highly meaningful caricatures
of body language interpretable in terms of hypotheses applicable to the
D.A.P. The results disclosed that S§_correctly matched cartoons and
captions at a high level of significance (p = < .0001). However, the
authors note several important qualifications: (A) clinicians were not

demonstrably more capable than bright normal S§_in matching (B) statistical

10
significance was largely contributed by only 6 of 32 cartoons; and (C)
generally there was no relation between S:§_confidence in his matchings
and objective correctness.

In the Murphy study (1957), the D.A.P. was administered to 53
male and 35 female adults who were applicants for employment at a state
colony. Drawings were rated for sexual differentiation and graphic
maturity. The results showed that females differentiated female figure
drawings from male figure drawings more adequately than did males, al-
though there was little difference in graphic maturity between the two
groups. Results were interpreted as suggesting that same-sex human
figure drawings of adults tend to be self-portraits rather than an
index of degree of identification with appropriate sex role.

Wiggenhorn (1957) sought to investigate changes in human figure
drawings as a function of changes in self-concept. The D.A.P. was
administered to stress and nonstress groups before and after the stress
condition. The results indicated that the redrawn figures for the
stress group showed several changes in the drawings corresponding to
lowered self-concept as measured by personal ranking sheets, while the
reinforced group did not.

Despite some confirming evidence from the above studies, at
least two studies did not support the hypothesis that human figure
drawings are valid projections of the drawer. Silverstein and Robinson
(1956) obtained human figure drawings from 22 children in the chronic
stages of poliomyelitis on the assumption that the drawings would re-
flect the child's physical disability. The drawings were studied by
three different methods to determine whether the disability was reflected.

Although by inspection more than 3/4 of the S§_appeared to represent

11
their disability either directly or indirectly, a comparison of disabled
children's drawings with those of normal controls showed a nonsignificant
number of significant differences between the two groups on a series of
55 scoring signs. It was noted that even judges experienced in the
diagnostic use of figure drawings were unable to differentiate the
drawings of the two groups at a level better than chance.

In the Wacks and Zaks study (1960) the drawings of 30 males with
spinal cord injury were analyzed according to 24 criterian and compared
to the drawings of 30 controls, i.e. chronically ill, long term,
hospitalized male veterans comparable to the experimental group in
age, education and length of illness. While it was expected that the
physically disabled would represent their disability in their drawings,
in most respects there were no significant differences between the two
groups in body image. The authors point out however, that failure to
find pronounced differences may be a reflection of general degree of
disturbance in psychological functioning in any chronically ill patient.

These few studies suggest that the assumption regarding the

figure drawn as a valid projection of the drawer is at best equivocal.
Diagnostic Use of the D.A.P.

General.--Lubin (1959) was interested in seeing if the D.A.P.
could discriminate between stable and unstable negro psychiatric aides.
Stability was defined in terms of having remained at the job for a
period of at least one year. The first drawing of each newly hired
negro aide was scored according to a 100 item check list. A comparison
of stable and unstable aides disclosed that only in test-taking attitude

did the two groups differ unequivocally.

12

Staltz and Colthrop (1961) sought to discover whether clinicians
could predict intelligence, sociability and emotional adjustment from
figure drawings. Three clinicians rated the drawings of 60 fourth graders
on a nine point scale. The clinicians were able to predict intelligence,
but were unsuccessful in predicting sociability and emotional adjustment.

Hardy (1960), using a global approach, investigated the relation-
ship between emotional maladjustment as it is classified into diagnosis
and figure drawings. Fifteen judges representing 3 levels of figure
drawing analysis experience sorted the drawings of 90 adult males
(i.e. normals, neurotics and psychotics) both separately and as a
battery using the global approach. The following results were obtained:
(A) although all judges showed significant ability to identify drawings,
it was the correct identification of psychotics that significantly
contributed to differentiation; (B) diagnostic errors were less prevalent
when diagnostic range increased; (C) a linear trend between judging
experience and correct identification was shown; and (D) the use of a
drawing battery reduced the validity of figure drawings as a diagnostic
tool. The authors conclude that the results tend to negate the use of
figure drawings as a general tool for assigning adjustment level, but
support their value in the hands of an experienced clinician.

Sherman (1958) collected 52 sets of male and female figure
drawings from hospitalized psychiatric patients and nursing assistants,
both groups equivalent in age and I.Q. Psychologists experienced in
drawing analysis then judged the drawings as to patient or nonpatient
status while commercial artists evaluated the drawings for artistic
merit. The following conclusions were reached: (A) psychologists

were unable to distinguish between patient and nonpatient drawings;

 

 

13
(B) a significant relation was found between art quality and judged
adjustment status. Sherman advises in view of these results that
psychologists obtain some independent measure of drawing ability when
evaluating drawings.

Singer (1958) sought to answer the question of whether varia-
tions in human figure drawings provide a reliable and valid person-
ality measurement. His sample consisted of 102 male psychiatric
patients. S§_were given both the M,M.P.I. and the D.A.P. Four drawing
features were evaluated on a 3 point scale, (1) head to trunk ratio;
(2) vertical size of figure; (3) clothing versus body emphasis; and
(4) vertical placement of figure. These features were related to
specific hypotheses based on postulates by drawing analysis experts.

A point biserial correlation was computed between each M.M.P.I. item
answered "true" and the drawing features. Only head to trunk ratio

and vertical size of the figure were significantly related to variations
in M.M.P.I. items response after cross-validation. It was concluded
that these drawing features may have some meaningful possibilities in
personality measurement.

A part of the wallon study (1959) had as its purpose to test
the validity of the D.A.P. as an index of broad levels of adjustment.
Drawings were obtained from 25 normal, 25 neurotic and 25 psychotic
adult males and then given to 5 experienced clinicians to assign to
3 diagnostic groups. Four of 5 judges achieved a number of successes
which exceeded chance at or beyond the .03 level of probability;
collectively, the judges exceeded chance at the .001 level. Despite
this, however, less than half the judgements for any one judge were in

agreement with the criterion.

14

The D.A.P. and Sexual Devianqy,--Machover has advanced several
hypotheses about the significance of the first-drawn figure on the
D.A.P. and its relation to sexual adjustment. Brown and Tolor (1957),
after reviewing 28 pertinent studies came to the following conclusions:
(A) Normal adults of both sexes draw their own respective sexes first
in making human figure drawings. Adult males make the first drawing
a male more frequently than females make the first-drawn figure a female.
The average figures for drawing the same-sex figure first for normal
college students are 91% for males and 63% for females. (B) Male
and females with behavior disorders make like-sexual figures more fre-
quently than they draw figures of the opposite sex. (C) What little
evidence is available indicates that the great majority of male homo-
sexuals don't draw opposite-sex figures first; this is contrary to the
popular belief that drawing the opposite-sex first is indicative of
homosexuality. There is no information whatever concerning the rela-
tion between female homosexuality and sex of the first-drawn figure.
(D) There is much confusion in the studies of children. When these are
taken as a whole, nearly equal proportions of boys and girls draw a
figure of their own sex first. (E) Analysis of human figure drawing
characteristics, apart from the sex of the figure as such, may provide
important data about homosexuality and psychosexual identification.
The authors conclude "There is no convincing evidence that adequacy of
psychosexual identification or adjustment is reflected in choice of
sex in human figure drawings. At present the only valid conclusion is
that the basis or significance of drawing a person of the opposite-sex
first is not known."

More recently, Vilhotti (1958) too has come to a negative

15
conclusion about a popular interpretation regarding sex of the first-
drawn figure. He compared the drawings of 50 male homosexuals to 50
male non-homosexuals. All S§_were mentally defective. Vilhotti con-
cludes that the sign of drawing a female figure first as diagnostic
of male homosexuality is not useful in differentiating homosexuals
from nonhomosexuals in institutionalized mental defectives.

Fisher (1959) examined the D.A.P. test performances of 277 male
and 232 female psychiatric patients and 32 male sexual deviates for the
sex of the first-drawn figure. In both the male psychiatric sample
and male sexual deviate sample 88% of S§_drew the male figure first,
while in the female psychiatric group, 62% of the RE drew the female
figure first. The data indicated that male and female psychiatric
patients and male sexual deviates do not differ significantly from
normals in the sex of the first-drawn figure. This finding was inter-
preted as supporting the earlier conclusions of Brown and Tolor (1957)
regarding the lack of evidence to support the interpretive hypothesis
on sex of first-drawn figure.

Grams and Rinder (1958) investigated the validity of 15 signs in
figure drawing which Machover had listed as predictive of homosexuality.
Fifty adolescent inmates were divided, on the basis of homosexual
experience, into two groups matched for age, I.Q., race and schooling.
Their drawings were given to three psychologists to score for signs of
homosexuality. None of the signs were found to have individual
validity. To determine the predictive significance of the 15 signs
taken as a whole, the total number of signs present in a drawing were
correlated with homosexuality and nonhomosexual experience. Again

the signs studied failed to validly predict to criterion.

16

While the above studies have largely failed to support Machover's
position, some supportive evidence is forthcoming from the two studies
below.

Frank (1955) asked 18 male and 56 female normal college students
to draw a picture of a whole person. The first drawing of 16 males and
38 females were of a figure of the same sex as themselves. This result
was interpreted as supporting Machover's assumption that drawing a
person of the opposite sex first indicates a problem of sexual identifi-
cation.

Finally, in the Whitaker study (1961), 236 males referred to a
court clinic were rated on the characteristics homosexuality and
effaminancy by a clinician on the basis of life history interviews.
Each S was then given an extended D.A.P. in which he chose the sex of
two of the three figures drawn. (In the extended D.A.P., after the
usual procedure of having S draw two figures, the second figure being
of the opposite sex as the first, S.is then instructed to draw a third
figure where again he has the option of choosing the sex of the figure
drawn.) The results were seen as supporting the hypothesis that
psychosexual identity is projected into free choice drawings. However,
the author noted that the psychometric signs were not more efficient

overall than the base rate in predicting the criterion characteristics.

The D.A.P. and Paranoid Pathology,--Among the most prominent

 

features of a paranoid's drawings, according to Machover, is emphasis
on eyes and ears. The studies cited below deal with these variables.
The Reznikoff and Nichols study (1958) attempted to check the

usefulness of 26 drawing indicators of paranoid pathology which were

17

compiled from Machover's list of indicators and the authors' clinical
experience. Upon comparing the drawings of 30 paranoid and 30 non-
paranoid patients, they found that of the 26 indicators, only emphasis
on outline of eyes and over-all heavy line quality occurred more fre-
quently in the paranoid group. However, as this result could occur by
chance, the groups were compared for the total number of indicators
present in individual drawings. No significant differences between the
two groups were found.

Ribler (1957) selected 20 pairs of D.A.P. test protocals from a
V.A. hospital files to determine whether judged eye and/or ear emphasis
could differentiate paranoid schizophrenics from unclassified schizo-
phrenics, anxiety neurotics and normals. The protocals were submitted
to four judges independent of each other. Although inter-judge
reliability was adequate, the judges could not differentiate the drawings
of paranoids from the other groups on the basis of eye-ear emphasis.

The Griffith and Peyman study (1959) used a somewhat different
approach than usual. Its purpose was to relate eye-ear emphasis to
the trait ideas of reference. The drawings of 18 S§_were selected
which showed both eye and ear emphasis and compared to the drawings of
a control group of 51 Sg, Two clinicians were asked to select the
drawings of S§_having ideas of reference. The results confirmed the
hypothesis that one could judge ideas of reference from eye-ear
emphasis. The authors conclude, however, that eye-ear emphasis on the

D.A.P. is valid, but inefficient for predicting ideas of reference.

Psychopathy and Aggression.--Craddick (1962) was interested in

 

the value of the D.A.P. for predicting psychopathy. The drawings of a

18
man by 28 male psychopathic prisoners and 28 male college students were
compared in terms of characteristic features which Machover suggested
as indicators of psychopathy (i.e. large drawings, two drawings centrally
located or shifted to the left side, shading evidenced, person in
drawing shown with a hand in a pocket). None of the features dif-
ferentiated the two groups except shading which was in the opposite
direction as expected, the college population evidencing more shading
than the psychopathic group.

Goldstein and Moss (1957) studied the effects of experimentally
induced feelings of aggression on figure drawing in terms of four
variables, line pressure, figure size, a group of seven specific drawing
details (i.e. slash-line mouth, detailed teeth, spiked fingers, clenched
fists, nostril emphasis, squared shoulders, toes in non-nude figures
and overall subjective judgement of aggression.) Induced aggression
'was interpolated between two figure drawing presentations. They found
that of the four variables, only the seven specific drawing details as
a group significantly related to aggression. The authors state that a
dichotomy appeared to have formed, with non-significant results being
found in the areas measuring aggression in the form of psychomotor
tension and with significant results being found in the areas in which

aggression is expressed symbolically.

Summary

On the whole, the evidence presented in this review does not
seem to support Machover's claims regarding the value of the D.A.P. as
a diagnostic tool. Although the studies reviewed have employed a number

of different approaches, from attempts to validate individual signs to

19
the more global attempts at evaluation, throughout all, the D.A.P.
comes through, at best, with only questionable validity. Despite these
difficulties, the D.A.P. continues to enjoy a great deal of popularity
as a diagnostic instrument. Perhaps this is due, as one writer states,
to the fact that it ". . . proves to be no worse than any other common

personality assessment procedure (Blum, 1954 pg. 125)."

 

’MV- ‘1‘..—

 

E'fla—n- ‘- '

 

.4 a: ..
-m- 'I" i! .-_l.- ”Eur

1"!

APPENDIX B
The Standard Adjective Q;Sort (Corsini, 1956)

Characteristics.--The S.A.Q.S. consists of 50 personal adjectives
printed on cards which may be sorted for perceptions of self and others.
These 50 adjectives are laid out in a rectangular distribution of ten
columns of five words each. With the help of a special table provided
in the manual, any two sorts may be correlated to obtain an index of
similarity.

The word list itself has been devised so that it contains items
which are familiar to adults of average intelligence. In addition, steps
were taken by the author to keep the list free of ambiquity, and stereo-
typed words and to insure that there was representation of the seven
major areas or factors of personality. Finally, every word has a

favoribility index which represents the favoribility ratings of 25 adult

subjects.

Raliability,--The reported test-retest reliability for two
college groups for self-sort a week apart was .81 for females and .79

for males (Corsini, 1956). No other reliability estimates are avail-

able at present.

validity.--Few validation studies on this instrument have been
reported to date. However, Corsini (1956), summarizing two of his own

investigations, makes a reasonable case for the usefulness of the

20

21
S.A.Q.S. in marriage counseling. Using the Burgess Happiness Test as
the criteria, the S.A.Q.S. was demonstrated to have a predictive

efficiency of .75 as a measure of marital adjustment.

Instructions.--"Here are 50 cards, each with a word that can be

 

used to describe people. You are to lay these cards out on the table

in such a manner that you can see all of them in ten piles or columns,
five words in each column-pile. Those words that will be in the last
column to the left are those that you think describe you l§§§£_. . .

and the five words in the pile to the extreme right will be those that
you think describe you best. The other words you will distribute in the
same manner in terms of how well they describe you. Before you finish,
look over all of the words to make sure that as they go to the right

they describe you better and better (Corsini, 1956)."

S.A.Q,S. Word List

 

 

 

Life Style 11. Resourceful
l. Daring 12. Clever .
2. Cautious 13. Clear-Thinking
3. Aloof Emotional Stability_
4. Adventurous l4. Responsible
5. Secretive 15. Dependable
Intelligence 16. Self-Controlled
6. Foresighted 17. Moderate
7. Witty l8. Persevering
8. Shrewd l9. Jealous
9. Imaginative 20. High-Strung

10. Original 21. Nervous

22. Suspicious
23. Immature
Dominance

24. Aggressive
25. Forceful
26. Independent
27. Stubborn
28. Dominant
29. Outspoken
30. Submissive
31. Dependent
Activity

32. Quick

33. Hasty

34. Hurried
35. Talkative

36. Active

22

Social Sensitivity_

 

37. Warm-Hearted
38. Soft-Hearted
39. Gentle

40. Appreciative
41. Discreet

42. Unselfish
43. Insensitive
44. Sensitive
45. Sarcastic

Mood

 

46. Easy-Going
47. Calm

48. WOrrying
49. Emotional

50. Excitable

APPENDIX C

Table I

Means and Standard Deviations for Ulcer and Orthopedic groups

on age, education, and I.Q. N = 20 in each group.

 

 

 

 

 

Orthopedic Ulcer
Mean S.D. Mean S.D.
Age 40.90 9.62 40.85 8.24
Education 8.85 3.08 9.15 1.80
I.Q. 99.60 14.96 99.75 13.95

 

23

APPENDIX D

TABLE II

Inter-judge comparisons on average Q sort
ratings for Ss.

 

 

 

Group 2 Value
Ulcer .464*
Orthopedic .33l*

 

*Differences non significant at .05 level.

24

REFERENCES

Blum, R. H. The validity of the Machover DAP technique. J. clin.
Psychol., 1954, 10, 121-125.

Brown, D. G. and Tolor, A. Human figure drawings as indicators of
sexual identification and inversion. Percept. mot. Skills,
1957, 7, 199-211.

 

Corsini, R. J. S.A.Q.S. (Chicago Q Sort) Manual of instructions.'
Chicago, 111.: Psychometric Affiliates, 1956.

Craddick, R. A. Draw-A-Person Characteristics of psychopathic prisoners
and college students. Percept. mot. Skills, 1962, 15, 11-13.

 

Fenichel, O. The psychoanalytic theory of neurosis. New York:
W. W. Norton and Co., 1945.

 

Fisher, G. M, Relationship between diagnosis of neuropsychiatric
disorder, sexual deviation, and the sex of the first-drawn
figure. Percept. mot. Skills, 1959, 9, 47-50.

 

Frank, G. H. A test of the use of a figure drawing test as an in-
dicator of sexual inversion. Psychol. Rep., 1955, 1, 137-138.

 

Friedman, I. Phenomenal, ideal and projected conceptions of self.
J. abnorm. soc. Psychol., 1955, 51, 611-615.

 

Goldstein, A. P. and Moss, R. L. The validity of interpretive signs
of aggression in the drawings of the human figure. J. clin.
Psychol., 1957, 13, 169-171.

Grams, A. and Rinder, L. Signs of homo-sexuality in human figure
drawings. J. consult. Psychol., 1958, 22, 394.

 

Griffith, A. V. and Peyman, D. A. R. Eye-ear emphasis in the D.A.P.
as indicating ideas of reference. J. consult. Psychol., 1959,
23, 560.

 

Hardy, M. W. The interaction of patient's intelligence and other
factors with clinicians' skill in the diagnostic use of human
figure drawings. Dissertation Abstr., 1960, 20, 4438.

 

Kamano, D. K. An investigation on the meaning of human figure drawings.
J. clin. Psychol., 1960, 16, 429-430.

 

25

26

Lakin, ML Certain formal characteristics of human figure drawings
by institutionalized aged and by normal children. J. consult.
Psychol., 1956, 20, 471-474.

 

Lubin, B. Differentiation of overtly stable and unstable psychiatric
aides by means of the D.A.P. test. Psychol. Rep., 1959, 5, 26.

 

Machover, Karen. Personality Projection in the drawing of the human
figure. Springfield, 111.: Charles C. Thomas, 1949.

 

McGuire, D. and Moss, C. S. An indirect validation study of the
Draw-A—Person Test through the cartoons of William Steig.
J. proj, Tech., 1962, 26, 88-95.

 

Murphy, M; M. Sexual differentiation of male and female job applicants
on the Draw-A-Person Test. J. clin. ngchol., 1957, 13, 87-88.

Reyher, J. Unpublished Manuscript. 1961.
Reznikoff, M, and Nichols, A. L. An evaluation of human figure drawing

indicators of paranoid pathology. J. consult. Psychol., 1958,
22, 395-397.

 

Ribler, R. 1. Diagnostic prediction from emphasis on the eye and ear
in human figure drawings. J. consult. Psychol., 1957, 21,
223-225.

 

Sherman, L. J. The influence of artistic quality on judgements of
patient and nonpatient status from human figure drawings.
J.yproj. Tech., 1958, 22, 338-340.

Silverstein, A. B. and Robinson, H. A. The representation of orthopedic
disability in children's figure drawings. J. consult. Psychol.,
1956, 20, 333-341.

 

Singer, R. H. Various aspects of human figure drawings as a personality
measure with hospitalized psychiatric patients. Dissertation
Abstr., 1958, 18, 290.

 

Staltz, R. E. and Colthrop, F. C. Clinical judgements and the Draw-A-
Person Test. J. consult. Psychol., 1961, 25, 43-45.

 

Swenson, C. H., Jr. Empirical evaluations of human figure drawings.
ngch. Bull., 1957, 54, 431-466.

Vilhotti, A. J. An investigation of the use of the D.A.P. in the
diagnosis of homosexuality in mentally deficient males. Amer. J.
ment. Defic., 1958, 62, 708-711.

Wacks, H. and Zaks, M. S. Studies of body image in men with spinal
cord injury. J. nerv. ment. Dis., 1960, 131, 121-127.

 

27

Walker, Helen M. and Lev, J. Statistical inference. New York:
Henry Holt and Co., 1953.

 

Wallon, E. J. A study of criteria used to differentiate the human
figure drawings of normals, neurotics and psychotics.
Dissertation Abstr., 1959, 20, 1873.

 

Whitaker, L. Jr. The use of an extended Draw-A-Person Test to identify

homosexual and effeminate men. J. consult. Psychol., 1961, 25,
482-485.

Wiggenhorn, A. H. An investigation of changes in human figure drawings
as a function of changes in the "self-concept". Provo. £9p.,
1957, l, (4), 15-37 (Abstract).

 

Wylie, Ruth C. The self-concept. University of Nebraska Press, 1961.

 

, 3a

2;"

iii?

«a, a
25.31.

USE 0313i?

flefififii‘
‘ A,"'.

i E f.‘
'1 s I“?
dfl‘. 3 L

~~¢ All

 

.. ‘.
{Eff-.5

Ufa

i

h:

7