THE EFFECTS OF STRESS ON THE
DRAW A PERSON TEST

Thaw: kw the Dogma of M. A.
MlCHIGAN STATE UNIVERSE”
Leonard Handler
1963

ABSTRACT
THE EFFECTS OF STRESS ON THE DRAW A PERSON TEST
by Leonard Handler

External stress was hypothesized to increase mani-
festations of anxiety on the DAP. The anxiety was hypothe-
sized to be a function, in part, of the symbolic meaning
of the figures.

Fifty-seven male college students drew a male,
female and automobile (control figure) under stress and
nonstress conditions. Both hypotheses were supported;

15 of the 21 anxiety and conflict indices significantly
differentiated between the stress and nonstress conditions
for the male drawing, 11 for the female drawing, and 5 for
the automobile drawing. Five indices significantly
differentiated in the opposite direction.

Suggestions were made for the use of the automobile
drawing as an aid in checking out clinical hypotheses in a
diagnostic evaluation--i.e., the role of drawing skill.

The data also offer support for the validity of the projective

hypothesis for the DAP.

THE EFFECTS OF STRESS ON THE DRAW A PERSON TEST

BY

Leonard Handler

A THESIS

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

MASTER OF ARTS

Department of Psychology

1963

G)

V
g,\

. ‘...“\.

ACKNOWLEDGEMENTS

I would like to thank my committee chairman, Dr.
Joseph Reyher, for his thoughtful guidance and assistance.

I am also very grateful to the other members of my committee,
Dr. Charles Hanley and Dr. Bill Kell, for their support

and their suggestions for improving the study. Their
comments also did much to clarify the presentation of
findings.

Herbert Potash deserves a Special thank—you for
the tedious and time consuming task of rating the drawings,
and for his suggestions at the inception of the study. My
wife Barbara read the manuscript and offered a number of
cogent, insightful suggestions. I would like to thank
her for this help, and for her infinite patience.

Dr. Bertram Karon was extremely helpful: I am grateful
to him for his constant support and encouragement, and for
his editorial assistance.

Several other people have generously given their
assistance. I would like to thank Jospeh Levine, Dr. Harold
Rosen and Larry Kipperman for their assistance in various
phases of the study.

ii

TABLE OF CONTENTS

Page

METHOD . . . . . . . . . . . . . .’. . . . . . . 3
Nonstress Procedure 3
Stress Procedure 4
Scoring 5
Automobile Drawing 6
RESULTS . . . . . . . . . . . . . . . . . . . . . . . 7
Reliability 7
Hypothesis One 8
Hypothesis Two 9
DISCUSSION . . . . . . . . . . . . . . . . . . . . . ll
Hypothesis One ' 11
Order Effects 20
Drawing Patterns 21
Hypothesis Two 24

The Projective Hypothesis 27
REFERENCES . . . . . . . . . . . . . . . . . . . . . 34

iii

LIST OF TABLES

Table Page
1. Reliability of the anxiety indices for the
male, female, and automobile drawings . . 29
2. Comparison of stress and nonstress drawings . 30
3. Figure drawing anxiety indices investigated in
this study, compared with prior findings . 32

iv

THE EFFECTS OF STRESS ON THE DRAW A PERSON TEST
Leonard Handler
Michigan State University

A recent article by Sundberg (1961) reveals that the
DAP is the second most frequently used test in hospitals,
clinics, and counseling services throughout the country.
However, research results leave serious doubts regarding
the validity of many hypotheses concerning symbolic
representation of traits, tendencies, and conflicts in
figure drawings (Swensen, 1957). Despite these doubts,
evidence in the individual case frequently points to the
test's utility in clinical practice.

The equivocal nature of research results on the
DAP is not surprising in light of the comparatively crude
investigatory methods that have usually been employed.
For example, anxiety indices are often related to
questionable external criteria of anxiety. Differences are

sought among nosological categories, which are known to be

unreliable and imprecise. Problems also arise in matching
subjects, and in dealing with the high degree of intersubject
variability in drawing. As Nichols and Strfimpfer (1962)
point out, "If there are, indeed, aspects of figure drawings
which are related to the personality of the person making

the drawing, such a relationship is likely to be obscured

in any group comparison by the large individual differences
ixioverall.qua1ity"i(p. 161). Accordingly, an approach which
uses the subject as his own control would be improvement

in design.

In the course of another research project 3 observed
that anxiety might be easily manipulated in connection with
figure drawings. When §s were attached to a large six
channel polygraph, in a soundproofed room, in the presence
of g, the drawings seemed to manifest more indices of anxiety,
disorganization, and privitivization than in the standard
testing situation. .E felt these disturbances were a product
of the stress produced by the surroundings, and the
experimental procedure, as well as reflecting the intrapsychic
stress produced by the drawing of the human figure itself.

The above observations led to two hypotheses:

l. Externally induced anxiety increases manifestations of

anxiety on figure drawings. 2. There are two sources of

these manifestations of anxiety: a) the laboratory stress
situation b) anxiety producing intrapsychic processes,

activated by drawing the human figure.

METHOD

Fifty—seven male students, from introductory psychology.
classes, were tested under stress and nonstress conditions.
Twenty-one of the gs were tested under stress conditions
first, while 36 were tested under nonstress conditions

first.

Nonstress Procedure

 

I§s were tested in a group. They were given four
sheets of 8-1/2 x 11 unlined paper, and were asked to number
the sheets consecutively. Names were omitted until all the
drawings were completed, and § did not know that he would
later be required to identify himself. .§s were also given
four pencils, a separate one to be used on each page, to
insure that each drawing was made with a freshly sharpened
pencil. .gs were told that except for the series of drawings
on the first page, the other three drawings were to be made
on separate sheets of paper, one drawing per page. On the

first page gs were requested to draw a series of eight

simple geometric figures, as a warm-up procedure. Following

this, §s drew, in turn, a male, an automobile, and a female.

Stress Procedure

§_was ushered into a tiny, dimly lit room. The
soundproofing material on the walls, and the double doors
gave many §s the impression of a "padded cell." The room
was crowded with apparatus, including a rather large and
complicated Grass six channel (Model 5) Polygraph. .g
was seated in an odd looking chair, with double armrests.
lg then attached GSR electrodes to the fingers of the
nonpreferred hand. A fifteen minute waiting period followed,
where g sat quietly and E busied himself manipulating the
numerous controls on the polygraph. All six amplifiers were
then turned on, and the ink writing oscillograph was set in
motion. .§ appeared quite startled when he observed the
record the pen was tracing. The whining, whirring and
cliCking of the motor and timer, the hum of the bank of
amplifiers, and their gleaming red indicator lights all
were calculated to increase S's anxiety level. The warm—up
procedure used in the nonstress session was repeated, except
that now, § wrote his name on the first sheet. He was then

asked to draw, in turn, a male, an automobile, and a female.

While §_drew, E busied himself operating the p01ygraph,
writing on the chart paper, and occasionally glancing at the
drawing over S's shoulder.

Although gs were not directly asked if they were
anxious, many volunteered this information. Others expressed
such feelings indirectly. Many §s hesitated to even enter
the room, and almost all §s remarked “anxiously" about the
surroundings and the polygraph. .g observed hand tremors,
exaggerated swallowing, blushing, and heavy, labored
breathing. GSR responses, although not employed here in
any controlled manner. were quite massive. The total
impression was that the SS were extremely anxious, because
of the vis-a-vis stress situation where §'s identity was
known, the intimidation in the testing session, and the

stress involved in the drawing task itself.

Scoring

All drawings were first coded and were then scored
for the 21 anxiety indices, one index at a time.) The
drawings were scored for degree of presence or absence of
each index, using a modified and extended Hoyt-Baron
scoring manual (Hoyt, 1955: Hoyt '&. Baron, 1959). Hoyt

and Baron report very small mean differences between two

raters, while Mbgar (1962) reports scorer reliabilities
ranging from .84 to 1.00.

Score values usually ranging from zero to three
were assigned. Several of the indices were instead scored
by assigning a plus to the drawing of a pair which showed
relatively more of a factor than the other drawing.

Previous research (Bradshaw, 1952; Lehner &
Gunderson, 1952) has shown that the test-retest reliability
of the indices of the kind noted in this study are high
enough for research purposes (the lowest reliability

reported was .45, while the highest was .93).

Automobile Drawing

The automobile drawing was included as a control
figure; Reyher (1959) has hypothesized that the automobile
drawingis a more neutral figure than the drawing of a
person, and is therefore less subject to projection than
either the male or the female drawing. If the same
disturbances occur on the drawings of the human figures and
the automobile, the disturbances merely reflect anxiety
produced by the task of drawing itself. However, when the
disturbances are associated only with the human figures,

they are likely to reflect intrapsychic sources of anxiety.

Before the figure could be used, however, the task
difficulty had to be compared with the difficulty of drawing
a person. Sixty—two male college students were asked
to draw a man and an automobile. A counterbalanced procedure
was employed. .§s were asked to rate comparatively the
automobile drawing as: a) easier, b) more difficult, or
c) equally as difficult to draw as the male drawing. A
nonsignificant chi square (for independent samples) of
.26 was obtained. The results indicated that there was
no significant difference in task difficulty between

drawing an automobile and drawing a male.

RESULTS

Reliability

 

Insert Table 1

about here

 

Table 1 shows the percent agreement between two
raters, scoring 90 drawings for degree of presence of each
indicator. The reliabilities, comparable to those reported

by Mogar (1962), and by Hoyt & Baron (1959), demonstrate

that the indices may be scored with a high degree of ;

reliability.

Hypothesis One

Hypothesis one was tested by comparing the stress
and nonstress drawings of each subject, on each of the indices.
A two tailed sign test was used to compare the degree of
difference on each of the indices. Table 2 summarizes the

results for the male, female, and automobile drawings.

 

Insert Table 2

about here

 

Male Drawing. Fifteen of the 21 indices significantly
differentiated the stress from the nonstress drawings, for
the total sample.

Female Drawing. .Eleven of the 21 indices significantly
differentiated the stress from the nonstress conditions, for
the total sample.

Automobile Drawing. Five out of 18 indices
significantly differentiated the stress from the nonstress
drawings. For obvious reasons, four of the indices used

in the analysis of the other two drawings (head size, head:

body ratio, head simplification,'and hair shading)..
were not applicable for the automobile drawing analysis.

Line pressure, erasure, and loss of detail
differentiated on all three drawings. Although shading,
hair shading, erasure, reinforcement, placement, and
emphasis lines, did differentiat significantly between stress
and nonstress drawings, they did so in the direction opposite
to that hypothesized by Buck and Machover. There was more
shading, hair shading, erasure, reinforcement, emphasis
line and placement in the upper left hand corner in the
nonstress rather than in the stress drawings. The other
significant indices differentiated the stress from the
nonstress drawings in the expected direction; they were
present more frequently in the stress situation. The

results therefore indicate that hypothesis one is supported.

Hypothesis Two

If the role of intrapsychic stress is involved in
differing degrees for the three drawings, one would expect
significant differences between the stress and nonstress
findings for the three drawings. Therefore, hypothesis two
was tested by comparing the significance levels for the

three drawings, using the Friedman TWO-Way Analysis of

10

Variance test. An g;2 of 11.6 was obtained, significant
between the .01 and .001 levels of confidence. The results
indicated that the number and degree of significance levels
for the three drawings was significantly different. The
automobile drawing had the fewest number of significant
indices (five), the female drawing, 10, while the male
drawing had the most (13). The results therefore seem to
support hypothesis two. The automobile drawing, assumed

to pose less intrapsychic stress than either the male or
female drawings, showed far less change under externally
induced stress than did the other two drawings. Individual
comparisons were done, using the sign test. The male
drawing was significantly different from the automobile
drawing (p é .001) in number and degree of significance
levels. When the female and automobile drawings were
compared, the significance level was .006, while the male-
female comparison resulted in a p value of .058. These
findings also support Reyher's formulations concerning

the more neutral nature of the automobile drawing.

11

DISCUSSION

Hypothesis One

Opposite Findings: shading, hair shading, erasure,
reinforcement, and emphasis lines. Such findings are quite
surprising, since presence, and not absence of these indices
has traditionally been interpreted to denote anxiety and
conflict. On the other hand, 11 of the other indices
differentiated between the stress and nonstress groups in
the expected direction, thereby supporting traditional
clinical interpretation. More important, however, is that
other experimenters, using hospital populations, have found
similar trends in the "opposite" direction for shading,
hair shading, erasure, and reinforcement. Table 3 summarizes
these findings, along with other research findings (findings
which agree with those of the present study, findings in
the opposite direction, and nonsignificant findings)

for the remainder of the indices.

 

Insert Table 3

about here

 

-,12
The results of the present study, and other
supporting research indicate the need to exercise extreme
caution in the interpretation of anxiety or conflict from
the presence of shading, hair shading, erasure, or
reinforcement. While the presence of these indices may
possibly denote anxiety, they may just as frequently 4
represent S's appropriate attempts to make the figures as
true to life as possible, to differentiate body areas,
and to give the figures more substance. Hammer (1959)
notes that some erasure, with subsequent improvement,
is a sign of adaptiveness and flexibility. The erasures
in the present study seemed to improve rather than distort
the figures. Thus, it appears that an absence of shading,
hair shading, erasure, reinforcement, and emphasis lines
rather than their presence may indicate anxiety or conflict.
Similarly, a reported finding that erasure was significantly
and positively related to intelligence (Mogar, 1962)
suggests that erasure (and probably also shading, reinforcement,
and emphasis lines) is a concomitant of general awareness
of oneself as he relates to, and yet is separate from his
environment. The presence of the above indices (along with

the absence of distortion, head and trunk simplification, etc.)

'13
may mean that S meets and reacts to reality situations
appropriately.

There may be an impairment of awareness under
stress, and a loss of differentiation from the surrounds,
which in turn becomes reflected in S's drawings. Such a
formulation is consistent with Mogar's findings of a positive
relationship between the estimation of head size relative
to figure size and both Manifest Anxiety Scale and Rorschach
Content Test scores. These findings suggest that "anxiety
is a condition which lessens the differentiation between
self and environment" (1962, p. 6). As Mogar points out,
this is essentially the position of Werner and Wapner
(1952: Wapner and Werner, 1958), whose sensory-tonic perceptual
theory states that perceptions of one's own body parts and
of the "self" tend to be overestimated under conditions
which decrease differentiation of self and world. The
findings of increased distortion, disturbed vertical
balance, lack of body area delineation lines, body and head
simplification under stress all tend to support this
formulation. Similar results have been reported for changes
in figure drawings under LSD, a drug noted for producing

disturbances resembling those found in severe ego disturbance.

14

A number of studies summarized in Table 3 also support the
above formulations.

Opposite Findings: placement. The finding that
placement upward and to the left was more frequent in the
nonstress male drawings is also contrary to traditional
clinical interpretation. An explanation for these results
is suggested in the work of Dennis (1958), and of Dennis
and Raskin (1960). In studies of various linguistic groups,
the authors found a significant relationship between the
quadrant in which g began to write, and the location of the
drawings. Thus, although the location of drawings may be
related to anxiety, habit transfer from learning to write in
a particular way may account for the placement of nonstress
drawings. However, habit transfer does not account for the
"movement downward" of the drawings in the stress group.
Perhaps the reSults mirror the disorganizing effects of.
stress.

Opposite Findings: An Alternative Interpretation.
One possible explanation for the lack of shading, erasure,
reinforcement, and emphasis lines, as well as for the loss
of detail in the stress drawings is that §_wanted to escape
from the stress situation, and therefore complied minimally.

§ would not take the time to detail the figure, to erase

15

and fix the parts that were not drawn well, and so forth.

The findings for the automobile drawing cast doubt on this
formulation, for although detail loss and erasure were
significant for the automobile drawing, the results for

shading, reinforcement, and emphasis lines were not significant.
In addition, both the stress and nonstress automobile drawings
were of high quality, while the male and female stress

drawings were generally of poorer quality than the corresponding
nonstress drawings. It would appear, then, that while some

of the difference can be ascribed to S's desire to quickly
leave a threatening situation, this by no means accounts for
the total differential results for the three drawings on

the remaining indices.

g Expected Findings: body simplification and head
simplification. Signs of regression were evident in the
results obtained for these two indices. Head and body were
more frequently drawn in a primitive and schematic manner
under stress. Table 3 indicates that similar findings have
been reported by other investigators. Decrements in mental
age scores on the DAP have been found under conditions of
frustration (Seashore & Bavelas, 1942), while regression
in mental age scores of young children as a result of stress

has also been demonstrated (Barker, Dembo & Lewin, 1941).

16

Expected Findings: light line, heavy line, and
line pressure. The significant findings for line pressure
on all three drawings seems to reflect an increase in energy
output due to external stress. Similar findings have been
reported by Ford (1929), Sherman (1942); and a number of
others (summarized in Poffenberger, 1942). The significant
finding for heavy line in the female drawing is probably,
in part, a reflection of increased work level due to external
stress. The significant finding for light line in the male
drawing seems to reflect intrapsychic stress; light line
did not differentiate between stress and nonstress in either
of the other two drawings. Although heavy line did not
differentiate between the stress and nonstress male drawings,
there was a definite trend in that direction (p = .09).

The results suggest that both light and heavy line
quality may be indicative of anxiety, but that different
processes underlie their presence. It is possible that
heavy line quality reflects feelings of external stress,
or pressure from without, while light line represents a
feeling of stress eminating from with g. Gutman (1952)
found that patients who did not improve in psychotherapy
tended to draw continuous and reinforced lines, while

those who improved tended to draw their figures with a light,

l7

sketchy line. The present interpretation of the differential
meaning of heavy and light line is consistent with Gutman's
findings: psychotherapists have stated that those who experience
anxiety about themselves make better patients, and improve
more quidkly than those who see their problems as emanating
from outside sources. Similarly, Reznikoff and Nichols
(1958) found that heavy line significantly differentiated
drawings of paranoid schizophrenics from a group of
undifferentiated schizophrenics. A paranoid typically
views the world about him as the source of his fears.

Expected Direction: size. Under stress, the
drawings both increased and decreased in size, depending
upon the order of administration of the stress procedure
(p = .02 for increase, and .02 for decrease). The
differential results for size suggest that both increases
and decreases in size may indicate anxiety. Thus, Wallach
(1960) reports that high anxiety social introverts were
more expansive, graphically, than nonanxious social introverts,
while social extroverts high in anxiety were more constricted
than nonanxious extroverts.

Nonsignificant Findings: transparency. It is not
surprising that transparency did not differentiate between

the stress and nonstress drawings, since the presence of

18

transparency has traditionally been said to indicate rather
serious reality testing impairment, extremely poor ego
defenses, and to be the most pathological form of conflict
expression (Machover, 1951). The results would probably be
very different with a neurotic or psychotic group placed
under stress.

On the whole, the findings of this study are in
substantial agreement with past studies 6n the DAP. A
total of 106 findings were in agreement with the results
reported here, while only 28 findings were in disagreement,
and 46 findings were nonsignificant.

The summarized research in Table 3 indicates that
omission, distortion, detail loss, line pressure increase,
heavy line,-size increase and decrease, head simplification
and trunk simplification have consistently yielded results
similar to those found in the present study. Evidence is
somewhat less consistent for reinforcement, line discontinuity,
light line, vertical imbalance, and delineation line absence.
With most of these indices, much of the disagreement has
been that in some studies the indices do not significantly
differentiate anxious from nonanxious gs. HOwever, at
least as many studies report that these indices g2

differentiate significantly between anxious and nonanxious

l9

I§s, in agreement with the present findings.

With shading, hair shading and erasure the evidence
in agreement with the present findings is balanced by a
similar number of studies reporting findings in the opposite
direction, and a similar number of studies reporting
nonsignificant findings. Although such findings do not seem
unusual to anyone familiar with DAP research literature,
these results are more atypical than they appear at first.
The present study found shading, hair shading,and erasure to
be present more frequently in the nonstress situation rather
than in the stress situation, a finding contrary to
Machover and Buck's predictions, and contrary to generally
accepted clinical interpretation. A number of other i
independent research findings are in substantial agreement
with these "opposite" results.

The summarized research for placement, head size,
head:body ratio, emphasis lines and transparency offer
some support for the present findings, but in general,
'research evidence does not support these findings. Perhaps
the discrepancy lies in differences in degree of pathology,

as in the findings reported for transparency.

20
Order Effects

With only one variable (size), and in only one
drawing (male), did the order of administration result in
statistically significant opposite results for the "stress
first" vs the "stress second" administration. When the
stress drawing was done first, the nonstress drawings were
bigger: when the stress drawing came second, the nonstress
drawings were smaller. In each instance the first drawing
administration, be it stress or nonstress, resulted in
smaller drawings. It was possible, therefore, that what
appeared to be differential results due to stress and order
might be due, instead, to either a second test administration,
or to chance. To determine whether "second administration"
was responsible for the differential results another set
of drawings was collected from 40 male introductory
psychology students, under two nonstress testing conditions,
one week apart. The results indicated that there was no
difference in size from the first to the second administration
(p = 1). Thus, the explanation for the order effect
probably rests in §'s interpretation of the drawing procedure,
and in the difference in the degree of familiarity he had

with the task. Second administration, per se, was not a

21

factor in producing the order effect.

Drawing Patterns

Two drawing patterns seemed to emerge under stress
conditions--constriction, and expansion. The constricted
drawings had lines that were heavier, more even and continuous,
despite mechanical breaks in the line where the pencil was
lifted from the paper. The lines appeared far less sketchy,
the body parts appeared as if they were not quite attached.

In addition, there was a decrease in size in the stress
drawings, compared with their nonstress counterparts. This
pattern seems to reflect increased rigidity and constriction

in response to stress. Increased rigidity under stress has
been reported by other investigators (Cowen, 1952: Beier,

1951: Ross, 1952: Marquardt, 1948). Rawn (1957) found that
omission of essential detail, and intrusion of line (similar

to transparency), characterized a rigid group and differentiated
it from a flexible one. She explained the results in terms

of poor ego boundaries for the rigid.

Pattern two was marked by increased diffusion of
body boundaries, and increased vagueness of body parts. Lines
were extremely sketchy, and loosely bound together, if at all.

‘ The lines were lighter, and the drawings tended to be bigger

22

than their nonstress counterparts. There was also a third
group, composed of §s who did not show either pattern, and
whose drawings changed comparatively little under stress.

The first two patterns roughly correspond to Fisher
and Cleveland's (1958) high and low boundary body image
groups, respectively, and to Witkin's high and low
differentiation groups (Witkin, Dyk, Paterson, Goodenough
& Carp, 1962). Witkin and his associates speak of the
"Sophistication—of—Body—Concept" scale, which is similar to
the following scales used in the present study: distortion,
omission, lack of delineation lines, and detail loss. The
scale is seen as a continuum, although both extremes are
emphasized. The more primitive pattern is likened to
hysterical defenses, while the overly structured pattern
is associated with defenses often found in obsessive-compulsive
personalities. Poorly articulated drawings were positively
correlated with global and undifferentiated perception of
the field, while drawings that were highly articulated were
correlated with an analytical field approach.

Similarly, Kagan (in press) hypothesized that an
analytical style represents a cognitive manifestation of
impulse control. He found that people who maintain control

over impulses more frequently exhibit an analytic cognitive

23

response than do people who control their impulses to a
lesser extent. Kipperman (1963) reports significant
differences in use of detail in conceptual organization
under stress and nonstress conditions.

Thus, it appears that figure drawings are capable
not only of reflecting anxiety and conflict, but are also
capable of reflecting S's predominant defensive style, and
his typical style of cognitive approach and organization.
Similarly, Goldworth (1950) has emphasized that the DAP can
be used to determine control, contact with reality, and
accuracy of perception, much like F+ is used on the Rorschach.

Moreover, it is not so much the presence of anxiety
that should be the clinician's focus of interest, but
rather, what is done with the anxiety seems of more importance.
Other more important questions concern the extent, strength
and effectiveness of S's defenses. Such indices as
distortion, body boundary delineation lines, and
primitivization of head and body seem to offer more clues
to S's level of functioning than the more traditional
anxiety indices (shading, erasure and reinforcement).

In addition, the degree of change under stress may
also indicate something about the quality of_§'s defenses.

Those whose drawings changed drastically in response to

24

stress would probably be more unstable than those whose
drawings changed only moderately, in appropriate response

to an external threat. Those whose drawings showed no

change, however, might either be well defended, or perhaps

too well defended, and rather rigid. In deciding which of
these alternatives is appropriate, it would be helpful to
examine the quality of both drawings. Moreover, the automobile
drawing can be extremely helpful in sorting out extraneous
drawing variables. The procedure follows immediately

below, in discussing hypothesis two.

Hypothesis Two

 

The differential results for the three drawings
suggests that the effects of internal stress can be separated
from the role of external stress.

Product-Moment correlations were computed for the
male and automobile drawings in both stress and nonstress
conditions, and for the male drawings in the stress and
nonstress situations. The low correlations between the
male and automobile drawings (.10 for the stress situation
and .20 for the nonstress situation) seems especially
significant in light of the marked disparity from the

high test-retest reliabilities reported in the literature for

25

the indices used in this study. Thus, the low correlations
seem to indicate the independence of the male and automobile
drawings, and reinforce the validity of the differential
findings for the two drawings (i.e., the male drawing
reflects anxiety, while the automobile drawing does so to a
far lesser degree).

The low correlation found for the stress and nonstress
male drawing comparison (.14) suggests that perhaps the
stress enhances the conflict, making it more obvious and
apparent in the stress drawings. Such a formulation seems
likely with a non—hospital group, whose concerns and
conflicts would tend to be less extreme. That something
in addition to the external stress situation is involved seems
highly probable from the differential results obtained for
the three drawings, and from similar findings on other
research comparing anxious and nonanxious groups (Table 3).

The present findings seem to hint, however, that
the DAP perhaps is less sensitive to conflict or anxiety
under nonstress conditions, with the present method of
administration and with less disturbed individuals. A
follow-up study with various psychiatric populations seems
indicated at this point, especially since it may be argued

that college students have relatively minor anxieties and

26

conflicts. Nevertheless, it seems possible to increase the
precision of the DAP, or of the clinician using the test,

so that he will be able to make finer and more accurate
predictions about a patient. This may perhaps be possible
by modifying the administration procedure. For example, a
drawing or series of drawings may be obtained as soon as

the client enters the testing room, and another series
obtained after the client and clinician have had a chance to
become acquainted with each other and rapport established.
Extreme care would be needed in sorting situational anxiety
from pathology, and perhaps at first such an approach should
be used only as a research tool.

Thus, erasure, line pressure, detail loss, and
perhaps head and body simplification probably reflect the
role of external stress more than internal stress, since
they were significant for all three drawings.2 Sorting
out these two sources of stress may perhaps be done with
the use of the automobile drawing as a more neutral control

figure. Those indices that differentiate only in the male

2Head and body simplification were scored for the
male and female drawings only, and were significant for
both of them.

27

drawing, or in both the male and female drawings probably
represent internal conflict. Such an approach allows

the clinician to make finer distinctions about a patient's
functioning, above and beyond the uSual statement that the
patient is anxious.

Another important contribution of the automobile
control figure is that the clinician can now determine the
contribution of drawing style and drawing quality in
interpreting DAP records, and can distinguish these factors
from more dynamic factors. A control would be especially
important here, since it has been demonstrated that
clinicians often make judgments more in terms of artistic
quality than in any other way (Whitmyre, 1953; Nichols &
Strfimpfer, 1962). Even though some aspects of style and
quality relate very closely to personality factors, a more
neutral drawing, used as a control or base rate from which
a clinician can generate hypotheses about a patient, becomes

vital in sorting fact from artifact.

The Projective Hypothesis

 

The differential results obtained for the three
drawings, plus the fact that the male drawing had the greatest

number of indices which significantly differentiated the

28

stress from the nonstress drawings, constitute evidence for
support of the projective hypothesis in figure drawings.
Male drawings done by males changed the most under stress,
while the reaction to the female drawing was less extreme.
The reaction to the automobile drawing was far below the
reaction to either of the other two drawings.

The above statement does not imply that the drawing
of a male, done by a male, is necessarily a representation of
the self. It may very well be that the drawing is that of
an idealized self or of a parent. The data suggest, however,
that S does project more of his conflicts and anxieties
into a same sex drawing than into either an opposite sex
drawing or a more neutral drawing of equal task difficulty.

In conclusion, the results of the present study
suggest that §s do reflect the results of external stress
in their drawings, and that above and beyond this, under
stress the drawings also reflect anxiety stemming from
personal conflicts. In addition, the study also indicates
that a great deal of diagnostic information can be obtained
from the DAP with the use of a control figure of equal task
difficulty, as Reyher (1959) has suggested. The data also
offers support for the validity of the projective hypothesis

for the DAP.

29

TABLE 1

RELIABILITY OF THE ANXIETY INDICES FOR THE MALE,
FEMALE, AND AUTOMOBILE DRAWINGS

 

 

Percent Agreement

 

 

Anxiety Indices Male Female .Automobile
Shading .93 .93 .87
Hair shading .93 .80 ---
Erasure .87 .87 .87
Reinforcement .90 .97 .80
Light line .96 .83 .93
Heavy line .92 .93 .93
Line pressure (increase) .98 .92 .87
Line disc0ntinuitya .93 1.00 .73
Emphasis lines .80 .80 .87
Delineation line absence .87 .73 .80
Transparency3 .93 1.00 1.00
Omissiona .97 .87 .73
vertical imbalance .67 .73 .87
Detail loss .93 .87 .80
Distortiona .73 .80 .80
Head simplification .73 .87 ---
Body simplification .73 .73 .80

 

NOte.--Reliabi1ities are not reported for size,

head size, head:body ratio, and placement, since a ruler
and grid were used for measurement on these indices.

aReliability may be spuriously high since presence
of this index was relatively rare.

. 30'
TABLE 2

COMPARISON OF STRESS AND NONSTRESS DRAWINGS

 

 

 

 

p value
Anxiety Indices Male Female Auto Reference

Findings in the
opposite direction:
(Present in non-
stress drawings
more frequently)
Shading .04 .0001 .07 Machover (1949)
Hair shading .001 .097 --- Machover (1949)
Erasure- .00006 .00006 .0002 Machover (1949)
Reinforcement .00006 .057 .13 Machover (1949)
placement .004 1.00 .13 Buck (1948)

(upper left) .
Emphasis lines .00006 .00006 .13 Machover (1949)
Findings in the
expected direction:
(Present in stress
drawings more
frequently)
Size (increase -

and decrease) .62a . .48 .69 Machover (1949)
Light line .008 .29 1.00 Machover (1949)
Heavy line .09 .02 .58 Buck (1948)
Line

discontinuity .008 .36 .75 Caligor (1957)
Line pressure

increase .02 .0001 .00006 Caligor (1957)
:Delineation

line absence .058 .004 .02 Witkin et a1. (1954)
Omission .02 .37 .05 Machover (1949)

(Table continued on next page)

31

TABLE 2 (Continued).

 

 

 

 

 

p value

Anxiety Indices Male Female Auto Reference
Vertical

imbalance .001 .002 .62 Caligor (1957)
Detail loss .00006 .003 .02 Caligor (1957)
Distortion .001 .01 1.00 Machover (1949)
Head

simplification .001 .0001 --- Goldworth (1950)
Body

simplification .001 .00006 .06 Goldworth (1950)
Nonsignificant
findings:
Head size .42 1.00 --- Machover (1949)
Head:body

ratio .48 .62 --- Machover (1949)
Transparency .18 .18 .42 ' Machover (1951)

aStress first drawings significantly smaller (p = .02):

stress second drawings significantly larger (p = .02).

‘32

TABLE 3

FIGURE DRAWING ANXIETY INDICES INVESTIGATED IN THIS STUDY,
COMPARED WITH PRIOR FINDINGS

 

 

 

(Table continued on next page)

Disagreement
With Present Disagreement
Findings: With Present
Agreement Findings in the Findings:
Anxiety with Present Opposite Nonsignificant
Indices Findings Direction Findings
~ Shading 13a, 14, 18, 3, 11, 13b, 30 6, 20, 29, 35,
43e 40
‘Hair shading 19 30 29
‘Erasure 9, 14, 18, 3, 22, 44 20, 29, 35
43e
\ Reinforce— 18, 29, 368, 36 20, 35
ment 43¢, 45
‘ Placement 7, 8, 44 3, 9, 20 15, 18, 29, 35,
40, 43e
h-Size (increase 3, 4d, 9, 16, 18, 43e 35, 40, 44
and decrease)20c' 22’ 23c,
24, 25C, 266,
29°, 36d, 42
‘ Head size 20, 29 14, 22, 45 18, 29, 43e
Head:body 20 S, 14, 23, 29
ratio
Light line 3, ll, 31, 18, 20, 29, 35,
36 43e
Heavy line 3, 34, 36 35
Line 3, 16, 36e,
pressure 44
(increase)

73,3

TABLE 3 (COntinued).

 

 

 

Anxiety .
Indices Agree Disagree Nonsignificant
Line dis- 3, 369, 44e ' 18, 35, 436
continuity
Emphasis
lines 3 18, 43e
Delineation 45, 46 18, 43e
line absence
Transparency 35 18, 21, 37,
439, 44e
- Omission 1, 12, 13, 2 20, 35
14, 17, 18,
21, 23, 28,
29, 33, 39
40, 43a, 449,
46
vertical 3, 23 . 40
limbalance
Detail loss 2, 3, 5, 10,
12, 21, 28,
36, 46
Distortion 3, 10, 12, 20, 29, 32
13, 14, 21,
23, 27, 36,
37, 40, 44a,
45, 46
Head 3, 5, 10, 35, 44
:igfilifica- 14’ 17' 18, i"
37, 43¢, 46

(Table continued on next page)

34

TABLE 3 (Continued).

 

 

 

Anxiety

Indices Agree Disagree Nonsignificant
Trunk 3, 10, 14, 17 - 32, 35
Simplifl- 37' 46

cation

 

1. Alexander, in press. 2. Anastasi & Foley, 1944.
3. Berman et al., 1951. 4. Brengelmann, 1955.' 5. Brill,
1937. 6. Craddick, 1962. 7. Dennis, 1958. 8. Dennis &
Raskin, 1960. 9. Eisen, 1951. 10. Elkish, 1945.
ll. Exner, 1962. 12. Fingert, 1939. 13. Fox et. al., 1958.
14. Goldworth, 1950. 15. Goodman & Kotkov, 1953.- 16. Gutman,
1952. 17. Hinrichs, 1935. 18. Holzberg & wexler, 1950.
19. Hoyt, 1955. 20. Hoyt & Baron, 1959. 21. Hozier, 1959.
22. Koppitz, 1960. 23. Koppitz et. al., 1959. 24. Lehner
& Gunderson, 1953. 25. Lembke, 1941. 26. Liebert et. al.,
1958. 27. Meyer et al., 1955. 28. Modell & Potter, 1949.
29. Mogar, 1962. 30. Morganstern, 1959. 31. Pfister, 1940.
32. Prater, 1950. 33. Rawn, 1957. 34. Reznikoff & Nicholas,
1958. 35. Royal, 1949. 36. Silverstein & Klee, 1958.
37. Springer, 1941. 38. Traube, 1937. 39. Utsugi &
Ohtsuki, 1955. 40. vane & Eisen, 1962. 41. Waehner, 1942.
42. Wallach & Grahm, 1960. 43. wexler & Holzberg, 1952.
44. Wiggenhorn, 1957. 45. Witkin et. al., 1954.
46. Witkin et. al., 1962.

 

b
agirls boys Csmaller only dlarger only

e .
compOSlte score

REFERENCES

Alexander, T. A study of perception and integration in
normality and psychopathology: children's drawing
of the human figure. J. genet._ggyghol., in press.

Anastasi,_Anne- & Foley, J. An experimental study of the
,drawing behavior of adult psychotics in comparison
with that of a normal control group. J..Exper.
Psychol., 1944, 34, 169-194.

Barker, R., Dembo, Tamara, & Lewin, K. .Frustration and
regression: an experiment with young children.
Univ. Ia. Stud. Child welf., 1941,18, No. l.

Beier, E. The effect of induced anxiety on flexibility
of intellectual functioning. Psychol. Monogr.,
1951, 65 (Whole No. 325).

Berman, A., Klein, A., & Lippman, A. Human figure drawings
as a projective technique. J. gen. Psychol., 1951,
45, 57-70.

Bradshaw, D. A study of group consistencies on the Draw—A-
Man test in relation to personality projectibn.

Unpublished master's thesis, Catholic Univer., 1952.

35

36

Brengelmann, J. Size and change of size of reproduction as a
measure of expressive movement. Z. Diggnost. ngchol.,
1955, 3, 23-33. (Psychol. Abstr., 30:1014).

Brill, M. A study of instability using the Goodenough
Drawing Scale. J. abnorm. soc. Psychol., 1937, 32,
288-302.

Buck, J. The HTP technique: a qualitative and quantitative
scoring manual. J. clin. Psychol., 1948, 4, 317—396.

Caligor, L. A new approach to figure drawing. Springfield,
111.: Charles Thomas, 1957.

Cowen, E. The influence of varying degrees of psychological
stress on problem solving rigidity. J. abnorm. soc.
Psychol., 1952, 47, 2, (Supplement, 512-519).

Craddick, R., Leipold, W., & Cacavas, P. The relationship
of shading on the Draw-A-Person test to manifest
anxiety scores. J. consult. Psychol., 1962, 26, 193.

Dennis, W. Handwriting conventions as determinants of human
figure drawings.‘J.con§glt. Psychol., 1958, 22, 293-295.

Dennis, W., & RaSkin, Evelyn. Further evidence concerning
the effect of handwriting habits upon the location

of drawings. J. consult. Psychol., 1960, 24, 548-549.

37

Eisen, Virginia. Comparison of human figure drawings by
behavior problem and normal control boys. Unpublished
doctoral dissertation, Fordham Univer., 1951.

Elkish, Paula. Children's drawings in a projective
technique. Psychol. Monogr., 1945, 58 (Whole No.
266).

Exner, J. A comparison of the human figure drawings of
psychoneurotics, character disturbances, normals,
and subjects experiencing eXperimentally induced
fear. J.gproj. Tech., 1962, 26, 329—337.

Fingert, H., Kogan, J., & Schilder, P. The Goodenough
test in insulin and metrazol treatment of
schizophrenia. J. gen. Psychol., 1939, 21, 349-365.

Fisher, 3., & Cleveland, S. Body image andpersonality.
Princeton: Van Nestrand, 1958.

Ford, A. Attention, automatization: an investigation of
the transitional nature of mind. Amer. J; Psychol.,
1929, 41, 1-32.

Fox, Cynthia, Davidson, K., Lighthall, F., Waite, R.,

& Sarason, S. Human figure drawings of high and
low anxious children. Child Develpm., 1958, 29,

297-301.

38

Goldworth, S. A comparative study of the drawings of a man
and a woman done by normal, neurotic, schizophrenic
and brain—damaged individuals. Unpublished doctoral
dissertation, Univer. of Pittsburgh, 1950.

Goodman, M. & Kotkov, B. Predictions of trait ranks from
Draw-A-Person measurements of obese and non-
obese women. J. clin. Psychol., 1953, 9, 365-367.

Gutman, Brigette. An investigation of the applicability of
the human figure drawing in predicting improvement
in therapy. Unpublished doctoral dissertation,

New York Univer., 1952.

Hammer, E. The clinical application ofgprojective drawings.
Springfield, 111.: Charles Thomas, 1958.

Hinrichs, W. The Goodenough drawing test in relation to
delinquency and problem behavior. Arch. Psychol.,
1935, No. 175, 82.

Holzberg, J. & Wexler, M. The validity of human form
drawings as a measure of personality deviation.
J.gproip Tech., 1950, 14, 343-361.

Hoyt, T. Same-sex drawings as a measure of anxiety in
institutionalized women. Unpublished master's

thesis, Kent State Univer., 1955.

39

Hoyt, T. & Baron, M. Anxiety indices in same sex drawings
of psychiatric patients with high and low MAS

scores. J. consult. Psychol., 1959, 23, 448—452.

 

Hozier, Ann. On the breakdown of the sense of reality:
a study of spatial perception in schizophrenia.
J. consult. Psychol., 1959, 23, 185-194.

Kagan, J., Moss, H. & Sigel, I. The psychological significance
of styles of conceptualization. Monogr. Res. Child
_Develpm., in press.

Kipperman, L. Effect of verbal attack on cognitive style.
Unpublished doctoral dissertation, Univer. of
Michigan, 1963.

Koppitz, Elizabeth, Sullivan, J., Blyth, D. & Shelton, J.
Prediction of first grade school achievement with
the B-G test and human figure drawings. J. clin.
Psychol., 1959, 15, 164—168.

Koppitz, Elizabeth. Teacher's attitude and children's
performance on the Bender Gestalt and human figure
drawings. J. clin. Psychol., 1960, 16, 204-208.

Lehner, G. & Gunderson, E. Reliability of graphic indices
in a projective technique (Draw-a-Person). J. clin.

PSYChOloI 1952: 81 125-128.

40

Lehner, G. & Gunderson, E. Height relationships on the
Draw-A—Person test. J. Pers., 1953, 21, 392-399.

Lembke, W. Uber de Zeichnungen "frechen" und ”von Schficternen".
Kindern. Zscht. f. Pad. Psychol., 1939, 31, 459-462.
Cited by Anastasi, Anne & Foley, J. A survey of
the literature on artistic behavior in the abnormal.
IV. Experimental investigations. J. gen. Psychol.,
1941, 25, 187-237.

Liebert, R., werner, H. & Wapner, S. Studies in the effect
of lysergic acid diethylamide (LSD 25): Self and
object—size perception in schizophrenics and normal
adults. AMA Arch. Neurol. Psychiat., 1958, 79,
580-584.

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

Machover, Karen. Drawing of the human figure: A method of

I personality investigation. In H. Anderson &
Gladys Anderson (Eds.j)., An introduction to prcfiective
techniques. Englewood Cliffs, N.J.: Prentice-Hall,

1951.

"0'
r

41

Marquardt, Dorothy. The pattern of punishment and its
relationship to abnormal fixation in adult human
subjects. J. gen. Psychol., 1948, 39, 107-144.

Meyer, B., Brown, F. & Levine, A. Observations on the
HTP drawing test before and after surgery.

Psychosom. Med. 1955, 17, 428-454.

 

Modell, A. & Potter, H. Human figure drawings of patients
with arterial hypertension, peptic ulcer and
bronchial asthma. Psychosom. Med., 1949, 11, 282-292.

Mogar, R. Anxiety indices in human figure drawings: A
replication and extension (extended report). J.

consult. Psychol., 1962, 26, 108.

 

Morganstern, Frances. The effect of an experimental
situation involving failure and disparagement on
certain features of children's figure drawings.
Unpublished doctoral dissertation, New York
Univ., 1959. .(Dissert.4§bstr., 20, 304). .

Nichols, R. & Strfimpfer, D. A factor analysis of Draw—A-
Person test scores. J. consult. Psyphol., 1962,
26, 156-161.

Pfister, O. Farbe und bewegung in der zeichnung geisteskronker.
Schweiz. Arch. f. Neurol. u. Psychiat., 1934, 34, 325-

365. Cited by Anne Anastasi & J. Foley. A survey

42

of the literature on artistic behavior in the
abnormal: Iv. Experimental investigations.
J. gen. Psychol., 1941, 25, 187-237.

Poffenberger, A. Principles of appliedppsychology. New
York: Appleton-Century, 1942.

Prater, G. A comparison of the head and body size in the
drawing of the human figure by hemiplegic and non—
hemiplegic persons. Unpublished master's thesis,
Univer. of Kentucky, 1950.

Rawn, Moss. Degree of disturbance in figure drawings as
related to mode of approach to problem solving.

J. genet. Psychol., 1957, 91, 191-196.

Reyher, J. Use of figure drawings in differential diagnosis.
Paper read at the State Diagnostician's Conference,
Michigan, 1959.

Reznikoff, M. & Nicholas, Alma. An evaluation of human
figure drawing indications of paranoid pathology.
J. consult. Psychol., 1958, 22, 395-397.

Ross, B., Rupel, J. & Grant, D. Effects of personal,
impersonal, and physical stress upon cognitive'
behavior in a card sorting problem. J. abnorm.

soc. Psychol., 1952, 47, 2, Supplement, 546—551.

43

Royal, R. Drawing characteristics of neurotic patients
using a drawing-of-a-man-and-woman technique. .g.
clin. Psychol., 1949, 5, 392-395.

Seashore, H. & Bavelas, A. A study of frustration in
children. J. gen. Psychol., 1942, 61, 279-314.

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

Siegel, S. NOnparametric statistics. New York: McGraw-
Hill, 1956.

Silverstein, A. & Klee, G. A psychopharmacological test of
the body image hypothesis. J. negv. ment. Dis.,
1958, 127, 323-329.

Springer, N. A study of the drawings of maladjusted and
adjusted children. J. genet. Psychol., 1941, 58,
131-138.

Sundberg, N. The practice of psychological testing in
clinical services in the United States. ‘Amer.
Psychologist, 1961, 16, 79-83.

Swensen, C. Empirical evaluations of human figure drawings.
Psychol. Bull., 1957, 54, 431-466.

Traube, T. La voleurdiagnostique des dessins des enfants
difficiles. Arch. de Psychol.,_l937, 26, 235-309.

Cited by Anastasi,.Anne & Foley, J. 'A survey of

44

the literature on artistic behavior in the
abnormal: IV. Experimental investigations. .J.
gen. Psychol., 1941, 25, 187-237.

Utsugi, E. & Ohtsuki, K. A study on the human figure drawing
of children. Tohpkpp_Psychol., 1955, Folia 14:
131—146. (Psychol. Abstr-, 30: 1067).

Vane, Julia & Eisen, Virginia. The Goodenough Draw-A-Man test
and signs of maladjustment in kindergarten children.
J. clin. Psychol., 1962, 18, 276-279.

Waehner, Trude. Formal criteria for the analysis of
children's drawings. Amer. J. Orthopsychiat.,

1942, 12, 95-103.

Wallach, M. & Grahm, Ruthellen. Personality functions of
graphic constriction and expansiveness. J. Pers.,
1960, 28, 73-88.

Wapner, S., werner, H. & Comalli, P. Effects of enhancement
of head boundary on head size and-shape. Percept.
mot. Skills, 1958, 8, 319-325.

Werner, H. & wapner, S. Toward a general theory of perception.
Psychol. Rev., 1952, 59, 324-338.

Wexler, M. & Holzberg, J. A further study of the validity
of human form drawings in personality evaluation.

J._per. Tech., 16, 1952, 249-251.

45

Whitmyre, J. The significance of artistic excellence in the

judgment of adjustment inferred from human figure

drawings. J. consult. Psychol., 1953, 17, 421—424.

Wiggenhorn, A. An investigation of changes in human figure

Witk in I

Witkin,

drawing as a function of changes in the "self
concept." Provo. Pap., 1957, 1(4), 15—37.

H., Lewis, H., Hertzman, M., Machover, Karen, Meissner,
Pearl & wapner, S. Personality throughpperception.
New York: Harper & Bros., 1954.

H., Dyk, R., Faterson, H., Goodenough, D. & Karp, S.
Psychological differentiation. New York: John Wiley

& Sons, 1962.

.‘x
'1

0,.

"7'11? llllllfilllll W