This article was downloaded by: [New York University]

On: 11 April 2015, At: 11:48

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

Journal of Projective
Techniques and Personality
Assessment
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/hzpa20

An Investigation on the
Reliability of Human Figure
Drawings
a a
James F. Guinan & John R. Hurley
a
Michigan State University , USA
Published online: 16 Nov 2010.

To cite this article: James F. Guinan & John R. Hurley (1965) An Investigation on
the Reliability of Human Figure Drawings, Journal of Projective Techniques and
Personality Assessment, 29:3, 300-304, DOI: 10.1080/0091651X.1965.10120212

To link to this article: http://dx.doi.org/10.1080/0091651X.1965.10120212

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.
 This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [New York University] at 11:48 11 April 2015
 An Investigation on the Reliability of Human Figure Drawings
JAMES AND JOHN R. HURLEY
F. GUINAN
Michigan State University
Slctnmnry: A review of thc literature showed that prior reliability studies on the DAP
have generally utilized atomistic graphic indices, while clinicians have claimed that
the use of the DAP, in practice, has been to obtain global, personalized, impressions.
Twenty Ss were twice administered the DAP with a five-week interval. Results showed
that each of 9 judges were able to match the drawings of the same S from the two
administrations at well beyond the .001 level of significance. T h e three judges with
greatest DAP experience matched the 20 sets of drawings perfectly. T h e mean dif-
ferences in correct matches for the three groups of judges (Ph.D. Clinical Psycho-
logists, graduate students in Clinical Psychology, and college freshmen) although sug-
gestive, were not significant. Results were contrasted with the findings of other
approaches discussed in relation to other studies, and a number of research implica-
Downloaded by [New York University] at 11:48 11 April 2015

t i o k were noted.

PROBLEM reliability based on repeated admin-

Sundberg’s (1961) recent survey of
clinical tools used in psychological
testing, indicated that the Draw-A-
Person Test ranks second only to the
Rorschach in frequency of usage.
Sundberg also noted that research
publications dealing with the DAP
were a sharp exception to a general
positive correlation between rank of
“frequency of usage” of a psycho-
logical test, and number of publica-
tions. T h e dearth of material on the
DAP seems particularly glaring in the
area of reliability studies.
T h e reliability of scores is basic to
the use of any test. Granick and
Scheflen (1958), however, have em-
phasized the difficulty of devising
satisfactory objective scoring pro-
cedures for projective techniques and
noted that “the reliability of scoring,
in effect, is dependent on the
reliability of judgments (p. 138) .”
T h e problem, essentially seems to
be the development of objective,
communicable scores which reflect the
integrity of “Gestalt” of the perform-
ance of the individual tested. In dis-
cussing the problems of reliability of
projective tests, Granick and Scheflen
point out that there have been three
traditional approaches to reliability:
(a) judgment reliability as measured
in terms of agreement between judges
rating the same data; (b) temporal
istrations of the same projective test to
the same individual are expected to
produce stable components (but not
identical responses) ; and (c) internal
reliability-comparison of two or more
separate parts of the same test data.
These writers conclude that an overall
statistical reliability for a projective
test is not feasible. An alternative of-
fered is to study the reliability of cer-
tain aspects of a projective device.
“The traditional concepts of reliabil-
ity need not be lost in dealing with
projective tests. What is necessary is
the adaptation of test data to avail-
able techniques without destroying
the global or clinical quality of the
test response (p. 141) .”
In introducing the DAP as a pro-
jective device, Machover (1949) states
that “structured and formal aspects of
drawings-size, line, placement, etc.-
are less subject to variability than
content aspects (p. 6) .” Such an infer-
ence suggests the reliability of specific
graphic indices, but does not neces-
sarily suggest a reliability of the pro-
jective aspects of the test itself.
There have been a number of in-
vestigations on the reliability of
specific graphic indices oE human
figure drawings. Gunderson and Leh-
ner (1949) suggest a line of research
employing reliability of ratings of
graphic traits used for diagnosis. In a
 JAMES F. GUINAN
A N D JOHN R. HURLEV
later publication (1958), the same
authors report that percent of agree-
ment between judges of graphic traits
was 90.5%. Bradshaw (1952) utilized
the test-retest method. He considered
twenty-five body areas that were
scored for 17 possible kinds of treat-
ment. Percent of agreement of two
judges was reported for each treat-
ment. Results showed that the highest
percent of agreement was in rating
the drawing as a whole. Swenson
(195i) tends to disparage the use of
the DAP as a projective tool, Specific-
Downloaded by [New York University] at 11:48 11 April 2015
ally, he criticizes the “percentage of
agreement” method used by Lehner
and Gunderson, and Bradshaw, and
argues persuasively that “percent of
agreement does not equal reliability.”
Moreover, neither Lehner and Gun-
derson, nor Bradshaw reported the
base rate of occurrence of each struc-
tural aspect observed. Thus, concludes
Swenson, “these studies do not show
reliability of the DAP (p. 463) .”
Since validity implies reliability,
some researchers have stressed validity
and thereby bypassed direct approach
to the problem of reliability. Accord-
ingly, a number of authors have point-
ed out the validity of various projec-
tive indices of the DAP. Hoyt and
Martin (1959) found placement and
size of figure drawn to be related to
manifest anxiety scores. Schmidt and
McGowan (1959) reported that figure
drawings differentiated between phy-
sically disabled and normals. Bodwin
and Bruck (1960) found a significant
correlation between self-concept as re-
flected in the DAP and as assessed by
an interview. Kamana (1960) found
similar results when measures of self-
concept obtained by a Semantic Dif-
ferential and DAP were correlated.
Goldstein and Rawn (1957) reported
that specific aggressive details were
related to other measures of aggres-
sion. There have also been reports of
negative findings when specific indices
were applied to the DAP. Stan- and
Marcuse (1959) in a replication of
Bradshaw’s study, failed to find sig-
nificance. Strumpfer and Nichols
(1962) reported that sixteen DAP
measures failed to differentiate nor-
mals, neurotics, and schizophrenics.
Sherman (1958, 1958a) and Whit-
myre (1953) have both found support
for the hypothesis that artistic ability,
and not personality projection, is the
primary factor in human figure
drawings.
There is a basic and general criti-
cism that can be leveled a t all of the
cited studies. Namely, research in the
area of evaluation of figure drawings
has tended to deal with specific vari-
ables in the drawing (shading, cen-
teredness, etc.) while use of the test in
clinical settings has been to obtain
general impressions of behavior trends
or personality organization. Arbit,
Lakin, and Mathis (1959) succinctly
point this out by suggesting that the
dimensions of the response, i.e., degree
to which a figure drawing reflects
aspects of self image and/or personal-
i t y problems, have not yet been speci-
fied. They question whether any clin-
icians rely upon specific indices and
attempt to relate them to specific con-
cepts. It seems more likely that clin-
icians view the drawings as a personal-
ized expression of an individual‘s be-
havior orientation. However, research
in the area of diagnostic evaluation
from figure drawings has tended to
deal with specified factors in the
drawings. T h e fact that many studies
show negative results may only indi-
cate that the difficulty is one of am-
biguity of diagnostic usage which has
resulted in research unrelated to the
more relevant aspects of the drawings.
Other writers, Zimmer (1956),
Rosenzweig (1960), and Laane
(1960) have all criticized studies
which have attempted to measure
specifically described or narrowly de-
fined variables in working with the
DAP. Albee and Hamlin (1949) in-
dicate that the degree to which inter-
pretation is determined by the specific
elements, rather than intuitive overall
impressions is not clear. “Presumably,
interpretation depends on specific
elements and their synthesis. Global
 302 An Investigation on the Reliability of Human Figure Drawings
judgments may or may not involve a The interval of time between admin-
synthesis of specific elements. Clin- istrations was 5 weeks. After each ad-
icians make judgments and interpre- ministration, drawings were stapled
tations with little conscious attention together and assigned a number from
to specific molecular factors of con- a table of random numbers.
stant significance, but rather as a re- Nine judges were chosen to match
sult of global impressions of unverbal- the drawings. Three were Ph.D.’s from
ized comparison of a present drawing the clinical section of the Department
with past experience of “intuitive” or
insightful impressions (Albee and of Psychology; three were graduate
students majoring in clinical psycho-
Hamlin, 1949) .” logy; and three were freshmen college
To date, no reliability study of the students who had not yet had a psy-
DAP that has utilized such global im- chology course. All judges were male.
pressions as criteria for judgments is
Downloaded by [New York University] at 11:48 11 April 2015

Drawings were separated into two

known to the authors. Presumably, groups. Each group of drawings con-
the personality of the individual has sisted 01 the twenty sets of drawings
an enduring organization which obtained from the two atlministra-
should be reflected in his human tions. T h e sets of drawings within
figure drawings. Thus, the drawings of each group were in the same initial
the same person over a period of time order for all judges: Instructions to
should reflect the more enduring, the judges were as follows: “Here
stable components or characteristics are two groups of drawings. Each
of the personality. More specifically, itgroup contains 20 sets of drawings
can be hypothesized that human figure that were obtained from the same
drawings of the same individuals twenty subjects at two different times.
should exhibit a sufficient degree of Your job will be to match the draw-
stability over a period of time to be ings that were done by each subject.”
identified as having been drawn by Judges were then given the drawings
the same person. T h e purpose of this without further directions. Time and
study was to test such an hypothesis. methods of matching were not re-
METHOD corded.
RESULTS
Ss were 20 freshmen and sopho-
mores (11 female, 9 male) from an T h e number of correct matches for
Introductory Psychology course. All each judge was recorded. The Ph.D.’s
Ss volunteered. T h e DAP was group- matched 18, 20, and 20, out of 20 re-
administered on two occasions. Pro- spectively. T h e three graduate stu-
cedure was identical in both adminis- dents matched 16, 18, and. 20; and
trations. Ss were given three sheets of the three freshmen matched 8, 10, and
Sl/, x 11 plain white paper and a No. 18. According to the binomial dis-
2 lead pencil. Instructions were simp- tribution, all nine judges matched
ly: “You will be asked to make two the drawings at far beyond the chance
drawings, and given six minutes for level (P<.OOl that more than 5 cor-
each. .. Draw a picture of a person.” rect matches could be made).
Any questions were reflected by “Just Six sets of drawings were matched
draw it as you think a person looks.” correctly by all nine judges. Of these,
Six minutes after Ss were told to start, four were drawn by male Ss. Of Ss
E instructed them: “Now take a sec- whose drawings were incorrectly
ond sheet of paper and draw a person matched, mean number of incorrect
of the opposite sex.” Six minutes later, matches for females was 1.91, for
Ss were told to stop, and to put iden- males 1.44. There were no sets of
tifying information on a third sheet. drawings incorrectly matched by more
T h e two drawings were then inserted than 4 judges. The mean number of
in the folded third sheet and collected. correct matches for the three groups
 JAMES F. GUINAN
AND JOHN R. HURLEY
combined was 16.4. For each group,
the number of correct matches was
12.0 for freshmen, 18.0 for graduate
sludents, and 19.3 for Ph.D.’s. T h e dif-
ference between means of the Ph.D.
and freshmen groups was treated
statistically, but results were not
significant.
(t=1.66, d f r 2 , .15<P<.lO).
DISCUSSION
T h e procedures used in this study
do not rely on specific indices, and
Downloaded by [New York University] at 11:48 11 April 2015
therefore the results are not neces-
sarily contradictory to the results of
former studies of DAP reliability
(Sherman, 1958; Starr and Marcuse,
1959; Swenson, 1957) which em loyed
only specific indices. Whereas fefore,
the methods of criteria of judgrrient
were in question, here the ability to
judge was measured. Whereas specific
indices have been found to be unre-
liable in these other studies, here the
ability to match drawings done by
the same individuals was found to
indicate a high reliability. However,
the conclusion reached by the authors
just cited, i.e. that the DAP is an un-
reliable test, was not supported by
this study. I n that there was not one
S whose drawings were not matched
by at least 5 out of the nine judges,
and in that judges were able to match
a n average of 16.4 drawings of 20
(P<.OOl) , the data certainly indicate
that there are reliable aspects to hu-
man figure drawings. However, the
conclusions drawn by certain authors
are in direct opposition to the results
of this study. Starr and Marcuse
(1959), and Strumpfer and Nichols
(1962) have argued that the lack of
consistency of specified aspects of hu-
man figure drawings suggests the in-
efficiency of the DAP. Swenson (1957)
concluded that the DAP as a tool of
personality assessment is of doubtful
value. These studies however have all
employed methods which measured
only parts o r aspects of the drawings.
I n this study the drawing as a whole
was judged without specific reference
to narrowly defined indices. It is,
303
therefore,. suggested that a method
which employs a global o r “intuitive”
impressions of the clinician as the
basis of judgment is more appropriate
than the atomistic methods which use
more carefully defined but perhaps
-
useless and irrelevant indices.
Of note is the markedly different
mean number of correct matches for
the three groups of judges. T h e fresh-
man judge who matched most of the
drawings is a n art student with some
experience in judging drawings. T h e
graduate student who was able to
match all of the twenty drawings has
had extensive experience with the
DAP. Hardy (1960) has suggested a
linear relationship between experi-
ence with the DAP and diagnostic
accuracy in using the DAP. T h e data
of this study seem to indicate support
for such a relationship.
I n this study the human figure
drawings themselves, rather than their
projective aspects, were the central
core of interest. T h e results would
suggest that a similar approach to the
reliability of measures of personality
variables in human figure drawings
may also prove fruitful. T h e suggested
relationship between experience and
reliability and/or validity of human
figure drawings as technique of per-
sonality assessment is obviously in
need of further verification.
REFERENCES
Albee, G . and Hamlin, R. An investigation of
of the reliability and validity of judgments
of adjustment inferred from drawings. J .
d i n . Psychol. 1949,5,389-392.
Arbit, J., Lakin, M., and Mathis, A. Clinical
psychologists diagnostic utilization of hu-
man figure drawings. J . clin. Psychol. 1959,
15, 325-327.
Bodwin, R. and Bruck, M. The adaptation
and validation of the DAP as a measure of
self concept. J . d i n . Psychol. 1960, 16, 427-
429.
Bradshaw, D. A study of group consistencies
on the DAP test in relation to personality
projection. Unpublished M. A. thesis, Cath-
olic University, 1952.
Goldstein, A. and Rawn, M. The validity of
interpretive signs of aggression in the draw-
ing of the human figure. J . d i n . Psychol.
1957, 13, 169-171.
 304 A n Znvestigation on the Reliability of Human Figure Wrawings
Granick, S. and Scheflen, N. Approaches to
reliability of projective tests with special
reference to Blacky Pictures Test. J . con-
sult. Psychol. 1958, 22, 137-141.
Gunderson, E. and Lehner, F. Reliability of a
projective test (the DAP) . Amer. Psycholo-
gist. 1949, 4 , 387.
Hardy, M. T h e interaction of patients’ I.Q.
and other factors with clinicians’ skill in
the diagnosis of human figure drawings.
Unpublished doctoral dissertation, Florida
State University, 1959.
Hoyt, T. and Martin, R. Anxiety indices in
same-sex drawings with psychotic patients
with high and low manifest anxiety scores,
J. consult. Psychol. 1959,23,448-452.
Downloaded by [New York University] at 11:48 11 April 2015
Kamana, D. An investigation on the meaning
of human figure drawings. J. clin. Psychol.
1960, 16, 429-430.
Laane, C. Clinical experience with the figure
drawing test. J. clin. exp. Psychopath. 1960,
21, 129-141.
Lehner, F. and Gunderson, E. Reliability of
graphic indices in a projective test (the
DAP). J . d i n . Psychol. 1958, 8, 125-128.
Machover, K. Personality projection in the
drawing of the human figure. Springfield,
Ill.: Charles C. Thomas, 1949.
Rosenzweig, S. T h e Rosenzweig Picture Frus-
tration Study: Children’s Form. I n Rabin
and Haworth (Eds.) , Projective techniques
with children. New York: Grune and Strat-
ton, 1960, Pp. 149-176.
Schmidt, L. and McGowan, J. Differentiation
of human figure drawings. J. consult. Psy-
chol. 1959, 23, 129-133.
Sherman, L. T h e influence of artistic quality
on judgments of patient and non-patient
status from human figure drawing. J. proj.
Tech. 1958, 22, 338-340.
Sherman, L. Sexual differences or artistic abil-
ity. .I. clin. Psychol. 1958, 14, 170-177.
Starr, S. and Marcuse, F. Reliability on the
Draw-A-Person Test. J. proj. Tech. 1959,
23, 83-86.
Strumpfer, D. and Nichols, R. A study of
some communicable measures for the
evaluation of human figure drawings. J.
proj. Tech. 1962, 26, 342-353.
Sundberg, N. T h e practice of psychological
testing in clinical services in the United
States. Amer. Psychologist. 1961, 16, 79-83.
Swenson. S. An emDirical evaluation of hu-
man figure drawings. Psychol. Bull. 1955,
54, 431-466.
Whibnyre, J. T h e significance of artistic ex-
cellence in the judgment of adjustment in-
ferred from human figure drawing. J. con-
sult. Psychol. 1953, 17, 421-424.
Zimmer, H. Validity of sentence completion
tests and human figure drawings. In D.
Brower and L. Abt (Eds.), Progresc in
clinical psychology. New York: Grune and
Stratton, 1956.
James F. Guinan
Michigan State University
E. Lansing, Mich.
Received November 27, 1964