Research in Developmental Disabilities, Vol. 17, No. I, pp.

1-13, 1996
Copyright 1995 Elsevier Science Ltd
Pergamon Printed in the USA. All rights reserved
0891-4222/96 $15.00 + .00

0891-4222(95)00032-1

The Draw-a-Person Task in Persons

With Mental Retardation-
What Does It Measure?

Elisabeth Dykens

Yale Child Study Center

As a widely used, easy-to-administer, and nonthreatening task, the Draw-a-

Person (DAP) holds particular promise as a nonverbal index of intelligence in
persons with mental retardation. The DAPs of 108 adolescents and adults with
mental retardation were reliably scored using Naglieri's cognitive and emotional
disturbance scoring systems. Contrary to expectations, visual-motor skills
emerged as the best predictor of DAP cognitive scores. Intelligence was correlat-
ed with DAP scores, but had considerably less predictive value than visual-
motor skills. DAP emotional indicator scores were only modestly associated with
social adaptation; even less support was found linking DAP emotional indicators
to psychopathology. Findings point to considerable caution in using the DAP as
an index of intelligence or as a screen f o r adjustment problems or specific psy-
chopathology. Although related to intelligence, this popular task seems predomi-
nantly to measure visual-motor development in adults with mental retardation.

The Draw-a-Person (DAP) task is often used in routine educational and psycho-
logical assessments of children and adults with various learning, behavioral,
and developmental problems. Indeed, well over half of practicing school and

This research was supported, in part, by NICHD Grant #730Z.

The author thanks Robert Hodapp, Ph.D., Donald Cohen, M.D., and Sara Sparrow, Ph.D. for
their helpful comments on an earlier draft of this manuscript; Ms. Elizabeth Hamrick for her assis-
tance in coding; and David Evans, Ph.D., for his statistical consultation.
Requests for reprints should be sent to Elisabeth Dykens, UCLA, Neuropsychiatric Institute,
760 Westwood Plaza, Los Angeles, CA 90024-1759.
2 E. Dykens

clinical psychologists use the DAP and other projective drawings in their stan-
dard test batteries (Piotrowski, 1984; Prout, 1983). Although the DAP is often
used to assess individuals with mental retardation or other "special" needs,
research to date has focused almost exclusively on the DAPs of typically devel-
oping children. It thus remains unclear to what extent the DAP measures similar
cognitive or emotional features in persons with and without mental retardation.
There are many features of the DAP that make it a particularly attractive
assessment tool for persons with mental retardation. It is a nonverbal, nonin-
trusive, easy-to-administer task with inherent appeal to most children and to
many adults. Further, while research generally supports the DAP as a nonver-
bal measure of ability (e.g., Naglieri, 1988), the DAP may be most accurate
when it is used in lower-functioning persons (Sattler, 1988; Scott, 1981).
Specifically, the sensitivity of the task may be optimal in children in the 5- to
12-year age range, or in persons with mental ages in this range (Sattler, 1988).
For these reasons, the DAP seems to hold particular promise for persons with
mental retardation.
Indeed, research to date supports the DAP as a nonverbal estimate of intelli-
gence in persons with mental retardation. Early studies using the Goodenough
(1926) system and Harris (1963) revision of these criteria found correlations
between DAP IQs and WAIS, WISC, or Stanford-Binet IQs in children and
adults with mental retardation (Gayton, Bassett, & Bishop, 1970; Gunzberg,
1955; Kay, 1980; Silverstein, 1966; Wells & Pedrini, 1971). Stronger correla-
tions were typically seen between DAP scores and performance IQs as
opposed to verbal IQs.
Although encouraging, these early studies evaluated drawings with scoring
criteria and norms that are now considered outdated. To address this concern,
Naglieri (1988) developed reliable, up-to-date scoring criteria that are normed
on a large, representative sample of subjects.
Naglieri (1988) asserts that his newer DAP criteria "may be a useful test of
nonverbal ability in exceptional populations, particularly the . . . mentally
retarded" (p. 3). Research has yet to determine exactly how the new criteria
and norms fare among persons with mental retardation. It remains unclear, for
example, whether drawings can be reliably scored in this population and to
what extent they serve as nonverbal indicators of intelligence. Practitioners are
thus caught in a bind: they should probably not rely on outdated DAP norms,
yet research has still not tested relations between the new norms and intelli-
gence in persons with mental retardation.
In addition to being used as a nonverbal estimate of intelligence, the DAP is
often used to assess personality features and emotional problems in children
and adults. Although distorted or omitted features in the DAP do not seem to
be valid indicators of specific emotional or psychiatric problems (e.g., Kahill,
1984; Martin, 1983), these unusual features may collectively indicate overall
levels of adjustment (Kahill, 1984; Naglieri, McNeish & Bardos, 1991). In this
vein, Naglieri and his colleagues (1991) have developed reliable and standard-
 The Draw-a-Person Task 3

ized scoring criteria that distinguish normal from disturbed populations. Their
criteria are designed to screen children and adolescents for gross levels of
adjustment and to identify subjects in need of further evaluation.
It remains unclear whether the DAP can serve as a useful screening tool for
global adjustment or adaptation in persons with mental retardation. An early
study found no relations between drawing quality, Goodenough DAP scores,
and "behavioral adjustment ratings" in persons with mild mental retardation
and borderline intelligence (Taylor, 1966). There may be distinctive drawing
features in persons with mental retardation and schizophrenia compared to
schizophrenics of average intelligence (Kay, 1980.) Data are limited, however,
that support or refute the DAP as an indicator of global adjustment or malad-
justment in persons with mental retardation.
This study has three aims: (a) to identify salient features of the DAP in
adults with mental retardation, including the reliability in scoring these draw-
ings using the Naglieri systems (1988, 1991); (b) to evaluate relations between
the Naglieri (1988) DAP scoring criteria and intelligence; and (c) to explore
relations between the Naglieri et al. (1991) DAP emotional disturbance crite-
ria, and adaptation and psychopathology in persons with mental retardation.

METHOD
Subjects
Subjects included 108 adolescents and adults (68 males, 40 females) with
mental retardation. Participants ranged in age from 12 years, 10 months to 51
years, 4 months, with a mean age of 26 years, 8 months (SD = 7.79). As mea-
sured by the Kaufman Brief Intelligence Test (K-BIT; Kaufman & Kaufman,
1990), the IQs of this group ranged from 40 to 70, with a mean IQ of 52
(SD = 9.33), and a mean K-BIT age-equivalent score of 7 years, 3 months
(SD = 1.34). While subjects' chronological ages exceeded Naglieri's (1988)
recommended age limit of the DAP task (17 years), subjects' mental ages were
well within the developmental range of this task.
Previous WISC-R or WAIS-R Full Scale IQ scores were obtained on 41
subjects through their educational or medical records. Their mean Wechsler-
based IQ was 57 (SD = 11.57), and their mean K-BIT IQ was 52 (SD = 10.53).
For these 41 subjects, the correlation between their K-BIT and Wechsler-based
IQs was .83, commensurate with the correlations between these measures
reported in the K-BIT manual. This correlation, consistent with correlations
between most IQ tests, suggests that the K-BIT is an adequate research esti-
mate of intelligence in this population.
As reported by caregivers, 55% of the subjects had no known causes for their
developmental delay. The remaining 45% were reported by their caregivers to
have a variety of etiologic conditions, including: Prader-Willi syndrome (16%),
Down syndrome (12%), fragile-X syndrome (10%), and Smith-Magenis
4 E. Dykens

syndrome (7%). Medical verifications were not made of caregiver reports of

known or unknown etiology. Preliminary analyses in groups with known and
unknown etiologies showed similar relations between DAP scores and depen-
dent measures. As such, subjects were combined into one "mixed" etiologic
group in subsequent data analyses. Participants were recruited from several
educational, recreational, and vocational programs in four states in the New
England area (Massachussetts, Connecticut, New York, and Pennsylvania).
Subjects with the four genetic syndromes previously mentioned were recruited
for the present study from our ongoing research on behavioral phenotypes
(Dykens, 1995). Remaining subjects were specifically recruited for this study.
All subjects, as well as their primary caregivers, gave consent to participate in
the project. Forty-six percent (46%) of the sample lived in group homes, 13%
lived in a residential care facility, and 41% resided at home.

Procedures and Measures

Subjects. Subjects were individually administered a test battery by the primary
investigator or a trained research assistant. The battery took from 30 to 60 min
to complete; it consisted of the Kaufman Brief Intelligence Test (K-BIT;
Kaufman & Kaufman, 1990); the Developmental Test of Visual-Motor
Integration (VMI; Beery, 1989); and the Draw-a-Person (DAP) task as de-
scribed by Koppitz (1968) and Naglieri (1988).
Naglieri's (1988) DAP instructions were modified slightly to optimize com-
pliance and to accommodate the range of intellectual functioning in the sample.
Specifically, subjects were given a pencil and a sheet of 8 1/2 x 11 in. paper and
told, "On this paper I would like you to draw a picture of a WHOLE person.
Make the very best person you can. Take your time and work carefully." The first
sentence of these instructions is from Koppitz (1968), the next two sentences are
from Naglieri (1988). Instructions were repeated as necessary to facilitate both
task compliance and comprehension. To further optimize performance in this
developmentally delayed sample, subjects were not given Naglieri's (1988) rec-
ommended 5-min time limit, nor were they told "I'11 tell you when to stop."
Upon completion of their drawing, subjects were asked three questions
from Buck's (1964) house-tree-person postdrawing interview: "Is your draw-
ing of a person a man (boy) or a woman (girl)?"; "Is the person in your draw-
ing doing anything?"; and "Is this a drawing of someone that you know?"
These types of postdrawing inquiries are often used to embellish DAP data.
Unlike Naglieri's (1988) recommended guidelines, subjects were not for-
mally instructed to draw themselves or an opposite-sex person. This was done
to avoid issues of task refusal, disinterest, or diminished motivation - - features
routinely encountered in our clinical use of the DAP in this population. Sub-
jects were, however, given the option of drawing a picture of themselves. If
they indicated interest, subjects were administered the self-drawing in a similar
manner as the DAP.
 The Draw-a-Person Task 5

All drawings were independently scored by the primary investigator and a

trained student assistant using Naglieri's (1988) cognitive criteria. In this sys-
tem, 14 features are rated on scales that range from 0 up to a maximum of 3,
4, 5, or 7 points. Raw scores were converted to standard scores and age-
equivalent scores using tables for individual drawings supplied in the Naglieri
(1988) manual. DAP age-equivalent scores were used in data analyses as they
reflect subjects' level of drawing development instead of their performance
relative to others.
Drawings were also independently scored using the Naglieri et al. (1991)
emotional disturbance criteria. In this system, drawings are scored for the pres-
ence or absence of 55 criteria. Emotional disturbance raw scores were used in
all data analyses, or the total number of indicators per drawing. Raw scores
were not converted to T scores (Naglieri et al., 1991) as T scores are based on
three as opposed to one or two drawings.

Caregivers. Parents or primary caregivers were asked to complete two mea-

sures: the Child Behavior Checklist (CBCL; Achenbach, 1991); and the
Vineland Adaptive Behavior Scales (Sparrow, Balla, & Cicchetti, 1984).
CBCLs and Vinelands were typically completed at the same time. The CBCL
was used to assess psychopathology and was scored for internalizing prob-
lems (withdrawn, somatic complaints, and anxious/depressed) and for exter-
nalizing problems (social problems, thought problems, attention problems,
delinquent behavior, and aggressive behavior). Raw scores were used in all
data analyses.
Vinelands were individually administered in the semistructured interview
format described by Sparrow et al. (1984) and were used to examine levels of
adjustment and social adaptation. The Vineland offers standard and age-equiv-
alent scores in three domains: communication, daily living skills, and social-
ization. The optional maladaptive behavior domain of the Vineland was not
administered. Consistent with other measures, Vineland age-equivalent scores
were used in all data analyses, reflecting subjects' developmental levels
instead of their performance relative to "normal" individuals.

RESULTS
DAP Reliability and Features
lnterrater reliability. Although high interrater reliability was found in the
Naglieri (1988, 1991) normative samples, interrelater reliability was established
in this study as it examined a "special" as opposed to "normal" population. For
the cognitive scores, the correlation between the two raters was .98 for the DAP
and .96 for the optional self-drawing. Correlations for the 14 individual cognitive
scoring items ranged from .76 to 1.00. Using the emotional disturbance scoring
criteria, the correlation between the two raters was .96 for the DAP and .95 for
6 E. Dykens

the self-drawing. The range in individual emotional indicator items was .66 to
1.00. All correlations were in the excellent-to-good range (Cicchetti, 1984).

Person versus self-drawings. When given the choice, 78% of the sample opted
to draw a picture of themselves. These subjects did not differ in age, sex, IQ,
visual-motor scores, or adaptive or maladaptive behavior from those subjects
(22%) who did not want to produce a second drawing.
Mean age-equivalent scores were compared across the DAP and self-draw-
ings in a paired t-test. This proved nonsignificant, with a mean DAP age-
equivalent score of 6 years (SD = 1.62) and a mean self-drawing score of 6
years, 2 months (SD = 1.74). The correlation between the DAP and self-draw-
ing cognitive scores was .92. Comparisons also proved nonsignificant between
mean emotional indicator scores on the DAP (M = 4.88; SD = 2.77) and self-
drawings (M = 4.35; SD = 2.39). The correlation between DAP and self-draw-
ing emotional indicator scores was .89.
Given similarities in DAP and self-drawings, only the DAP was used in
subsequent data analyses. This made optimal use of subjects without apparent
compromise to findings. To check this assumption, separate analyses were per-
formed using mean scores based on both drawings; these analyses showed
results similar to those of analyses using the DAP only.

DAP features. In the cognitive scoring of the DAP, subjects earned the most
points for the degree of completion of the head and eyes and for the attach-
ment of various body parts. They received the fewest points for clothing, neck,
and fingers. Table 1 lists DAP raw scores, or the number of points earned for
each of the 14 scoring system items, as well as the mean score for each item.
The majority of subjects (94%) showed one or more emotional indicators.
The mean number of indicators was 4.88 (SD = 2.77), with a range of 0 to 11
indicators. Table 2 shows the percentage of subjects showing the number of
indicators in this range.
Thirty-nine of the 55 emotional disturbance indicators (71%) were relative-
ly infrequent, i.e., they occurred in less than 10% of the drawings. The remain-
ing 16 indicators (29%) were reasonably prevalent, seen in 10% to 42% of the
sample. Seven of these more common indicators were various types of omis-
sions (e.g., no fingers, no feet). Table 3 summarizes the numbers and relative
frequencies of the 16 reasonably prevalent emotional indicators.
No sex differences were found in either the cognitive or emotional DAP
scores. Furthermore, drawing features did not differ based on any of the sub-
ject's responses to the postdrawing inquiries. Specifically, no DAP differences
were found in whether subjects drew the same sex (73% of the sample) or
opposite sex (27%), a person that they knew (39%) or did not know (61%), or
a person showing action (34%) or no action (66%). There were also no differ-
ences across these responses in IQ, Vineland, or CBCL scores.
 The Draw-a-Person Task 7

TABLE 1
Raw Scores and Means for the 14 DAP Cognitive Scoring Criteria

Drawing Variables Raw Score Mean Score

Attachment (scored 0 to 5) 2.88 .58

Head (0 to 3) 1.59 .53
Eyes (0 to 5) 2.31 .46
Mouth (0 to 4) 1.78 .44
Arms (0 to 5) 2.21 .44
Legs (0 to 4) 1.68 .42
Nose (0 to 4) 1.62 .40
Trunk (0 to 4) 1.46 .37
Hair (0 to 4) 1.37 .34
Feet (0 to 5) 1.59 .32
Ears (0 to 5) 1.17 .23
Fingers (0 to 7) 1.57 .22
Neck (0 to 4) .85 .21
Clothing (0 to 5) .61 .12

Relationship of DAP to Other Domains

A correlation matrix was generated that examined relationships between
DAP age-equivalent scores, DAP emotional disturbance raw scores, and the
following domains: intelligence (using the vocabulary and matrices age-equiva-
lent scores from the K-BIT); visual-motor integration (using the age-equivalent
score from the VMI); adaptive behavior (using the Vineland's overall Adaptive
Behavior Composite age-equivalent score); and psychopathology (using the
CBCL's total raw score). Correlations are summarized in Table 4.
Differences between correlations were calculated using the t-test for nonin-
dependent correlations (Edwards, 1973). In the cognitive area, the correlation
between the DAP and the VMI was significantly higher than the correlation
between the DAP and the K-BIT's vocabulary domain, t(103) = 2.56, p < .05,
and the K- BIT's matrices domain, t(103) = 2.01, p < .05. DAP emotional
indicators correlated more highly with the Vineland than the CBCL, yet the
difference between these two correlations was nonsignificant, t(101) = 1.29.

TABLE 2
The Number and Proportion of Subjects Showing
Emotional Indicators

Number of Emotional Indicators N %

0-1 11 10
2-3 22 21
4-5 28 27
6-7 26 25
8-9 10 10
10-11 7 7
8 E. Dykens

TABLE 3
Number and Percentage of Subjects Showing the 16
Relatively Common DAP Emotional Indicators a

Emotional Indicators N %

Tall figure 45 41.7

No fingers 45 41.7
Big figure 42 38.9
Vacant eyes 40 37.0
No feet 39 36.1
No hair 26 24.1
No torso 26 24.1
Hands cut off 24 22.2
No arms 20 18.5
No legs 18 16.7
Top placement 17 15.7
Short figure 17 15.7
Rotated page 16 14.8
Transparencies 15 13.9
No nose 13 12.0
Teeth showing 11 10.0

aThe remaining 39 indicators were infrequent, seen in

less than 10% of the sample.

Relation of DAP to Intelligence. Significant relationships between DAP cogni-

tive scores, IQ, and the VMI were further examined in two regressions. In the
first regression, a stepwise model was used with the DAP cognitive score as
the dependent variable and the VMI and K-BIT vocabulary and matrices
domains as independent variables. The VMI emerged as the only significant
predictor, accounting for 33% of the variance, F(1, 106) = 52.41, p < .001. As
the VMI is significantly correlated with the DAP (r = .57) and with the K-BIT
IQ (r = .50), a hierarchical regression was conducted that accounted for these

TABLE 4
Correlations Between DAP Cognitive and Emotional Features and
the K-BIT, VMI, Vineland and CBCL

DAP

Cognitive Age- Emotional

Equivalent Score Disturbance Score

K-BIT Vocabulary .34*** -. 13

K-BIT Matrices .41 *** -.08
Visual-Motor Integration .57"** -. 14
Vineland Composite .27* .36***
Child Behavior Checklist -. 13 .24*

*p < .05, <**p < .01, ***p < .001.

 The Draw-a-Person Task 9

relationships. An analysis was performed with the DAP cognitive score as the
dependent variable and with the K-BIT IQ forced into the equation prior to the
VMI. The proportion of variance was then calculated that was uniquely
accounted for by the VMI (Pedhazur, 1982). In this model, 10% of the vari-
ance was uniquely associated with intelligence, F(1, 106) = 11.53, p < .001.
After accounting for this effect, 23% of the variance was uniquely associated
with the VMI, F(2, 105) = 26.38, p < .001.

Relationship of DAP to adaptation and psychopathology. Significant relation-

ships between DAP emotional scores, the ineland, and CBCL were also
examined in two regression equations. In the stepwise model, DAP emotional
indicators served as the dependent variable and the composite CBCL and
Vineland scores were independent variables. The Vineland was the only signifi-
cant predictor, F(1,106) = 12.56, p < .001, accounting for 10% of the variance.
As the CBCL was related to the Vineland (r = -.50) and to DAP emotional
indicator scores (r = .24), a hierarchical regression was constructed that further
examined these relations. The DAP emotional score served as the predictor,
and the CBCL was forced into the equation prior to the Vineland. The propor-
tion of variance was then calculated that was uniquely associated with each
independent variable. The CBCL accounted for only 3% of the variance,
F(1,106) = 2.77, p < .05. After accounting for this modest effect, the Vineland
accounted for just 7% of the variance, F(2, 105) = 6.55,p < .01.
Additional correlations were performed to explore possible relations
between DAP emotional indicators and two types of emotional problems.
Based on a review of the drawing literature, the 16 more common emotional
indicators in Table 2 were divided into two categories: those indicators hypoth-
esized in the literature to symbolize internalizing difficulties and those drawing
features typically associated with externalizing problems. Internalizing drawing
indicators were: short figure, transparencies, top placement, vacant eyes, and all
omissions. These drawing features are hypothesized to represent internal states
such as insecurity, anxiety, depression, somatic concerns, poor self-concept, and
self-absorption (see Ogdon, 1979 for a review). Externalizing drawing indica-
tors were: tall and big figure, rotated page, and teeth showing - - all thought to
symbolize aggression and acting-out behaviors (Ogdon, 1979).
Drawing indicators of internalizing difficulties were modestly associated
with the CBCL internalizing domain, r(89) = .23, p < .05, and not with the
externalizing domain (r = .16). Conversely, drawing indicators of externalizing
problems were modestly correlated with the CBCL externalizing domain,
r(89) = .21, p < .05, but not with the internalizing domain (r =. 11).

DISCUSSION
This study is the first using Naglieri's cognitive (1988) and emotional dis-
turbance (1991) scoring systems to describe features of the DAP in persons
10 E. Dykens

with mental retardation. Specifically, this work identified DAP interrater relia-
bility and the feasibility of a shorter version of this measure. Findings also
shed new light on relationships between the DAP and intelligence, adaptation,
and psychopathology.
Regarding the first area of work, interrater reliability was high in this study.
These findings, consistent with Naglieri's normative data (1988, 1991), support
the use of both scoring systems in this "special" population. Preliminary support
was also found for a shorter version of the DAP in adults with mental retardation.
In particular, to optimize performance and avoid task refusal or disinterest,
subjects were instructed to produce only one drawing (Koppitz, 1968); they were
then given the option of drawing themselves. Approximately 22% of the sample
refused this option, clarifying a clinical need to modify the task on an as-needed
basis in some adults with mental retardation. Further, high correlations between
the DAP and self-drawing in both the cognitive and emotional domains (.92 and
.89, respectively) suggest very little variability across drawings. In contrast, the
DAP normative sample showed an average correlation of .62 between the man
and self-drawings (Naglieri, 1988), suggesting a need for multiple drawings in
this group. In adults with mental retardation, then, one drawing as opposed to
three may suffice for certain clinical or research purposes.
The second aim of the study was to examine relations between the DAP and
intelligence. Similar to earlier work, this study found significant correlations
between the newer Naglieri (1988) DAP scoring system and intelligence, par-
ticularly the K-BIT's nonverbal matrices domain. Contrary to expectations,
however, visual-motor development emerged as a more powerful predictor of
DAP cognitive scores.
In particular, significantly higher correlations were found between the DAP
and the VMI than between the DAP and the matrices and verbal domains of
the K-BIT. While intelligence had some predictive relation to the DAP,
accounting for 10% of the variance, visual-motor development accounted for a
much larger proportion of the variance (23%). In this particular population,
then, the DAP appears more closely associated with visual-motor development
than with intelligence.
This study is the first to examine relations between visual-motor development
and the DAPs of subjects with mental retardation. Since their earliest use, DAPs
have been cast as indices of intelligence and "mental maturity" (Goodenough,
1926), not of visual-motor development. The process of drawing a person, how-
ever, necessarily involves visual-motor functioning. Most scoring systems as-
sume a certain level of visual-motor maturation, at the very least that the subject
has progressed beyond the "scribbling" stage of drawing development (see
Thomas & Silk, 1990). While scoring systems require visual-motor maturation,
they do not usually reflect visual-motor sophistication, instead emphasizing the
inclusion of basic, even crudely drawn, parts (e.g., Naglieri, 1988). That scoring
systems typically downplay motor precision makes it even more striking that
visual-motor skills were so strongly associated with DAP scores in this study.
 The Draw-a-Person Task 11

Relationships between the DAP and visual-motor development have also

not been widely studied in normally developing children. Historically, many
developmentalists assumed that normal children often show synchrony
between visual-motor, cognitive, and other behavioral domains (e.g., Piaget,
1952). This line of thinking suggests that if the DAPs of normal children are
associated with intelligence, then drawing performance and intelligence are
also roughly commensurate with visual-motor functioning.
Although the extent to which normal children show this synchrony is open
to debate (e.g., Gardner, 1993; Humphreys, 1979), many people with mental
retardation and other learning difficulties are vulnerable to uneven develop-
ment, showing relative strengths and weaknesses across various domains (e.g.,
Bellugi, Wang, & Jernigan, 1994; Dykens, Hodapp, Walsh, & Nash, 1992). An
extreme example of this unevenness is seen in Williams syndrome, a develop-
mental disorder associated with significant visual-motor weaknesses and lin-
guistic strengths (e.g., Bellugi et al., 1994). Although rare, this example shows
how the visual-motor functioning of persons with mental retardation need not
be commensurate with their intelligence or other behavioral domains.
Clearly, then, practitioners should not automatically assume that the DAP
predominantly assesses intelligence in persons with mental retardation. While
the DAP is associated with intelligence, stronger relationships were found
between the DAP and visual-motor development. Even within the population
of persons with mental retardation, levels of visual-motor functioning may or
may not be comparable to levels of intellectual functioning. As such, the DAP
should not be administered in isolation as it is unclear what this task best mea-
sures for any particular individual with mental retardation. Instead, the DAP
seems well suited to supplement or clarify data from other cognitive and visu-
al-motor measures within a comprehensive test battery.
The third goal of this study was to examine relations between the DAP
emotional scoring system and global adjustment and maladjustment. Only
modest support was found for DAP emotional indicators as indices of adapta-
tion or adjustment in adults with mental retardation. Specifically, while the
Vineland emerged as the sole predictor of DAP emotional indicator scores, it
accounted for just 7% of the variance. More promising links of the DAP to
adaptation might be found in selected groups of subjects with and without
well-delineated emotional or adjustment problems.
Even less support was found for relations between DAP emotional indicators
and psychopathology. The CBCL had little predictive relation to DAP emotion-
al indicator scores, accounting for just 3% of the variance. Furthermore, draw-
ing features believed to symbolize internalizing versus externalizing problems
were only modestly correlated with the CBCL's internalizing and externalizing
domains. These data provide little support for linking specific drawing features
to specific kinds of psychopathology in persons with mental retardation, despite
early efforts to do so (e.g., De Martino, 1954; Kay, 1978).
12 E. Dykens

In addition to using the DAP as an index of cognitive and emotional func-

tioning, some practitioners expand DAP data with postdrawing interviews. The
three postdrawing questions administered in this study do not appear to be
diagnostically relevant. Specifically, most subjects drew a person of the same
sex that they did not know, and they did not depict the figure in action.
Subjects who drew the opposite sex, figures in action, or a familiar person
were not different from remaining subjects in drawing scores, IQ, visual-motor
skills, adaptive functioning, or psychopathology. While these three questions
were not particularly informative, clinicians may still opt to use postdrawing
inquiries that help identify subjects' thoughts or feelings about themselves and
others (see Karp, 1990).
Although it remains unknown how applicable findings from this study are
to young children with mental retardation, this work sheds new light on the
DAP in adolescents and adults with mental retardation. The DAP emerged as a
reliably scored task, and preliminary support was found for an adapted, shorter
version of the DAP task in this population. Findings point to considerable cau-
tion in using the DAP as a screen for global adjustment concerns or for spe-
cific psychopathology or as a nonverbal index of intelligence. Although related
to intelligence, this popular task seems predominantly to measure visual-motor
development in adults with mental retardation.

REFERENCES
Acbenbach, T. M. (1991). Manual for the child behavior checklist/4-18 and 1991 profile. Burlington,
VT: University of Vermont Department of Psychiatry.
Beery, K. E. (1989). The developmental test of visual-motor integration (3rd rcv.). Cleveland, OH:
Modem Curriculum.
Bellugi, U., Wang, P., & Jernigan, T. L. (1994). Williams syndrome: An unusual neuropsychologi-
cal profile. In S. H. Browman & J. Grafram (Eds.), Atypical cognitive deficits in developmental
disorders (pp. 23-56). Hillsdale, NJ: Erlbaum.
Buck, J. N. (1964). The house-tree-person (H-T-P) manual supplement: Administration and inter-
pretation of the H-T-P test. Los Angeles, CA: Western Psychological Services.
Cicchctti, D. C. (1984). On a model for assessing the security of infantile attachment: Issues of
observer reliability and validity. The Behavioral and Brain Sciences, 7, 149-151.
De Martino, M. (1954). Human figure drawings by mentally retarded males. Journal of Clinical
Psychology, 10, 241-244.
Dykens, E. M. (1995). Measuring behavioral phenotypes: Provocations from the "new genetics."
American Journal on Mental retardation, 99, 522-532.
Dykens, E. M., Hodapp, R. M., Walsh, K., & Nash, L. J. (1992). Profiles, trajectories and corre-
lates of intelligence in Prader-Willi syndrome. Journal of the American Academy of Child and
Adolescent Psychiatry, 31, 1125-1130.
Edwards, A. L. (1973). Statistical methods (3rd ed.). New York: Holt, Rinehart & Winston.
Gardner, H. (1993). Multiple intelligences: The theory in practice. New York: Basic Books.
Gayton, W. E, Bassett, J. E., & Bishop, J. S. (1970). The Harris revision of the Goodenough Draw-
a-Man Test: Suitability for a retarded population. Journal of Clinical Psychology, 26, 522-523.
Goodenough, E L. (1926). Measurement of intelligence by drawings. New York: World Book.
Gunzberg, H. C. (1955). Scope and limitations of the Goodenough drawing test method in clinical
work with mental defectives. Journal of Clinical Psychology, 11, 8-14.
 The Draw-a-Person Task 13

Harris, D. B. (1963). Children's drawings as measures of inteUectual maturity: A revision and

extension of the Goodenough Draw-a-Man test. New York: Harcourt, Brace, Jovanovich.
Humphreys, L. (1979). The construct of general intelligence. Intelligence, 3, 105-120.
Kahill, S. (1984). Human figure drawings in adults: An update of the empirical evidence. Canadian
Psychology, 25, 269-292.
Karp, S. A. (1990). Draw-a-Person questionnaire manual. Worthington, OH: International
Diagnostic Systems.
Kaufman, A. S., & Kaufman, N. L. (1990). Kaufman Brief Intelligence Test. Circle Pines, MN:
American Guidance Service.
Kay, S. R. (1978). Qualitative differences in human figure drawings according to schizophrenic
subtype. Perceptual and Motor Skills, 47, 923-932.
Kay, S. R. (1980). Progressive figure drawings in the development of mentally retarded psy-
chotics. Perceptual and Motor Skills, 50, 583-590.
Koppitz, E. M. (1968). Psychological evaluation of children's figure drawings. New York: Grune
& Stratton.
Martin, R. P. (1983). The ethical issues in the use and interpretation of the Draw-a-Person test and
other similar projective procedures. School Psychologist, 38, 6-8.
Naglieri, J. A. (1988). Draw-a-Person: A quantitative scoring system. San Antonio, TX; New
York: The Psychological Corporation.
Naglieri, J. A., McNeish, T. J., & Bardos, A. N. (1991). Draw-a-Person: Screening procedure for
emotional disturbance. Austin, TX: Pro-ed.
Ogdon, D. P. (1979). Psychodiagnostics and personality assessment: A handbook (2rid ed.). Los
Angeles, CA: Western Psychological Services.
Pedhazur, E. (1982). Multiple regression in behavioral research: Exploration and prediction. Fort
Worth, TX: Holt, Rinehart & Winston.
Piaget, J. (1952). The origins of intelligence in children. New York: International Universities Press.
Piotrowski, C. (1984). The status of projective techniques, or "wishing won't make it go away."
Journal of Clinical Psychology, 40, 1495-1502.
Prout, H. T. (1983). School psychologists and social-emotional assessment techniques: Patterns
and training in use. School Psychology Review, 12, 377-383.
Sattler, J. M. (1988). Assessment of children (3rd ed.). San Diego, CA: Jerome Sattler, Publisher.
Scott, L. H. (1981). Measuring intelligence with the Goodenough-Hams drawing test. Psychological
Bulletin, 89, 483-505.
Silverstein, A. B. (1966). Anxiety and the quality of human figure drawings. American Journal of
Mental Deficiency, 70, 607-608.
Sparrow, S. S., Balla, D. A., & Cicchetti, D. V. (1984). Vineland adaptive behavior scales: Survey
form. Circle Pines, MN: American Guidance Service.
Taylor, J. B. (1966). The use of human figure drawings with the upper level mentally retarded.
American Journal of Mental Deficiency, 71,423-426.
Thomas, G. V., & Silk, A. M. (1990). An introduction to the psychology of children's drawings.
New York: New York University Press.
Wells, D. G., & Pedrini, D. T. (1971). Relationships among Wechsler Adult Intelligence Scale,
Goodenough-Harris and Peabody Picture Vocabulary test with institutionalized retarded adults.
Perceptual and Motor Skills, 33, 227-232.