A NORMATIVE STUDY OF SEXUAL
DIFFERENTIATION IN D-A-P DRAWINGS

OF CHILDREN

TI’nesls Ior II1e Degree oI M. A.
MICHIGAN STATE UNIVERSITY

Cheryl J. Normington
19760

A NORMATIVE STUDY OF SEXUAL DIFFERENTIATION
IN D-A-P DRAWINGS OF CHILDREN

BY

Cheryl J. Normington

A THESIS

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

MASTER OF ARTS

Department of Psychology

1960

ABSTRACT

A NORMATIVE STUDY OF SEXUAL DIFFERENTIATION
IN D-A-P DRAWINGS OF CHILDREN

by Cheryl J. Normington

In a normative study, three aspects of children's figure
drawings were studied. These were: (1) deve10pment of sexual
differentiation; (2) the relationship between the increasing ability
to differentiate, and increasing mental age; and (3) male and female
emphasis of the figures. Data is reported for 312 seven-through
twelve-year-old children. At each age level, there were approxi-
mately 50 children, with the proportions of boys and girls being
similar.

The pairs of drawings were scored for the degree of sexual
differentiation present, as measured by Haworth's Sexual Differen-
tiation Scale for the Draw-A-Person test. This test consists of
four levels of differentiation, which are:

1. Figures nearly the same, no apparent sex.

2. Similar body configuration and facial features, minimal

differentiation.

3. One figure more clearly differentiated as to sex.-

4. Each figure well differentiated as to sex.

Sexual differentiation, as measured by this scale, was found

to improve as a function of age, with the growth curve for both

Cheryl J. Normington

boys and girls being moderately smooth. Girls tended to show
somewhat more differentiation of the figures than did boys.

The higher levels of sexual differentiation on the scale do not
seem merely to reflect increased mental age, for when chronological
age was held constant by correlating level of differentiation and I. Q.
scores at each of four age levels, the correlations were low and
non- significant.

Male or female emphasis was determined by noting sex of
the first drawn figure, sex of the larger figure, and sex of the more
clearly differentiated figure at level three. Relationships between
these measures of emphasis were considered.

Boys and girls were found to differ significantly in their
treatment of figures at level three, with girls showing greater
tendency to emphasize their own sex figure more (or to show equal
emphasis of the figures). Both boys and girls tended to draw their
own sex figure first, and their own sex figure larger, with girls
showing a greater tendency to do so than boys, except at the older
age levels, where boys showed equal, or greater tendency than girls
to draw their own sex figure first.

A significant relationship between the measures, sex of the
first drawn figure, and sex of the larger figure, was found to exist,
whereas no significant relationship was found between the treatment
of the figures at level three and either of the measures, sex of the
first drawn figure, or sex of the larger figure.

Level three merits further investigation, especially in light
of the significant sex differences in treatment of the figures at this
level, and the independence of this measure from the measures of

sex of the first drawn figure and sex of the larger figure.

Cheryl J. Normington

Cultural factors which may be pertinent to the understanding
of the sex differences existing both in ability to differentiate, and
in proportions emphasizing their own sex figure were discussed.

The scale utilized in the study was found to be appropriate
for use with children's drawings as a measure of the development

of sexual differentiation.

.1 ,4 Q i .
Approved LO//cz,j_r1 “ ‘— A/fl‘iww ;‘ r'—- (‘2:
Majdr Professor

Date 1:),/':4/,—-,,,.,._[1,}... Q' {'1 //' (r, if)

ACKNOWLEDGMENTS

Grateful acknowledgment is made to Dr. Mary Haworth
for her sincere and active interest in this research, whose
suggestions contributed immeasurably to the completion of
this thesis, and. to Dr. Albert Rabin, Dr. Joseph Reyher, and
Dr. Charles Hanley, for their helpful guidance and suggestions.

The cooperation of Dr. Henry Kennedy, and Miss Bock,
Mrs. French, Mrs. Guilford, and Mr. Sonders, in providing
facilities and time for this research is also gratefully
appreciated.

ii

TABLE OF CONTENTS

Page
I. INTRODUCTION ....................... l
A. The concept of identification . . . . . . . ..... 2

B. The application of drawings in the assessment of
identification and Sexual Differentiation ...... 5
11. METHOD . . . . . .'. .................... 16

A.Subjects................ 17
B. Data collected concerning the subjects. ...... 18
1.Age. ......... l8
2. Parental occupation . . . . . . . . . . . . . 19
3.I.Q.scores................. 21
4. Art instruction ..... . ...... . . . . 21
C.Testadministration................. 22
D. The scale being used and scoring of the drawings . 22
E. Reliability of the scale .......... . . . . . 23

III.RESULTS ....... .. ..... ............ 24

A. Levels of differentiation . . . . . . . . . . . . . . 24

B. Relationship between level of differentiation and
mentalage..................... 28

C. Male or female emphasis of the figures . . . . . . 30

IV. DISCUSSION . ..... ' . . . ......... . . . . . 42
V. SUMMARY AND CONCLUSIONS .............. 46
BIBLIOGRAPHY . ......... . . . . . . . . . . . . . . 48
APPENDICES ...... ....... 52

iii

LIST OF TABLES
TABLE _ Page
1. Subjects by school and grade . . .......... . . 17
II. Nmnber of children at each age . . . . . . . . . . . . 18
III. Occupational groupings . . . . . . . . . . . . ..... 20
IV. Mean I.Q. scores for nine- through twelve-year-olds. 21

V. Degree of relationship between level of differentiation
andI.Q.score,byages................ 29

VI. Treatment of figures at level three (boys) . ..... . 31
VII. Treatment of figures at level three (girls) ..... . . 32

VIII. Comparison between boys and girls as to their treat-
ment of figures at level three . . . . . . . . . . . . . 32

IX. Proportion of boys and girls drawing their own sex
figurefirst,ateachage................ 33

X. Proportion of boys and girls drawing their own sex
figurelarger,ateachage............... 34

XI. Patterns between sex of the larger figure and sex of
thefirstdra.wnfigure................. 37

XII. Patterns between sex of the larger figure and treat-
ment of figures at level three . . . . . . . . . . . . . 38

XIII. Patterns between sex of the first'drawn figure and
treatment of figures at level three . . . . . . . . . . 39

XIV. Patterns among treatment of figures at level three,

sex of the first drawn figure, and sex of the larger
figure ...... 40

iv

LIST OF FIGURES

FIGURE

1. Mean level of differentiation .

II. Proportion of children at each of the four levels of
differentiation, by age . . . .

III. Sex of the more (or equally) emphasized figure at
level three, sex of the first drawn figure, and sex of
thelargerfigure(boys). . . . . . . . . . . . . . . .

IV. Sex of the more (or equally) emphasized figure at
level three, sex of the first drawn figure, and sex of
the larger figure (girls) . ......... . . .

Page

25

36

36

APPENDICES

,Page

A. Sexual differentiation scale for the Draw-A-Person test

(forusewithchildren)................... 52
B. A comparison of the boys drawing their own sex figure

first atthe varying age levels . . . . . . . . . . . . . . . 55
C. A'test of independence of sex of first drawn figure and

sexoflargerfigure(boys).—.. . . . . . . . . . . . . . . 55
D. A test of independence of sex of first drawn figure and

sex oflarger figure (girls) . . . . . . . . . . . . . . . . 55
E. Figure more emphasized at level three by pilot study

boys 0 O 0 O O O O O O O O O I O O I O O O O O O O O O O O O 56
F. Figure more emphasized at level three by pilot study

girls 0 I O I O O O O O O O O O O O. O O O O O O O O O O O O O 56

G. A further breakdown of figure more emphasized at level
threebypilotstudyboys.................. 57

H. A further breakdown of figure more emphasized at level
threebypilotstudygirls................. 57

I. INTRODUCTION

Differential treatment given to the male and female figures
in the Draw-A-Person test has been the source of extensive
research. One approach to the exploring of the meaning of the
differential treatment has involved testing the hypotheses that the
sex of the first drawn figures, and the sex of the larger figure may
provide clues to sexual identification and sexual inversion.

A newer approach concerns the degree of differentiation,
or lack of it, between the male and female figures, and the extent
to which both figures are adequate representations of their sexes.
A progressive development of the ability to increasingly differentiate
the essential features of maleness and femaleness is theoretically
expected in the drawings of children, and drawings in which the two
sexes are clearly and correctly differentiated is theoretically
expected to indicate an individual whose sexual identification is well
established.

The development of sexual differentiation in seven- through
I twelve-year-old children, as it is revealed in their human figure
drawings, is being traced in the present study. The question of the
degree of relationship between increasing ability to differentiate
between Sexes, and increasing mental age is being explored. Male
or female emphasis of the figures will also be determined by noting
sex of the first drawn figure, sex of theilarger figure, and sex of
the more differentiated figure within a given level of sexual dif—
ferentiation. These measures will be related to children's deveIOp-

ment of sexual differentiation.

As an introduction to the present study, literature concerning
the meaning of identification, and the use of drawings in the assess-

ment of identification and sexual differentiation will be reviewed.

A. The Concept of Identification

Since the time of Freud (1922) the concept of parental identifi-
cation has been liberally used as an explanatory principle in
interpreting general personality development, and sex apprOpriate
behavior. Freud describes the process of identification in the
following way: For both boys and girls, the first object is the
mother, who provides both nourishment and pleasure to the child.
However, at about the age of two or three, the child discovers his
genital zone as a source of pleasure, and ati‘this time the boy desires
his mother and wishes to get rid of his father, who is his rival for
the mother's love. The girl, on the other hand, takes her father
as her love object, after she has given up her wish for a penis and
substituted in its place a wish for a child from him.

The threat of castration brings about, in the boy, repression
of his desire for his mother, and a strengthening of his identification
with the father. Identification of the girl with her mother takes
place more gradually. Her original identification with the mother,
who was her first love object, is Strengthened as she becomes a
”little woman'' in an attempt to gain her father's love.

Since the time of Freud, the concept of identification has been
in a prominent position on the stage of personality theorizing, and

has an important place in other psychological theories (Cave and

Rausch, 1952; Lazowick, 1955; Martin, 1954; Mowrer, 1953;
Sanford, 1955; Stoke, 1950; Tolman, 1943) as well as in psycho-
analysis.

Sanford (l955).suggests that a term which can be employed in
so many different theories can hardly mean anything very precise,
and he proposes that the term be given up, or that, "what kind" of
identification be specified.

Lynn (1959) has attempted to comply with Sanford's latter
suggestion, rather than throwing out the term ”identification"
altogether. He feels that such widesPread use of the term suggests
its potential utility with adequate clarification. A differentiation
is made by Lynn among the concepts of sex-role preference, sex-
role adoption, and sex-role identification.

Sex-role preference is defined by Lynn as the desire to adopt
the behavior associated with one sex or the other, or the perception
of such behavior as more desirable. It has been measured by
simply asking Ss whether they have ever desired to be of the opposite
sex, and by having children state their preference for objects, or
pictures of objects, which are characteristic of one sex or the other
(Benjamin, 1932; Brown, 1956; Fauls and Smith, 1956; Rabban,
1950).

Sex-role adoption refers to overt behavior and is the actual
adoption of behavior characteristic of one sex or the other, not
just the desire to adopt such behavior.

Sex-role identification refers to the actual incorporation of the
role of a given sex, and to the unconscious reactions characteristic
of that role. While sex-role adoption refers to overt behavior, sex-

role identification refers to a more basic process characteristic

of a given sex, Lynn feels. In clarifying the distinction between

the latter two concepts, he points out that a person may be identified
with the Opposite sex, yet adopt much of the behavior characteristic
of his own sex for expediency. On the other hand, a woman who
occasionally ad0pts aspects characteristic of the male-sex role is
not necessarily identified with the male role.

Lynn holds that, the early closeness of the girl to the same-
sex parent gives her an initial advantage in progressing toward
appr0priate identification. This initial advantage, he suggests, is
counterbalanced by later learning experiences in a masculine-
oriented culture. The prestige and privileges afforded males but
not females, and the lack of punishment for adopting aspects of the
masculine role, Lynn feels, have a slow, corrosive, weakening
effect on the girl's feminine identification.

Boys, it is postulated, must shift from initial identification
with the mother to identification with the masculine role. However,
a conventional masculine role is clearly spelled out for the boys.
And the prestige and privileges accorded the male, the rewards
offered for adopting the masculine role, and the punishment for not
doing so, are predicted by Lynn to have a strengthening effect on
the boy's masculine identification.

The concept of identification is difficult both to define and to
measure. Attempts to define it have been indicated above, and
attempts to measure it through the projective technique of human

figure drawings will now be discussed.

B. The Application of Drawings in the Assessment of
Identification and Sexual Differentiation

Historically, the use of drawings in understanding child
development is of long standing. Early interest was directed prim-
arily at the study of the mental development of the child, and
Goodenough (1926) points out that this interest dates back to Cooke
in 1885 and Ricci in 1887.

Techniques for modern quantitative treatment of drawings .
were first applied by Florence Goodenough (1926) who utilized the
figure drawings of children in deveIOping a non-verbal measure of:
intelligence. Interestingly, she found that drawings of a male
figure presented marked sex differences of a qualitative nature.
She presented a table of characteristics most often drawn by boys
and girls, and found that some children included, in the male figure,
characteristics more typical of those drawn by the opposite sex.
Current interest in differentiation between two figures had its
origins in her work.

Recent interest in drawings has shifted to their use as pro-
jective devices. Goodenough's focusing on the human figure led to
Karen Machover's (1949) development of the Draw-A-Person test,
a technique for evaluating personality through drawings of the human
figure, the basic key to analysis being projection of the body image
in its functional implication. Rather than utilizing just the male
figure as in the Goodenough test, Machover used both the male
and female figures, and attempted to explore the significance of

differential treafinent accorded the two figures.

An early assumption of Machover (1949) was that the sex of
the first-drawn figure, and the relative size of the figure, might
provide clues to sexual identification and sexual inversion.
Machover (1949) and Schilder (1935) made explicit the hypothesis
that the sex of the first drawn figure reflects the individual's sexual
identification, sexual orientation, and sex-role preference.

Many studies have been carried out to investigate this
hypothesis, and Brown and Tolor (1957) reviewed studies by Morris
(1955), Weider and Noller (1950, 1953), Knopf and Richards (1952),
Jolles (1952), and Tolor and Tolor (1955). Of the total number of
children in these studies, 82% of the 1677 boys, and 83% of the 1700
girls drew like-sex figures first. Contradictory results, however,
were obtained among the studies.

Morris (1955) using a sample of 25 boys and 25 girls with a
median age of 13, from a seventh grade class, reported that 96%
of the boys and 76% of the girls made like sex drawings first. Five
years later 38 of the Ss, plus 12 new $3 were examined, and when
92% of the boys and 76% of the girls still drew their own sex first,
Morris concluded that differences resulting from age were negligible.

Weider and Noller's (1950) sample was composed of younger
children. One-hundred fifty-three children, ages 8-10, 73 male
and 80 female, were selected from the upper, middle, and lower
socio-economic classes in Louisville. Ninety-seven percent of the
girls and 74% of the boys drew the figure-of their own sex first,
and 80% of the girls compared to 52% of the boys drew their own sex
larger in size. These findings were interpreted by Weider and
Noller as indicating that during latency period, ”girls . . . 'identify'

more with their own sex than the boys do. . . . " (p. 324)

The interpretation given by Weider and Noller is based on the
writings of Machover (1953) who, in tracing the personality develop-
ment of the child, states that at the age of five and six the girl
draws the opposite sex first. During the latency period, though,
she ". . . retains a solid and consistent identification with the female
role, with considerable confidence and price in this role. . . . "

(p. 89) After ages 10 and 11 ". . . there is a steady deterioration
of self-esteem and efficiency. . . . " (p. 88). In adult life the
female is said to express ambivalence in sex role, shifting in the
sex drawn first. Machover suggests that in the case of boys, latency
period is accompanied by weak sexual differentiation, whereas
previously at the age of five and six the boy compares favorably with
the girl in sexual maturity. At five and six the boy frequently draws
his own sex first, she says, but during latency often draws the
opposite sex first, which, Machover believes, demonstrates a
pattern of feminine identification. Machover suggests that in our
culture the latency period is more conducive to the role played by
girls than by boys.

Continuing their work, Weider and Noller (1953) confirmed
their previous findings, using a larger sample of 438 SS, 210 male
and 228 female third grade children in the Louisville public schools.
Ninety-four per cent of the girls and 70% of the boys drew their own
sex first, and 69% of the girls compared to 38% of the boys drew
their own sex larger.

Contrary results to those of Weider and Noller were reported
by Knopf and Richards (1952) who studied very small samples of
10 male and 10 female 55 at age level six and age level eight, from

a public school in Illinois. At both age levels 80% of the boys drew

their own sex first, while at age six, 50% of the girls drew female
figures first compared to 70% at the age of eight years. Knopf and
Richards interpreted their results as suggesting girls are generally
more ambivalent in their sex-role preference than boys, although
this tendency was thought to decrease from age six to eight. Their
small sample size, however, suggests that their results may not
be highly reliable.

Of a sample of 1282 male school children between the ages
of five and 12, enrolled in public schools in Illinois, Jolles (1952)
found 85% drew the male figure first, while 80% of 1278 female 55
drew the female figure first. Although at all ages the same-sex
drawings predominated both for boys and girls, younger males
tended to draw the opposite sex figure more frequently than did the
older boys, whereas girls at the 11- and 12-year-old level displayed
a preference for drawing the opposite sex more frequently than did
the boys at the same age.

Tolor and Tolor (1955) found results consistent with those of
Weider and Noller. Eighty-two per cent of 67 boys in a Brooklyn
public school drew figures of the same sex first, compared to 91%
of the 69 girls. Age ranged from nine to 12 for the sample.

Further studies with children were carried out by Butler and
Marcuse (1959), and Bieliauskas (1960). Butler and Marcuse (1959)
obtained drawings from 810 boys and 734 girls, ages five through 18.
At the younger years of five through seven, boys showed a greater
tendency than at later ages to draw the Opposite sex figure first.

At year eight and after, the proportion of boys drawing their own
sex figure first was found to have increased, and was greater than

for girls.

Bieliauskas (1960), with a sample of 1000 children, ages four
through 14, also found that it was at the younger ages boys showed
the least tendency to draw their own sex figure first. The boys'
tendency to draw the male figure first seemed to show a continuous
increase with the advancement in age, Bieliauskas reports, while
the girls' developmental line showed some slight instability.

Although the results from the studies are somewhat contra-
dictory, it does appear that both boys and girls do prefer to draw
a person of their own sex first. This preference seems to be less
strong for boys at the younger than at the older year levels.

Findings concerning sex of the larger figure are reported only
by Weider and Noller (1950 and 1953). From their results, the
tendency to draw one’s own sex figure larger appears to be less
strong than the tendency to draw one's own sex figure first, for both
boys and girls. Boys appear to show less tendency than girls to
draw their own sex figure larger.

A newer approach to the study of differential treatment given
to figures is concerned with the differentiation, or lack of it,
between the male and female figures drawn, and the extent to which
both figures are adequate representations of the respective sexes.
Machover (1949) has emphasized the differential treatment of male
and female figures by boys and girls throughout the latency years,
and points out that ”the attitudes of the subject toward the sexes is

. . the basic determinant in the differential treatment" (p. 102).
She also states, "the particular type of treatment . . . accorded the
pair of figures drawn by a subject may be associated with the degree
of identification with the male and female figure that is character-

istic of the subject" (p. 101).

10

An early attempt to assess developmental aspects by pairs of
drawings was made by Modell (1951) who studied changes in the
D-A-P's of 28 hospitalized regressed psychotics as they improved
in adjustment. Their consecutive pairs of drawings were rated
using a scale devised by him for "body'image maturation” and
"sexual maturation. " As patients recovered from regressed states,
the "sexual maturation" and ”body image maturation" of their
D-A-P's improved significantly. Modell's "sexual maturation"
score was based on the number of characteristics which differentiate
between the male and female figures. He suggested that the loss of
sexual differentiation of the body image "is related to an almost
complete return to a pregenital stage of development" (p. 595).

Fisher and Fisher (1952) deve10ped a four-point scale for
rating the femininity of the female figure drawn'by 76 female
psychiatric patients. The score on their scale (was related to the
sexual adjustment of these patients, I

Swensen (1955) devised a scale for assessing sexual differen-
tiation, with pairs of drawings. This scale, deve10ped in order
to investigate the relationship between sexual differentiation on the
D-A-P and various behavioral characteristics, consists of nine
points, which are as follows:.

1. Little or no sexual differentiation
There is little or no difference between the two figures,
and what difference exists between them does not particu-
larly suggest sexual differentiation.

3. Poor sexual differentiation
Longer hair on the female than on the male. There
may be a slight suggestion of difference in body contour
and/or clothing.

11

5. Fair sexual differentiation
The female definitely has longer hair then the male.
The female clearly has different body contour, with
either rounded hips or breasts, or both present. 'There
may be a suggestion of difference in the clothing of the
pair.

7. Good sexual differentiation
The female has longer hair than the male. The female

has rounded body contour, the male has angular contour.
Breasts and/or rounded hips present. There is a clear
difference in clothing, with the female wearing feminine
apparel, although the apparel may be copied after that of
the male, e. g. slacks. There may be the suggestion of
differentiation in minor details, such as eyelashes or
fuller lips on the female.

9. Excellent sexual differentiation
Female hair is longer than male hair, with definite

feminine hair styling in the female. The male body has
angular contour, the female body has rounded contour with
both breasts and rounded hips present. The male wearing
clothing that is definitely masculine, the female wearing
clothing that is clearly feminine. Minor details, such as
eyes, mouth, earrings, bracelets, etc. , clearly appr0priate
for the sex of the figure on which they are drawn. (Swenson,
1955, p. 38)

Points 2, 4, 6, and 8, are to be used for rating drawings that
appear to fall midway between those listed above.

Drawings of the patients of toneterans Administration Mental
Hygiene Clinics were used in developing this scale, and significant
discriminations between the scores of an adult out-patient population
and a psychiatric in-patient group were obtained. The scoring
reliability reported by Swensen is . 84 for two independent judges.

This scale has been used by Sipprelle and Swensen (1956), with
college students in psychotherapy; by Cutter (1956) with sexual psycho-

paths, to distinguish between persons with general deficits and

12

those with specific sexual disturbances, and by Murphy (1957) with
male and female job-applicants.

Swensen and Newton (1955) applied the Swensen scale to the
study of the development of sexual differentiation in children.

Sets of drawings of 163 children (grades one to eight) were rated
for sexual differentiation using the scale, and the sex of the first
drawn figure was also noted. Swensen found sexual differentiation
to improve in a negatively accelerated curve, as a function of age.
At the younger ages, girls tended to differentiate between the sexes
significantly better than boys, but boys' sexual differentiation caught
up with that of the girls at about age 13, and beyond this age, there
was no significant difference between the sexual differentiation of
boys and girls. Although Swensen found a fairly regular increase
in sexual differentiation with increased age, there was no compar-
able increase found in per cent drawing their own sex first.

A modification of Swensen's scale was used by Limuaco (1959)
in a cross-cultural study comparing American and Filipino Children
by means of the D-A-P test. The scale was reduced from nine to
five points, so as to be more easily applied to the simple, less
complicated drawings of children. Filipino children were found to
show a greater degree of sexual differentiation than American
children. The major role of the mother in the AInerican home and
the diminishing role of the father, especially as a disciplinarian,
the permissiveness of American parents concerning their childrens'
choice of a sex role, and the overlapping of masculine and feminine
roles in the American culture, are discussed as factors influencing
the American children's lower sexual differentiation scores.

Using Goodenough scale scores as a measure of identification,

Limuaco found Filipino boys to identify more with the male figure

13

than American boys. American and Filipino girls showed no signifi-
cant difference in degree of identification with the female figure.
The dominant role of the mother in rearing children in the American
family, Limuaco suggests, facilitates the identification of the girl,
whereas it results in confusion of the boy's sex identity.

In attempting to apply the Swensen scale to drawings of nine-
and ten—year-old Israeli and American children, Haworth and Rabin1
encountered difficulty. Although many of the drawings indicated
clear awareness of sex differences, the children were not emphasiz—
ing the characteristics necessary for being classified at advanced
stages of sexual differentiation. Also, there seemed to be a need
for a stage at which one figure was portrayed with more sex appro-
priate characteristics than was the other, and this was not provided
for by the Swensen scale.

A modification of the Swensen scale was developed by Haworth, z
in an effort to provide a measure more appropriate for use with
children. The four descriptive levels of the scale are:

1. Figures nearly the same, no apparent sex.

2. Similar body configuration and facial features, minimal

differentiation. 9

3. One figure more clearly differentiated as to sex.

4. Each figure well differentiated as to sex.

This scale will be used in assessing the degree of sexual
differentiation displayed by children in their human figure drawings,

and normative data for the scale will be provided.

 

1Unpublished data.

2Mary R. Haworth, ”Sexual Differentiation Scale For the
Draw-A-Person Test" (in preparation for publication).

14

An important question in the use of the scale is the effect of
drawing skill and intelligence on the ability to differentiate the
figures. Machover (1949, 1953) points out the wide discrepancy
frequently found in the drawing of the male and female figures, and
interprets this as indicating that skill in drawing is unrelated to
sexual differentiation. Swensen (1955) found that sexual differen-
tiation did not correlate significantly with another similar visual-
motor test, the Bender-Gestalt. Murphy (1957) found no significant
sex differences in Goodenough scores, which were used as measures
of drawing ability,’ although the males and females differed signifi-
cantly on the Swensen scale of sexual differentiation. Rabin and
Limuaco (1959) also found that differences with resPect to degree
of sexual differentiation could not be accounted for by differences in
Goodenough scores, which were used as a measure of drawing
ability. Elimination of the factor of drawing ability did not affect the
significant differences found between degree of sexual differentiation
shown by American and Filipino children.

There are eleven items used in scoring the male figure in the
present scale, and only four of these are in any way similar to
Goodenough items, so it does not seem likely that a high correspond-
ence would be found between level of differentiation and mental ability,
even when using Goodenough scores as criterion. In the present
study, the relationship between mental age and degree of sexual
differentiation is being investigated through the use of an external
measure of mental ability. Mental ages, obtained from scores on a
group intelligence test, are being correlated with level of differen-
tiation, to determine the degree of relationship.

Within each of the four levels of differentiation, there may be

greater emphasis upon one figure than the other, and this emphasis

15

within levels will be of interest. The sex of the more emphasized
figure within a level, the sex of the first drawn figure, and the sex
of the larger figure will be given consideration, and possible

interpretations of the findings will be discussed.

11. METHOD

In a preliminary study, drawings of 441 first through sixth
grade normal school children were obtained from the Psychological
Clinic at Michigan State University. These drawings had been
collected from public school groups in previous studies using the
D-A-P as a group test. The original sources of the drawings were
six schools located in the communities of East Lansing, Lansing,
Hickory Corners, and St. Johns. Both the grades sampled and the
test administrator differed at the various schools. No information,
such as LO. , or social-economic level, was available for any of
the children.

These drawings were scored for level of sexual differentiation,
using Haworth's Sexual Differentiation Scale for the Draw-A-Person
Test. The findings suggested that sexual differentiation did increase
with age. Fluctuations in the growth curve occurred, however, and
it was felt that these fluctuations might be due to extraneous factors
such as differences in the socio-economic backgrounds of the children
from the various localities, or differences in the personality of the
test administrator.

It was decided that in establishing norms for the scale being
studied, a sample would be drawn from one community and tests
administered to all children by the same examiner. In selecting a
community, the need for obtaining I. Q. scores of the children, and
information concerning the occupation of the children's parents, was
kept in mind. Owos so, a community of approximately 18, 000,

located in lower Michigan, was chosen for the normative study.

16

17
A. Subjects

Sets of figure drawings were'collected from 326 first through
sixth grade children from four public schools in Owosso, Michigan.
One first, third, and fifth grade class was selected from both the
Bryant and the Emerson schools, and one class of second, fourth,
and sixth graders were selected from both the Central and the
Washington schools. At each of these schools, there were two or
more sections of children at each grade level. The selection of the
section to be visited was made by the school principal. Basis for
selection is unknown, and no attempt to control for this factor was
made, other than that an ”average” class at each grade level was

requested.

TABLE 1. Subjects by School and Grade

 

 

 

Number Number
School and Grade of Boys of Girls Total
Bryant, lst Grade 10 13 23
Emerson, lst Grade l3 13 26
Central, 2nd Grade 15 11 26
Washington, 2nd Grade 15 12 27
Bryant, 3rd Grade 11 9 20
Emerson, 3rd Grade 12 18 30
Central, 4th Grade 15 17 32
Washington, 4th Grade 16 16 32
Bryant, 5th Grade 6 23 29
Emerson, 5th Grade 18 12 30
Central, 6th Grade 11 13 24

Washington, 6th Grade 15 1 2 27
Total ' 157 169 326

 

 

18

B. Data Collected Concerning the Subjects

Date of birth and father's occupation for each child were
obtained from the school records, and 1.0. scores from a group
administered test, the California Test of Mental Maturity, were
obtained for children in the third through sixth grades. Information
concerning the amount of art instruction each group received was

provided by the classroom teachers.

1. Age

 

Date of birth for each child was obtained, and the age of each
child was found, to the nearest month. The number of children at

each age level is indicated in Table 11.

TABLE II. Number of Children at Each Age

 

II

 

 

Number Number
Age in Years and Months of Boys of Girls Total
5-6 to 6-5 2 4 6
6-6 to 7-5 22 28 ' 50
7-6 to 8-5 29 22 51
8-6 to 9-5 25 26 51
9-6 to 10-5 29 35 64
10-6 to 11-5 20 29 49
11-6 to 12-5 24 23 47
12-6 to 13-5 6 2 8

Total 157 169 326

 

19

Because of the small number of children at years six and
thirteen, only the drawings of seven through twelve year olds are
being included in the present study. This lowers the total to 312,
the number of boys to 149, and the number of girls to 163.

There are approximately 50 children at each of the age levels

being studied, with the distribution of boys and girls being fairly even.

2 . Parental Occupation

 

The occupations of the children's fathers were obtained from
the school records and the occupations were classified, using the
groupings of the U. S. Census of Population for 1950. Because
parental occupations for 41% of the six grade children either were not
reported in the school records, or were not reported in sufficient
detail to be classified, information concerning parental occupations
of 6th grade children could not be used. The proportions for grades
one through five, the proportions for these grades combined, and the
national norms, are given in Table III.

In sampling, the four schools were paired, with two schools
being sampled from at each grade level. The schools were paired
in the way that they were, with the Bryant and Emerson schools being
combined, and Central and Washington schools being combined, in an
attempt to maintain a similar distribution of occupational groupings
at each grade level. This attempt was fairly successful.

In comparison to the national norms of 1950, the three top levels
of the present sample of parental occupations correspond closely.

The major discrepancy between the present sample and the national
norms is a shift from rural to industrial, which may reflect a present

trend in the distribution of occupations in the country during the decade.

20

TABLE III. Occupational Groupings

 

 

 

 

 

_ Grades National
lst 2nd 3rd 4th 5th Combined Norms1
1950

Professional, techni-

cal, and kindred

workers .08 . 11 .08 .05 .03 .06 .073
Managers, officials,

and proprietors,

except farm . 10 . 09 . 06 . 09 . 08 . 09 . 107
Clerical and kindred

workers, and sales

workers .20 .08 .16 .19 .16 .16 .128
Craftsmen, foremen,

operatives, and

kindred workers .52 .58 .54 .46 .58 .53 . 387
Private household

workers and service

workers .02 .04 .06 .05 .01 .04 .061
Farmers and farm

managers, laborers,

and foremen of farms .00 .02 .00 .00 .03 .01 .151
Laborers, except farm

and mine .04 .04 .00 .00 .03 .02 .081
Occupation not

reported ' .04 .04 .10 .16. .08 .09 .011

Total 1.00 1.00 1.00 1.00 1.00 1.00 1.00

 

1U. S. Census of Population, 1950, U. S. Department of Commerce
Bureau of the Census, Special Report P. E. No. 1C, Occupation by
Industry, p. IC-8.

 

21

3. I. Q. Scores

 

Shortly before the collection of the present data, the California
Test of Mental Maturity had been administered to the third through
sixth grade children attending all four schools. Scores from this
group administered intelligence test were available for 182 of the
211 children between the ages of nine through twelve. Mean I. Q. for
the group of 182 children was 108.46, with a standard deviation of
15. 7. Table IV indicates the mean I. Q. for the children at each of

four age levels, as well as the mean score for the total group.

TABLE IV. Mean I.Q. Scores for Nine- Through Twelve-Year-Olds

 

 

 

Age Number Mean I.Q.
9 years 41 107.58..
10 years 56 111.26
11 years 47 105. 36
12 years 38 109.10
Total 182 108.46
SD = 15.7

 

4. Art Instruction

 

Weekly time spent in art activities varies from one hour per
week to two-and-a-half hours per week in the classrooms of the
children in this study. All of the classrooms have a half hour of
supervision every three weeks by the visiting art supervisor, who
rotates among the various schools and grades. The classroom

teachers supplement this art instruction with varied activities.

22
C . Test Administration

The Draw-A-Person test was administered to each of the
twelve sections of children by the author. Sheets of plain white
paper (8%" x 11") and pencils were distributed, and instructions were
given to "Draw a person. " Further directions to draw the entire
person, and to draw any type of person they wished, were given to
the groups, if more clarification was needed. After the members
of a group had completed one drawing, and had labeled it number one,
the drawings were collected. In the early grades, the children were
also asked to indicate on their first drawing whether the figure was
a boy or girl, since it was not always possible to tell from the drawing
itself. Then, second sheets of paper were distributed, and the groups
were asked to draw a girl if they had previously drawn a boy, or to

draw a boy if they had drawn a girl. These drawings were collected.

_D. The Scale Being Used and Scoring
of the Drawings

After the Draw-A-Person test had been administered, the sets
of drawings were scored, using Haworth's Sexual Differentiation
Scale for the Draw—A-Person Test, as an index of sexual differentiation.
This scale consists of four descriptive levels, which are: I

1. Figures nearly the same, no apparent sex.

2. Similar body configuration and facial features, minimal

differentiation.
3. One figure more clearly differentiated as to sex.

4. Each figure well differentiated as to sex.

23

Rather than scoring by use of the descriptive categories,
types of items to be scored and criteria to be met at each level of
differentiation were established empirically. Haworth found that
an item by item tabulation revealed which items seemed to appear
earlier in the developmental process and which items characterized
the drawings which had been globally ranked at the highest level.
For example, guns, pipes and pockets were found to appear on
otherwise fairly primitive male figures, while mustaches, pocket
handkerchiefs, and coat lapels were rarely found until both figures
were drawn in clearly sex appr0priate terms. This provided a
basis for classifying the items into types. The type of items, and
nmnber of items on each figure determines the score for the figure,
and the pair of scores for the two drawings establishes the level of
differentiation. See TABLE A in the Appendix.

All pairs of drawings were scored for level of differentiation
in the way described. In addition, sex of the first drawn figure,

and sex of the larger figure were recorded for all sets of drawings.

E. Reliability of the Scale

To determine the scorer reliability for the scale being used,
100 pairs of drawings from the present sample were scored for level

1 These drawings were

of differentiation by two independent judges.
selected by choosing each third set from the total number of sets.

The drawings were arranged by grades, so that in the sample drawn,
all grades were represented in the appropriate proportion. Eighty-
eight of the 100 pairs of drawings were classified as being in the same

level by both independent judges, resulting in agreement of 88%.

 

1Dr. Mary Haworth in addition to the author.

III. RESULTS

Results concerning the following aspects of the drawings will
be discussed: (A) level of sexual differentiation, (B) correlation of
mental age with level of sexual differentiation, and (C) male or
female emphasis of the figures, as determined by'sex of the first
drawn figure, sex of the larger figure, and sex of the more

differentiated figure within a given level of differentiation.

A. Levels of Differentiation

Each set of drawings was scored for level of differentiation,
and the mean level of differentiation for each age was found, for boys,
for girls, and for the total group. Graphically, these means at each
of the age levels are pictured in FIGURE 1.

From the graph it can be seen that sexual differentiation
improves as a function of age. Slight but consistent sex differences
do occur, with girls tending to differentiate between the sexers_.better,
except at age 11.

Graphs showing the proportions of children, by age, scoring at
each of the four levels of sexual differentiation, are pictured in
FIGURE 11.

It can be seen that, at seven years, considering the total group,
the majority of the children are at level two, with some being at both
levels one and three. Level three gains, and is almost equal to level
two, at age eight. Level four is first represented at age nine; the
largest prOportion of the children, however, are still at the second

level.

24

25

FIGURE 1. Mean Level of Differentiation

Level
4. 0

 

 

 

 

 

Age 7 8 9 ‘10 11 12

Key: woo-m Total
Boys
_. _. _. Girls

 

26

FIGURE 11. Proportion of Children at Each of the Four Levels of

Differentiation, by Age.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Total Boys Girls
7Years N=50 N= 22 N= 28
100
75 — — .—
F 1r-
50.. z _
25 — I- r
0 [I II _ . A 1" -
l 2 3 4 1 2 3 4 1 2 3
8Years N=51 N= 29 N= 22
100
75.1— .p _
H r
50~ F ‘- —
25._ +- L.
0 I] II n
1 2 3 4 1 2 3 4 1 2 3
9Years N= 51 N:= 25 N=26
100
75._ — -
50 — "— — F ~ I-
25 — +— L
0 1‘1 1'! II 1 n 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

27

FIGURE 11, Continued

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Total Boys Girls
10 Years N=64 N: 29 N: 35
10 -
75. _ - r11
50!— n— - '-
K
PW
25.. _ ..
0 n n n 51 __ H
l 2 3 4 l 2 3 4 1 2 3
11 Y a N = 4 = 2 :
100 e rs 9 N o N
75 - ._ I'
50 .. .. - n
25 - .. I
0 1 1" mm I
l 2 3 4 l 2 3 4 1 2 3
12 Years N247 N=24 N:
100 I ..
75 _ __ _ p
‘F
50 ft __ 1-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2:" n n H

>21
:1

 

28

Level three is clearly the most characteristic at age ten, and
continues to be so through the years eleven and twelve. At age
eleven, level one drops out. At twelve years, level four is gaining,
but level two is still represented, and the largest pr0portion of the
children continue to be at level three.

A slight sex difference can be seen, as to pr0portions at the
variouslevels, with the girls tending to be slightly more advanced

than the boys .

B. Relationship Between Level of Differentiation
and Mental Age

It can be seen from the growth curves that as children grow
older, they tend to show increasing differentiation of their figure
drawings. An important question is the degree of relationship between
the level of sexual differentiation and intelligence. Does the advance
in ability to differentiate the figures merely reflect increasing mental
age? To investigate this, I. Q. scores for 182 nine- through 12-year-
olds were available from the California Test of Mental Maturity.

In order to hold constant chronological age, bi-serial correlations
between I. Q. scores and level of differentiation were computed at each
of the four age levels. Level of differentiation was dichotomized
between levels two and three, with simpler- drawings of levels one and
two being combined, and the more advanced drawings of levels three
and four being combined. The correlations obtained are shown in
Table V.

With chronological age held constant, the correlations between

intelligence and level of differentiation are low and non- significant.

29

TABLE V. Degree of Relationship Between Level of Differentiation
and I. Q. Score, by Ages

 

 

 

Bi-serial
Age N Correlation p
9 41 . 06 ’ N. S.
10 56 . 20 N. S.
11 47 . 08 N.S.
12 38 . 14 N. S.

Total: 182

 

To obtain a second estMate of the degree of relationship between
level of differentiation and mental age, the I. Q. scores of the 182 nine-
through 12-year-olds were converted to mental ages. A bi-serial
correlation was computed between these mental age scores, which spanned
eleven mental years, and level of differentiation. Level of differentiation
was dichotomized, with levels one and two being combined, and levels
three and four being combined. The bi-serial correlation obtained
between mental age and level of differentiation was . 261, which is sig-
nificant at the . 003 level.

A bi-serial correlation between chronological age in months, and
level of sexual differentiation was also computed. Here, the span of
chronological age was four years, in contrast to the span of eleven years
of mental age in the correlation above. For the correlation, level of
differentiation was dichotomized, as in the previous case. The bi-serial
correlation obtained was . 271, which is significant at the . 002 level.

The restriction in age range was thought to perhaps be lowering
the correlation obtained between chronological age and level of dif-
ferentiation. Information concerning the chronological age, although not

the mental age, of seven and eight year old children was available, and

30

in an effort to lessen the restriction of range in the correlation between
chronological age and level, a further bi-serial correlation was com-
puted, utilizing all of the children in the study, ages seven through
twelve. Again, level of differentiation was dichotomized, between
levels two and three. This bi-serial correlation between chronological
age and level of differentiation, when utilizing more subjects in an
effort to lessen restriction in range, was .452, in contrast to the corre-
lation of only . 261 obtained between mental age and level of differen-

tiation.
C. Male or Female Emphasis of the Figures

Within each of the four levels of differentiation it is possible
that one of the figures will have a greater number of sex appropriate
items than the other, and thus one figure may be more emphasized
within any given level. At levels one and two, both figures are still
relatively immature, and the possible discrepancy between the two
figures is not as marked as at the later levels. At level three, where
one figure is usually more clearly differentiated, several points
difference may exist between the figures. At level four, although both
figures are well differentiated as to sex, it is still possible for
figures to differ in number of items present. However, it is particularly
at level three where a marked discrepancy seems to occur. In view
of this, and also because of the small number of drawings available at
level four, it is level three which will be given particular attention in
the present study.

The drawings of 66 boys, and 86 girls could be classified as
being at the third level of sexual differentiation. At this level, one's
own sex figure may be more emphasized, one's opposite sex figure

may be more emphasized, or the figures may show equal emphasis.

31

All drawings at level three were scored as to which of these possi-
bilities had occurred, with a figure considered to show greater
emphasis if it received a larger number of points than the other
figure. In a few cases, the figures received an equal number of
points, but a "D item, " the most advanced type, was among the points
on one of the figures, whereas the other figure was without the more
advanced type of item. Although both figures had the same number.
of points, the figure with the more advanced item was considered
more emphasized.

The findings concerning treatment of figures at level three, and
the pr0portions of children drawing their own sex figure first, and
larger, will be presented. Relationships between these measures

will be discussed.

1. Treatment of Figures at Level Three

 

The proportions of the 66 boys and 86 girls emphasizing their
own sex figure, emphasizing the opposite sex figure, or emphasizing

the figures equally at level three, are shown in TABLES VI and VII.

TABLE VI. Treatznent of Figures at Level Three (Boys)

 

 

 

Own Sex Equal Opposite Sex

Age N Emphasis Emphasis Emphasis

7 3 67% 33% 0%

8 9 45 33 22

9 9 32 ll 57
10 15 13 67 20
l 1 13 39 15 46
12 17 53 6 41

 

Total 66 38% 27% 35%

 

32

TABLE VII. Treatment of Figures at Level Three (Girls)

 

 

 

 

Own Sex Equal Opposite Sex

Age N Emphasis Emphasis Emphasis

7 8 75% 12. 5% 12. 5%

8 13 46 31 23

9 10 70 20 10
10 27 67 15 18
ll 15 67 13 20
12 13 92 0 8

Total 86 69% 15% . 16%

 

Only slightly over one-third of the boys emphasized their own
sex figure, whereas two-thirds of the girls emphasized their own sex
figure. The boys showed a greater tendency than the girls both to
emphasize the opposite sex figure, which they did about as frequently
as emphasizing their own sex figure, and to treat the figures equally.

These differences between boys and girls are significant, as
indicated in TABLE VIII (Age levels were combined in computing the
chi square value. This was possible, since no significant relation-
ship was found between age and treatment of figures at level three,

for either boys or girls.)

TABLE VIII. Comparison Between Boys and Girls as to Their
Treatment of Figures at Level Three

 

 

 

Boys Girls X2 p
Own Sex Emphasis 25 59
Equal Emphasis 18 13

Opposite Sex Emphasis 23 14 14. 4 . 001

 

33

2. Proportions Drawing Their Own Sex Figure First

For each set of drawings, the sex of the first drawn figure
was noted. TABLE IX indicates the pr0portions of seven- through

twelve-year old boys and girls drawing their own sex figure first.

TABLE IX. Proportion of Boys and Girls Drawing Their Own Sex
Figure First, At Each Age

 

 

 

 

Boys Girls
Age N ‘ Own-Sex First N Own-Sex First

7 22 45% 28 68%
8 29 69 22 86
9 25 92 26 65
10 29 90 35 89
ll 20 85 29 76
12 24 87. 5 23 78

Total 149 79% 163 77%

 

Both boys and girls tend to draw their own sex figure first,
and the overall tendency seems to be of similar strength for both.
Fluctuation at the various age levels is slightly greater for the boys.
At the younger ages, boys seem to show less tendency to draw their
own sex figure first then at the older age levels. A significant
difference (p = . 001) does exist in the proportions of boys drawing
their own sex figure first at the varying age levels. (See Table B in
the Appendix.) This is not so with the girls, who show less

fluctuation.

34

3. Pr0portions Drawing Their Own Sex Figure Larger

The sex of the larger figure, as well as of the first drawn
figure, was recorded for each set of drawings. TABLE X indicates
the pr0portions of boys and girls drawing their own sex figure
larger. Two boys and nine girls drew figures of equal size, and the
drawings of these eleven children have been omitted from these and

all later tables .

TABLE X. Pr0portion of Boys and Girls Drawing Their Own Sex
Figure Larger, At Each Age

 

 

 

 

Boys ‘ Girls
Age N Own-Sex Larger N Own-Sex Larger

7 22 59% 28 64%
8 29 55 21 71
9 25 60 24 62
10 29 76 33 79
l l 20 50 28 75
12 22 68 20 60
Total 147 62 154 70

 

- Both boys and girls tend to draw their own sex figure larger,
with the tendency being slightly stronger in the case of the girls.
The proportions of children drawing their own sex figure larger are
less than the proportions drawing their own sex figure first.
Fluctuations at the various age levels are not significant for either -

boys or girls.

35

4. Relationship Between the Measures

 

In studying the relationship between the three measures,
emphasis at level three, sex drawn first, and sex drawn larger,
it was decided, for ease of presentation and statistical analysis, to
combine equal emphasis of figures at level three with emphasis upon
one's own sex figure, in order that there would be only two, rather
than three groupings. Clinically, emphasis upon the Opposite sex
figure would seem to be a more deviant re8ponse than emphasizing
one's own sex figure, or emphasizing the figures equally, and there-
fore, the dichotomy of emphasis upon the Opposite sex figure vs.
emphasis of one's own sex figure (or equal emphasis) was utilized.
Sixty-five per cent of the boys and eighty-five per cent of the girls
emphasized their own sex figure, or emphasized the figures equally.
Thirty-five per cent of the boys and sixteen per cent of the girls
emphasized the opposite sex figure.

The relationship between sex of the more (or equally) empha-
sized figure at level three, sex of the first drawn figure, and sex
of the larger figure can be seen in Figures 111 and IV. It should be
noted that at age seven, there are only three boys whose drawings
are at level three, so the results indicating that 100% of seven-year-
old boys show own sex (or equal) emphasis, is misleading.

For both boys and girls, analysis by chi square indicates
that a significant relationship does exist between sex of the first
drawn figure, and sex of the larger figure (See TABLES C and D in
the Appendix), with the degree of this relationship, as estimated by
phi coefficients, being . 26 for boys, and . 30 for girls. No significant
relationship was found between sex of the more (or equally) emphasized

figure at level three and sex of the first drawn figure, or between

36

FIGURE 111. Sex of the'More (or Equally) Emphasized Figure at
Level Three, Sex of the First Drawn Figure, and Sex
of theLarger Figure

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Boys
1009' \
75 -.
50 _
25 1.
0 m 1 L I- 'I. _1
Age 7 8 9 10 ll 12
Key: (N=66) — -— -- Own Sex (or equal) Emphasis
(N=149) Own Sex Figure ‘First
(N=l47) m Own Sex Figure Larger
FIGURE IV.
Girls
100%
75
5O .
25 L
0 _ 1 I I I
Age 7 8 9 10 11 12

Key: (N=86) — - - Own Sex (or Equal) Emphasis
(N=163) Own Sex Figure First
’IN=154) M Own Sex Figure Larger

 

37

sex of the more (or equally) emphasized figure at level three and
sex of the larger figure, for either boys or girls. In the latter tests
involving treatment of the figure at level three, the number of cases
available was, of course, smaller than the number available for the
tests of independence between sex of the first drawn figure and sex

of the larger figure.

5. Combinations of the Measures, Sex of the Larger Figure, Sex of
the First Drawn Figure, and Sex of the More (or Equally) Empha-
sized Figure at Level Three

 

 

 

Boys and girls were found to be highly similar both in the
proportions drawing their own sex figure larger, and in the pr0portions
drawing their own sex figure first. In combinations of these two

measures, they are also closely alike, as shown in TABLE XI. 1

TABLE XI. Patterns Between Sex of the Larger Figure, and Sex
of the First Drawn Figure

 

 

 

Boys Girls
N = 147 N = 154
Own Larger, Own First 55% 60%
Own Larger, Opposite First 11 10
Opposite Larger, Own First 20 17
Opposite Larger, Opposite First 14 13

 

The majority of both boys and girls draw their own sex figure
first and larger. Slightly over one-tenth draw the opposite sex figure

first and larger .

 

1The eleven children drawing figures of equal size are not
included in the table.

38

Combinations between sex of the larger figure and sex of the

more emphasized figure at level three are shown in TABLE XII.l

TABLE XII. Patterns Between Sex of the Larger Figure, and
Treatment of Figures at Level Three

 

-'m
m.—

 

 

 

Boys Girls
N = 64 N = 83
Own Larger, Own Emphasized
(or Equal Emphasis) 45% 59%
Own Larger, Opposite
Emphasized , 17 12
Opposite Larger, Own
Emphasized (or Equal Emphasis) 22 23
Opposite Larger, Opposite
Emphasized 16 6

 

Approximately half of both the girls and boys drew their own
sex figure both larger and more emphasized, with girls tending to do
so to a somewhat greater extent than boys. A small prOportion of
the girls drew the Opposite sex figure both larger and more empha-
sized, indicating that this, for girls, is a rarely occurring response.
For boys, this response occurred more frequently.

Combinations between sex of the first drawn figure and sex

of the more emphasized figure at level three are shown in Table X111. 2

 

lFive children whose drawings were at level three drew
figures of equal size, and these five cases are not included in the
table.

2As in the previous table, the five children whose drawings
were at level three, but who had drawn figures equal in size, are not
included.

39

TABLE XIII. Patterns Between Sex of the First Drawn Figure,
and Treatment of Figures at Level Three

 

 

Boys Girls

N = 64 N = 83
OwnFirst, Own Emphasized
(or Equal Emphasis) 55% 71%
OwnFirst, Opposite
Emphasized 25 11
OppositeFirst, Own Emphasized
(or Equal Emphasis) 12 11
OppositeE‘irst,
Opposite Emphasized 8 7

 

Over half of both the boys and girls draw their own sex
figure both first and more emphasized, with the tendency being
stronger in the case of girls. For both boys and girls, drawing the
opposite sex figure both first and more emphasized, appears to
occur rarely.

Slightly over half of the boys draw their own sex figure first
and larger, and the same proportion of boys draw their own sex
figure first and more (or equally) emphasized at level three. These
are the most characteristic combinations for boys.

A large proportion of girls, 71%, draw their own sex figure
first, and more (or equally) emphasized atlevel three; for girls,
this is the more characteristic combination.

It appears to be unusual for either boys or girls to draw the
opposite sex figure both first and more emphasized. It is also a
rarely occurring event for girls to draw the Opposite sex figure both

larger and more emphasized.

4O

Combinations of the three measures are shown in Table

XIV. 1

TABLE XIV. Patterns Among Treatment of Figures at Level Three,
Sex of the First Drawn Figure, and Sex of the Larger
Figure

 

 

 

Boys Girls
N = 64 N = 83
Own (or Equal) Emphasis,
Own First, Own Larger 39% 53%
Own (or Equal) Emphasis,
Own First, Opposite Larger 16 18
Own (or Equal) Emphasis,
Opposite First, Own Larger 6 6
Own (or Equal) Emphasis,
Opposite First, Opposite Larger 6 5
Opposite Emphasis, I
Own First, Own Larger 17 8
Opposite Emphasis,
Own First, Opposite Larger 8 2
Opposite Emphasis, Opposite
First, Own Larger 0 _ 4
Opposite Emphasis, Opposite .
First, Opposite Larger 8 4

 

The most characteristic grouping for both boys and girls is
own (or equal) emphasis, own first, and own larger. Girls, more
than boys, group in this way.

Second, in frequency for girls, is the grouping own (or equal)

emphasis, own first, and opposite larger. And third in frequency

 

1The five children at level three drawing figures equal in size
are not included.

41

is the grouping, opposite emphasis, own first, and own larger. For
boys, the second and third groupings of girls occur about equally in
frequency.

Seventy-one per cent of the girls and fifty-five per cent of
the boys are included within the two classifications of own (or equal)
emphasis, own first, own larger, and own (or equal) emphasis,

own first, opposite larger.

IV. DISCUSSION

The present study provides normative data concerning the
sexual differentiation of seven- through twelve-year-old children,
as measured by Haworth's Secual Differentiation Scale for the D-A-P
test. Sexual differentiation appears to increase with age, and the
growth curve for both boys and girls progresses fairly evenly.

Girls tend to differentiate slightly, but consistently, better than boys,
except at age eleven. This trend seems to be in agreement with

the previous findings of Swensen (1955) who reports that girls differen-
tiate between the sexes significantly better than boys, through about
age twelve.

The fairly even progression, and similarity between the growth
curves of the boys and girls, suggests there may be a gradual increase
in ability of boys and girls to differentiate between the sexes, during
the years seven to 12. This increasing ability is thought to reflect a
progressive awareness of sex differences. Well established sexual
identification may be implied by the obtainrnent of level four. The age
and universality of the attainment of level four needs to be investigated
through research with children older than those in the present study.

The higher levels of sexual differentiation on the scale do not
seem merely to reflect increased mental age. The correlation of
. 261 between mental age and level of differentiation was significant,
but low. In contrast, a correlation of .452 was obtained between
chronological age and level of differentiation (with restriction in range
being lessened by including seven- and eight-year—olds, as well as

nine- through twelve-year-old children). In order to hold chronological

42

43

age constant, I. Q. scores at each of the age levels where they were
available (ages nine through twelve) were correlated with level of
differentiation, and these correlations were non-significant.

The scale does appear to be a developmental index, with the
findings being applicable to children similar to those in the present
study.

Level three of the scale being used was given particular atten-
tion, because it was here that differential treatment of figures within
a given level could most readily be observed. Significant sex dif-
ferences as to the figure emphasized at level three were found, and
these sex differences were also present in the pilot study made
previously to the collection of data for the present thesis. Both in the
present and in the preliminary study1 approximately 65% of the boys,
in contrast to 85% of the girls emphasized their own sex figure, or
emphasized the figures equally. And when considering emphasis upon
one's own sex figure and equal emphasis of the figures separately, in
both studies two-thirds of the girls, in contrast to approximately
one-third of the boys, emphasized their own-sex figure.

Larger samples of drawings at level three need to be collected
and a longitudinal approach utilized so that the possibility of develop-
mental trends may be investigated. Questions such as whether most
children may emphasize first one, and then the other figure at level
three before progressing to level four, whether children will empha-
size the figures equally at some time, and whether emphasis on one's
own or Opposite sex does indeed reflect identification, need to be

investigated.

 

1This preliminary study is described on page 16, and TABLES
E through H in the Appendix indicate the treatment of the figures at
level three by children in this group.

44

Proportions of children drawing their own sex figure first
in the present sample are similar to the prOportions for the combined
sample of 1677 boys and 1700 girls from the studies reviewed by
Brown and Tolor (1957). The studies by Jolles (1952), Butler and
Marcuse (1959) and Bieliauskas (1960) give indication that at the
younger age levels, boys show less tendency than at later ages to
draw a person of their own sex first. Swensen (1955), however,
reports he found no tendency for the proportion of children drawing
their own sex figure first to increase with age, for either boys or
girls. In the present study, it was at the younger ages that boys
showed least tendency to draw their own sex figure first, and a signifi-
cant relationship was found between age, and sex Of the first drawn
figure by boys.

As was found by Weider and Noller (1950, 1953), the tendency
to draw ones own sex figure larger appears to be less strong than
the tendency to draw ones own sex figure first, for both boys and
girls, and boys seem to show less tendency than girls to draw their
own sex figure larger.

A significant relationship was found between the measures,
sex of the first drawn figure, and sex of the larger figure, for both
boys and girls, whereas no significant relationship was found between
sex of the more (or equally) emphasized figure at level three and sex
of the first drawn or larger figure, for either boys or girls.

Its independence from the other measures, and the significant
differences found between boys and girls in their tendency to emphasize
their own sex figure more (or equally) at level three, indicate that

the occurrence and meaning of this level deserve further exploration.

45

At all ages except 11, girls show more advanced differentiation
than boys, and girls also show a greater tendency to emphasize their
own sex figure more (or equally) and to draw their own sex figure
first, and larger. At the later ages, the tendency of the boys to draw
their own sex figure first seems to be equal in strength, or greater
than that of the girls. The data available indicate that girls are more
consistent in all aspects being studied.

Machover (1953) suggests that in our culture, the latency
period is more conducive to the role played by girls than by boys.
Lynn (1959) feels that the early closeness of the girl to the same-sex
.parent gives her an initial advantage in progressing toward appropriate
identification. However, he predicts that the prestige and privileges
accorded the male, the rewards Offered for ad0pting the masculine
role, and the punishment for not doing so, have a strengthening effect
on the boy's masculine identification. Limuaco (1959) suggests that
the dominant role of the mother in rearing children in the American
home and the diminishing role of the father, especially as a disciplin-
arian, may be factors which adversely influence the sexual differen-
tiation and identification of the boy.

The findings in this study are consistent with these formulations.
Definite interpretation of the sex differences found in children's
development of sexual differentiation, and interpretation of the meaning
of emphasis upon one's own sex figure at level three, and the drawing
of one's own sex figure first and larger must await further research,
however. The validity of each of these measures as indicators of

identification must be more fully investigated.

V. SUMMARY AND CONCLUSIONS

The Draw-A-Person test was administered to 326 first through
sixth grade public school children, and from this sample, normative
data concerning sexual differentiation is reported for the 312 seven-
through twelve-year-old children. At each of these age levels, there
are approximately 50 children, with the proportions of boys and girls
being similar.

The pairs of drawings were scored for the degree of sexual
differentiation present, as measured by Haworth's Sexual Differen-
tiation Scale for the Draw-A-Person test. This test consists of four
levels of differentiation, which are:

1. Figures nearly the same no apparent sex.

2. Similar body configuration and facial features, minimal

differentiation.

3. One figure more clearly differentiated as to sex.

4. Each figure well differentiated as to sex.

Sexual differentiation, as measured by this scale, was found to
improve as a function of age, with the growth curve for both boys and
girls being moderately smooth. Girls tended to show somewhat more
differentiation of the figures than did boys.

The higher levels of sexual differentiation on the scale do not
seem merely to reflect increased mental age, for when chronological
age was held constant by correlating level of differentiation and I. Q.
scores at each of four age levels, the correlations were low and
non- significant.

All pairs of drawings at the third level of differentiation were

classified as to whether the own sex figure or the opposite sex figure

46

47

showed greater emphasis. Sex of the first drawn figure, and sex of
the larger figure, were also recorded for each set of drawings.

Boys and girls were found to differ significantly in their
treatment of figures at level three, with girls showing greater
tendency to emphasize their own sex figure more. Both boys and
girls tended to draw their own sex figure first, and their own sex
figure larger, with girls showing a greater tendency to do so than
boys, except at the older age levels, where boys showed equal, or
greater tendency than girls to draw their own sex figure first.

A significant relationship between the measures, sex of the
first drawn figure, and sex of the larger figure, was found to exist,
whereas no significant relationship between the treatment of the
figures at level three and either of the measures, sex of the first
drawn figure, or sex of the larger figure, was found.

Level three merits further investigation, eSpecially in light
of the significant sex differences in treatment of the figures at this
level, and the independence of this measure from the measures of
sex of the first drawn figure and sex of the larger figure.

Cultural factors which may be pertinent to the understanding
of the sex differences existing both in ability to differentiate, and in
proportions emphasizing their own sex figure more at level three,
and drawing their own sex figure first and larger, were discussed.

The scale utilized in the study was found to be appr0priate
for use with children's drawings as a measure of the development of

s exual differ entiation .

BIBLIOGRAPHY

Benjamin, H. Age and sex differences in the toy preferences of
young children. J. genet. Psychol., 1932, 41, 417-429.

 

Bieliauskas, V. J. _ Sexual identification in children's drawings of
human figures. J. clin. Psychol., 1960, 16, No. 1, 42-44.

 

Brown, D. G. Sex-role preference in young children.
Psychol. Monogr., 1956, 70, No. 14, 1-19.

 

Brown, D. G. Masculinity-femininity development in children.
J. consult. Psychol., 1957, 21, 197-202.

 

Brown, D. G. Sex-role development in a changing culture.
Psychol. Bull. 1958, 55, 232-241.

 

Brown, D. G. and Tolor, A. Human figure drawings as indicators
of sexual identification and inversion. Perceptual motor skills,
1957, 7, 199-211. Monogr., suppl. 3. ‘

 

Butler, R. L. and Marcuse, F. L. Sex identification at different
ages using the Draw-A-Person test. J. proj. Tech. 1959,
23, 299-302.

 

Cava, E. L., and Rausch, H. L. Identification and the adolescent
boy's perception of his father. J. abnorm. soc. Psychol.,
1952, 47, 855-856.

 

Cutter, F. Sexual differentiation in figure drawings and overt
deviation. J. clin. Psychol., 1956, 12, 369-372.

 

Fauls, L. B., and Smith, W. D. Sex-role learning of five year
olds. J. genetic Psychol., 1956, 89, 105-117.

 

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

 

48

49

Fisher, S. , and Fisher, Rhoda L. Style of sexual adjustment in
disturbed women and its expression in figure drawings.
J. Psychol. , 1952, 34, 169-179.

 

Frank, L. K. Projective methods for the study of personality.
J. Psychol., 1939, 8, 389-413.

 

Freud, S. Introductory lectures on psychoanalysis.
London: Allen and Unwin, 1922.

 

Goodenough, Florence L. The measurement of intelligence by drawings.
Yonkers-on-Hudson, N. Y.: World Book Co. , 1926.

 

J olles, I. A Study of the validity of some hypotheses for the qualitative
interpretation of the H-T-P for children of elementary school
age: 1. Sexual identification. J. clin. Psychol. , 1952, 8,
113- 118.

 

Knopf, I. J., and Richards, T. W. The child's differentiation of sex
as reflected in drawings of the human figure. J. genet. Psychol. ,
1952, 81, 99-112.

 

Lazowick, L. M. On the nature of identification. J. abnorm. soc.
Psychol., 1955, 51, 175-183.

 

Limuaco, Josefina A. A comparison of American and Filipino children
by means of the Draw-A-Person test. M. A. thesis, Michigan
State University, 1959.

Lynn, D. B. A note on sex differences in the deveIOpment of masculine
and feminine identification. Psychol. Rev. 1959, 66, 126-135.

 

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

 

Martin, W. E. Learning theory and identification: III. The develop-
ment of values in children. J. genet. Psychol. , 1954, 84,
211-217. ‘

 

Modell, A. H. Changes in human figure drawings by patients who
recover from regressed states. Amer. J. Orthopsychiat.,
1951, 21, 584-596.

 

50

Mowrer, O. H. Learning theory and personality dynamics.
New York: Ronald, 1950.

 

Murphy, Mary M. Sexual differentiation of male and female job
applicants on the Draw-A-Person test. J. clin. Psychol. ,
1957, 13, 87-88. T

 

Rabban, M. Sex- role identification in young children in two diverse
social groups. Genet. Psychol. Monogr., 1950, 42, 81-158.

 

Rabin, A. I. , and Limuaco, Josefina A. Sexual differentiation of
American and Filipino children as reflected in the Draw-A-
Person test. J. soc. Psychol. , 1959, 50, 207-211.

 

Sanford, N. The dynamics of identification. Psychol. Rev., 1955,
62, 106-118.

 

Schilder, P. The image and appearance of the human body.
New York: International University Press, 1935,

 

Sipprelle, C. N., and Swensen, C. H. Relationship of sexual adjust-
ment to'certain sexual characteristics of human figure
drawings. J. consult. Psychol., 1956, 20, 197-198.

 

Stoke, S. M. An inquiry into the concept of identification.
J. genet. Psychol., 1950, 67, 163-189.

 

Swensen, C. H. Sexual differentiation of the Draw-A-Person test.
J. clin. Psychol., 1955, 11, 37-40.

 

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

 

Swensen, C. H. and Newton, K. R. The development of sexual
differentiation on the Draw-A-Person test. J. clin. Psychol. ,
1955, 11, 417-419.

 

Tolman, E. C. Identification and the postwar world. J. abnorm.
soc. Psychol., 1943, 38, 141-148.

 

 

Tolor, A. , and Tolor, B. Judgment of children's popularity from
their human figure drawings. J. proj. Tech., 1955, 19,
170-176.

 

51

U. S. Census of P0pulation, 1950. U. S. Department of Commerce,
Bureau of the Census, Special Report P. E. No. IC,
Occupation of Industry.

Walker, H. M., and Lev, J. Statistical inference, New York:
Henry Holt and Company, Inc. , 1953.

 

Weider, A., and Noller, P. A. Objective Studies of children's
drawings of human figures. 1. Sex awareness and socio-
economic level. J. clin. Psychol., 1950, 6, 319-325.

 

Wdider, A., and Noller, P. A. Objective studies of children's
drawings of human figures. 11. Sex, age, intelligence.
J. clin. Psychol., 1953, 9, 20-23.

 

APPENDIX

éanngenoaofinwwhomauzaga aopwnwmIIWHfiImHgQH

 

 

 

 

 

 

 

30053930005393 gaamuumagou: Eansufismagfls
gong goapasouo I .
32.2. A8300 2 30$ a 25 22 03... .800 0 on 0H
000.6 00305003 :23 .208 90308 -
00 E: 080% 3003 agar
830600 0880-580 .20 008 m - m 080 0o 003 0 61mm 0003 so 022 00::
300.000 H.” Q0>0V 09000.8 0.00033 no "3009042 .13 .. m
g is .. a i. 000 on 8 83030000003 303 0503 080 .s
gonad .33
002 0.333%
“03.800002 $03 1 m
003 08370800 .805 08 m .. m
0.6.03 093 .8003 3000 38 8000 0 .. m I I
00>00H0 0092 "0030000000. 13 .. an .030 0.503 23 no 80.: a 80 0.0.03 .82
004...“? Sun no 0903940303. 0.033 m .. s
000.00 0000303 no 00003.8 03039.0 no 033$ N .. in A030”: m no 000a no.“ nowuelmmasoo 03.3
0.230 8.3: Sauoneam 033.0% no 0.0.0308 .n .. in 0.503 009 0803 O elm MO 55088 00:
0300. .Ho 0pp0§m 0.33002 m .. m
6.88. .8 one 6.883 son 0.06: *0 - m an .. a m - m 8.0803
no. g 0A 1 m max 080 op 00 000.350.0930 .3830 0.88 0.53m 25 .m
*0 .. m 0002.. 0 mm
*H 1 m
o - m 0003 0 so. 03000808 on ass 00000 0 go 063
003 003 A008 8V has p.300 0 .. 0
9030 .. 00.03 a .. N 5303000033 H0833
0 0|.me o - m 04.. $33000 1800 000 0300000008 Noon 03050 .0
A. 0.300 on
05000.” 0:0 093030.59 000800.83 0.00 080.3 00083 050.: a no 0 M2.
07.3“..." 3.30 no 003.30 0.8.300 00.90am upon no mmm no 00.: m. 4013. 05.0: .00:
033.50. 33.50 a - a 5808.5» 330 0 0806.83
0.346 0.50m A .. O
g 0 .. 0 EA 800 000% on N05.00 009 3.900s 00.83pm .H
50H 0.30m 0603” 0.3: 00.300 figs 03096.6

 

 

 

 

 

022.80 .0 .08:
A0833". no? 8: no.3 8.30 08890-88 3.. 8.... 080 8303008003 1080

$4 Blah—<4.

 

 

.Ab099 909 :90 0000090.. 09990 0 0099 909909 003 0 009900009909 N.0.9V 900390099 39
0.0500 000 A000 .0000 .0980 009000-0090 .905 0009008 99 000900.00 00.8.9900 9308 0.89900 9000903
7900009 099 000900 03090 0099 0000999000 00 0099 909909 .0090

90080 0900 000 .8000 0c 00000 000 0090000 .8000 009000 .99 0900.5 .0080 .8 00099

909009 3099 90900099090 09 990099990 003990300 09 99 .003 909 09099090090909 900009

90900009009 90 0.9990 9009900 099 09 005 2000090 999000000900 900099 0050 990090 909909090009 909009
.000009 .8 008 no 090009 00909900 9900 “03900.8 000 0.89.98 00900 09.0.03

 

 

.A0099008 000909 0 .013 0.800 000 000009 0050 000 00000 09 .0900 03.000 .900 0.00:

"0099009999090 0090003 03099 0.90: .

.A00909
9.26 0090000900 00 00000090 0.900 9090000900 900 . 6.3 0099 000500-09 38009908 00 900: .9090: 009990
.000090 0099 000 80090
909009098 99 090000 .000 62.80.0003. 000000 6090090 90 :0? 0089 00 000009.009 99 990000
0090099 0.999099 30 00:09 00990 9002 200909 0099 .000 ..90903 009990.. 99 00930930000 0.90990 90: 0900099

 

 

.090 30090090 0009 4900-9000 £0039: .00000 00909900 609000
99 0900.6 .0000 0090000 0089 000009008 .8 09.9999 090090 0000 0.80 00 0000 0.9.900 0900 30.80090

 

 

 

 

”w "0099009999090 0990000 03099 0.90309 .
.0903 00 0099 0000000099 0903 09009000 99 99.00 090309 990090 .9090: 009.992 “9 0999.
.9909 00 9000099 99 «090.099 09.0309 00 900093099 0903 09 90.0: .0999 0mm090-wm0m
.0903 00 3099 90909.9 90 «090099 0.90309 00 9.30 9000099 09009 99 0.90309 990090 09000 no 099.9.
.0903 no 900090 000 0.90309 00 9000099 9909 99 90 90.0903 0099 900009 9909 090309 99
0.8099 0000 8. 0900.8 9 050 .0900 09 090009 0000 09 0900 .99 0.8099 000090 9900.8 000 on 090009 09m
.3009 2.00 0000 .99 0.8099 000090 090000 900 on .0090... - 0000.9 ”0 0999
"09.93000 .00-.9 0090099 0.09 099 0003909 3 099 90 03909 09 03099 09990090 90090000 00900.2 .

.3099 n 000 000

«03099 c 039 .03099 9 059 “030.9909 00 00059999090 09 00000 9003 09 030 00090990 09000 0539033 099 09 m .9960 9.0099 099 3H .

 

“9099.9.
6099.9

"0099.9.

"0099.9

099099 909 03099 0 .mMm 90 05390903 98.60 009390990 N03099 o 99 00900009300 09 .903 03099 9 90 900.9

.009909900909990 90 9050.9 009390900 09 09000 099 00 009000 0090009009900 09000.9 .

09909990 099 09 009090000 090099 9000 00 9000099 03099 90 909300 099 90.000 .

 

0900.0. 30990990090993 90.880 099 0099000 909 0009900999

Hairs—0m

 

 

: H259

 

 

v“...

 

 

"a“...

.m..... . a”... can.

0 ‘m~—»~~c~~v

4——~. -¢.~.-~ -‘

. ~ 3....
«‘0; 3 ”A ;.\‘.\ . .. 0 x4.“
h. . Av. A.» \A , ..
In. 1..., . .. a Na” \1 ~
M. 2 r E _,
xi .\
d. a -\
,( K - , m .353
-\ . . \‘w \u
. a . ..\

. 1.... \i
u . w m
. -2 -- x/. . a .
.k N _ a.
I fl : . h

55

TABLE B. A Comparison of the Boys Drawing Their Own Sex
Figure First at the Varying Age Levels

 

 

 

Own Opposite

Age First First Xz p

7 10 12

8 20 9

9 23 2
10 26 3
11 17 3
12 21 3 22. 3 . 001

 

TABLE C. A Test of Independence of Sex of First Drawn Figure
and Sex of Larger Figure (Boys)

 

 

Own Opposite
Larger Larger X‘2 p
Own Sex Figure
Drawn First 81 29
Opposite Sex
Figure Drawn
First 16 21 10. 1 .005**

 

>1< * Yates correction applied.

TABLE D. A Test of Independence of Sex of First Drawn Figure
and Sex of Larger Figure (Girls)

 

 

 

Own Opposite
Larger Larger X2 p
Own Sex Figure
Drawn First 92 27
Opposite Sex
Figure Drawn
First 15 20 13.6 .001>l<=:<

 

>I==i=Yates correction applied.

56

TABLE E. Figure More Emphasized at Level Three by Pilot

 

 

 

 

Study Boys
Own (or Equal) Opposite
Grade N Emphasis Emphasis
1 1 0% 100%
2 4 75 2.5
3 7 100 O
4 19 53 47
5 12 50 50
6 11 64 36
Total 54 61% 39%

 

TABLE F. Figure More Emphasized at Level Three by Pilot

Study Girls

 

 

 

Own (or Equal) Opposite
Grade N Emphasis Emphasis
1 2 50% 50%
2 4 75 25
3 4 75 25
4 22 91 9
5 6 83 17
6 14 93 7
Total 52 86% 14%

 

57

TABLE G. A Further Breakdown of Figure More Emphasized at
Level Three by Pilot Study Boys

 

 

 

 

Own Opposite Figures Equally
Grade N Emphasis Emphasis Emphasized

1 1 0% 100% 0%
2 4 25 25 50
3 7 57 0 43
4 19 37 47 16
5 12 25 50 25
6 11 55 36 9

54 39% 39% 22%

 

TABLE H. A Further Breakdown of Figure More Emphasized at
Level Three by Pilot Study Girls

 

 

 

Own Opposite Figures Equally
Grade N Emphasis Emphasis Emphasized

1 z 50% 50% 0%

2 4 50 25 25

3 4 75 25 0

4 22 68 9 23

5 6 83 17 0

6 14 7 1 . 5 7 21 . 5

 

52 69% 14% 17%

 

EX

1,1
111

1 11:11: (:1

1‘.
11'!

F111”
1’ k.

 

 

 

 

 

”71111111111111; [1111111117111111111'ES