This article was downloaded by: [Universidad Nacional Colombia]

On: 30 September 2012, At: 20:13

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

Art Therapy: Journal of the American Art Therapy

Association
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uart20

Children's Family Drawings: A Study of Attachment,

Personality, and Adjustment
a a
Limor Goldner & Miri Scharf
a
Haifa, Israel

Version of record first published: 14 Apr 2011.

To cite this article: Limor Goldner & Miri Scharf (2011): Children's Family Drawings: A Study of Attachment, Personality, and
Adjustment, Art Therapy: Journal of the American Art Therapy Association, 28:1, 11-18

To link to this article: http://dx.doi.org/10.1080/07421656.2011.557350

PLEASE SCROLL DOWN FOR ARTICLE

Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
 Art Therapy: Journal of the American Art Therapy Association, 28(1) pp. 11–18

Children’s Family Drawings: A Study of Attachment,
Personality, and Adjustment
Limor Goldner and Miri Scharf, Haifa, Israel
Abstract
This study examined the relationship between children’s at-
tachment security, as manifested in their family drawings, and
their personality and adjustment. Family drawings were col-
lected from 222 Israeli children, as well as data regarding their
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012
personality and adjustment. Each drawing was coded and clas-
sified into 1 of 4 attachment categories based on global and
individual characteristics such as completeness of figures, facial
expression, size, and degree of movement. Results showed that
drawings from securely attached children included more positive
markers of personality and reflected their superior psychosocial
functioning. Drawings from children in the ambivalent and
disorganized attachment categories reflected the children’s ad-
justment difficulties; children whose drawings were classified as
representing avoidant attachment were found to function rel-
atively well. The study demonstrated the effectiveness of using
family drawings to assess children’s attachment representations
and to identify children at risk for adjustment problems in the
school setting.
Introduction
Family drawings have a long history of clinical use be-
cause they provide a representation of a child’s view of fam-
ily dynamics and the self within the family (Leon, Wallace,
& Rudy, 2007). The nonverbal nature of drawings allows
children to express nonconscious aspects of their mental
representations that they may be unable to convey verbally
(Fury, Carlson, & Sroufe, 1997). The use of family draw-
ings generally relies on a clinical and theoretical base but not
on a solid empirical foundation. However, in recent years
children’s family drawings have become increasingly stud-
ied in empirical research (e.g., Deaver, 2009; Leon et al.,
2007; Milne, Greenway, & Best, 2005). For researchers as
well as therapists, the drawings can shed light on children’s
internal worlds and their perception of their families and
parents. For example, Piperno, Di Biasi, and Levi (2007)
found that drawings by physically and/or sexually abused
children revealed greater emotional distress than drawings
by children who had not been abused. In another study, chil-
dren whose drawings reflected higher levels of parent–child
Editor’s Note: Limor Goldner, doctoral candidate, and Miri
Scharf, PhD, are affiliated with the Department of Counsel-
ing and Human Development at the University of Haifa, Israel.
Correspondence concerning this article may be addressed to the
first author at limor.goldner@weizmann.ac.il
11
C AATA, Inc. 2011
alliances (which frequently indicate boundary dissolution)
were found to have less prosocial behavior and assertiveness
and a greater degree of social problems (Leon et al., 2007).
Scholars and researchers have used family drawings to
assess children’s attachment representations (e.g., Kaplan &
Main, 1986; Madigan, Ladd, & Goldberg, 2003; Pianta,
Longmaid, & Ferguson, 1999). The study described in this
article extends previous research by examining the relation-
ship between children’s attachment security, as manifested
in their family drawings, and their personality traits and ad-
justment. We focused on the family drawings of elementary
school children and attempted to identify the features of
drawings by children who might be at risk for adjustment
problems.
Attachment Theory
Attachment theory posits that, beginning in infancy
and continuing throughout the course of life, an individual’s
mental health and his or her capacity to form close relation-
ships are closely linked to prior relationships with attach-
ment figures that provided emotional support and protec-
tion (Belsky & Cassidy, 1994; Bretherton & Munholland,
2008). Accordingly, children’s actual experiences shape their
representational models, which subsequently serve to guide
their behavior in novel circumstances (Bowlby, 1980). Chil-
dren experiencing sensitive and responsive care will have se-
cure attachment, characterized by the ability to develop trust
in others, to find comfort with closeness, and to develop
adaptive ways of dealing with stress. Avoidant attachment is
associated with having discomfort with closeness and an in-
clination for self-reliance, whereas a child with anxious (am-
bivalent) attachment has an intense desire for closeness and
experiences constant concern about the parent’s availability
(Mikulincer & Shaver, 2006). Children with disorganized
attachment are characterized by an apparent lack or collapse
of a consistent and organized strategy for dealing with stress.
The particular forms and mixtures of disorganized behav-
iors tend to be idiosyncratic from child to child, but include
anxious, helpless, or depressed behaviors; unexpected fluctu-
ations of approach and avoidance toward the attachment fig-
ure; and other conflicted and unpredictable behaviors (Main
& Solomon, 1990).
Family Drawings for Classifying Attachment
Representations
Most known assessments that use family drawings are
not based on an empirically derived theory (Mostkoff &
 12 CHILDREN’S FAMILY DRAWINGS
Lazarus, 1983; Pianta et al., 1999). An early study by Kaplan
and Main (1986) used attachment theory to classify chil-
dren’s family drawings into four patterns of attachment. On
the basis of specific markers in the drawings and their over-
all configuration, Kaplan and Main found that drawings by
children with secure attachment were realistic; figures were
complete, grounded, and individuated (not drawn exactly
alike). There was a natural proximity among family mem-
bers and an impression of happiness in the family. Drawings
by children classified as avoidant tried to convey a positive
picture, emphasizing invulnerability and happiness. The fig-
ures’ arms were absent or drawn in a way that did not allow
holding, and there was lack of individuation of the figures
and lack of movement in the picture. Drawings by children
classified as ambivalent included figures that were extremely
large or small, and either overlapped or were separated from
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012
each other with barriers. Soft body parts and facial features
were exaggerated. Finally, drawings by children classified as
disorganized/disoriented included strange marks, threaten-
ing themes and fantasy themes, unfinished objects or figures,
and/or an impression of excessive and irrational sweetness.
Based on Kaplan and Main’s (1986) coding system,
Fury et al. (1997) developed a global approach to children’s
family drawings that considered each drawing as a whole in
an integrative manner. The eight rating scales they devel-
oped consist of two positive dimensions (vitality/creativity
and family pride/happiness) and six negative dimensions
(vulnerability, emotional distance/isolation, tension/anger,
role reversal, bizarreness/dissociation, and global pathology).
The global rating scales were correlated with children’s at-
tachment classifications as assessed in infancy. The authors’
analysis showed that even after controlling for IQ, current
life stress, and emotional functioning, the children’s early
attachment history made a significant contribution to the
prediction of negative dimensions in their drawings (Fury
et al., 1997).
In another study (Madigan et al., 2003), children whose
family drawings depicted higher levels of emotional dis-
tance, vulnerability, and parent–child role reversal were
found to have an insecure attachment history, whereas fam-
ily drawings that scored higher in family pride and lower
in global pathology were drawn by children with a secure
attachment history. Carlson, Sroufe, and Egeland (2004)
conducted a longitudinal study in which children’s family
drawings at age 8 were found to correlate with the chil-
dren’s attachment classification identified from interview
data gathered while the children were in preschool and again
at age 12.
Attachment and Personality
Personality refers to a person’s tendencies to behave,
think, and feel in certain consistent ways (Shiner & Caspi,
2003). Personality traits indicate inherited characteristics
and intrinsic maturation (McCrae et al., 2000), and re-
flect individual strengths and weaknesses. Five traits repre-
sent major domains of personality: (a) extraversion, which
refers to activity level and is expressed through sociability,
enthusiasm, assertiveness, and self-confidence; (b) agreeable-
ness, which is expressed through sensitivity toward others
and responsibility; (c) conscientiousness with respect to or-
derliness and reliability; (d) openness to experience as it
pertains to intellectual functioning, creativity, and imag-
ination; and (e) neuroticism, which refers to a tendency
to experience anxiety, depression, and emotional instability
(Measelle, John, Ablow, Cowan, & Cowan, 2005; Shiner &
Caspi, 2003). Research indicates that the characteristics of
these five traits are already present in childhood and that
personality plays an important role in an individual’s adjust-
ment capacity (Barbaranelli, Caprara, Rabasca, & Pastorelli,
2003; Measelle et al., 2005; Shiner & Caspi, 2003).
Several studies have examined the relationship between
attachment and personality using various measures among
adults. Attachment anxiety (characterizing the ambivalent
pattern) has been positively correlated with neuroticism
and negatively correlated with agreeableness (Bäckström &
Holmes, 2001; Roismann et al., 2007; Winjngaards-de Meij
et al., 2007), and attachment avoidance has been positively
correlated with neuroticism (Bäckström & Holmes, 2001;
Carver, 1997; Shaver & Brennan, 1992; Winjngaards-de
Meij et al., 2007) and negatively correlated with conscien-
tiousness (Becker, Billings, Eveleth, & Gilbert, 1997) and
with extraversion and agreeableness (Bäckström & Holmes,
2001; Becker et al., 1997; Roismann et al., 2007). Thus,
these studies found that the personality traits of neuroticism
and agreeableness in adults consistently correlated with at-
tachment styles, whereas other personality traits were less
consistently correlated with attachment styles.
Studies examining the relationship between attachment
security and personality traits among school-aged children
are fairly rare. Hagekull and Bohlin (2003) examined rela-
tionships between infant temperament, attachment security,
and the five personality traits described above. Using parent
ratings of their children’s temperament at age 20 months and
attachment to their mother at 15 months, and their moth-
ers’ and teachers’ ratings on the five personality traits, the
researchers found that attachment security predicted neu-
roticism and openness; however, there was no prediction of
agreeableness and conscientiousness.
Attachment and Children’s Adjustment
Individual differences in attachment security are ex-
pressed in emotional regulation and exploration, and have
important implications for social and personality develop-
ment as well as for adjustment (Weinfield, Sroufe, Egeland,
& Carlson, 2008). In general, research has shown that at-
tachment security facilitates resilient functioning and serves
as a buffer when coping with adversities, whereas insecure
attachment may hamper a child’s adjustment. However, not
all studies distinguish between the different types of attach-
ment insecurity.
Substantial evidence indicates that children with se-
cure attachment develop better social–emotional compe-
tence (Granot & Mayseless, 2001; Ranson & Urichuk,
2008). They have more constructive coping mechanisms,
better regulation of emotion in the classroom (Kerns, Abra-
ham, Schlegelmilch, & Morgan, 2007), and better behav-
ioral adjustment (Granot & Mayseless, 2001). Attachment
security is also correlated with social competence (Early
 GOLDNER / SCHARF
Child Care Research Network, 2006; Walls, 1998), social
support, and peer acceptance (Granot & Mayseless, 2001;
Kerns, Kelpac, & Cole, 1996). Children who have insecure
attachment have been found to experience more loneliness
(Kerns & Stevens, 1995; Kerns et al., 1996) and depression
(Graham & Easterbrooks, 2000). Two studies used Kaplan
and Main’s (1986) coding system to examine children’s be-
havioral adjustment: Pianta et al. (1999) studied the draw-
ings of preschool children and found indications of previous
and concurrent social and behavioral competence that were
independent of the children’s age, sex, socioeconomic sta-
tus, intelligence, and motor skills, and Carlson et al. (2004)
found that children’s drawings predicted the children’s social
functioning in early and late adolescence.
Insecure attachment has been associated with various
behavioral, social, and emotional problems (Weinfield et al.,
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012
2008). Sroufe (1983) suggested that both ambivalently
attached and avoidant children may behave aggressively.
Avoidant children may behave in a hostile way, as in bullying
and lying, whereas ambivalent children can be easily aroused
and may display impulsivity and attention difficulties
(Deklyen & Greenberg, 2008). Attention deficit/
hyperactivity disorder (ADHD) may be related to in-
secure internal models of attachment (Clarke, Ungerer,
Chahoud, Johnson, & Stiefel, 2002), with higher levels of
hyperactivity found in children with avoidant attachment
(Moss, Bureau, Cyr, Mongeau, & St-Laurent, 2004).
Children diagnosed with combined or predominantly
hyperactive-impulse ADHD may be more likely to have
anxious and ambivalent attachment styles than children
with predominantly inattentive type ADHD (Finzi-Dottan,
Manor, & Tyano, 2006). Ambivalent attachment also is
associated with anxiety problems and depression (Weinfield
et al., 2008).
Children with disorganized attachment styles are an es-
pecially vulnerable population. They may have greater so-
cial and behavioral difficulties in the classroom (Goldwyn,
Green, Stanley, & Smith, 2000) and may experience disso-
ciation (Main & Moran, 1996), as well as various externaliz-
ing problems such as noncompliance and tantrums (Lyons-
Ruth, 1996). A high frequency of disorganized attachment
has been found among boys with oppositional defiant dis-
order (Greenberg, Speltz, Deklyen, & Endriga, 1991).
The aim of the current study was to examine the re-
lationship between the attachment representations of el-
ementary school–aged children, as manifested in their
family drawings, and their personality traits and social and
behavioral adjustment. We hypothesized that children with
secure attachment representations would have higher lev-
els of agreeable behavior and lower levels of neuroticism,
more prosocial behaviors, fewer behavioral and social prob-
lems, a lower degree of hyperactivity, and fewer emotional
symptoms.
Method
Participants
A total of 222 children participated in the study. Chil-
dren were drawn from 19 elementary schools in northern
13
Israel that were located in both peripheral and central city
neighborhoods populated by families with lower socioeco-
nomic status. Of the total sample, 55% were boys (n = 121)
and 45% were girls (n = 101). The mean age of the children
was 9.70 years (range 8–12 years; SD = 1.14); the mean age
of their parents was 43 years (SD = 7.30). Fifty-three percent
of the children were from two-parent families, 28% had par-
ents who were divorced, and 19% were from single-parent
families. Children who were born in Israel made up 62% of
the sample and the others were immigrants (most from the
former Soviet Union). The children had an average of 2.38
siblings each (range 1–11; SD = 1.56)
Procedure
After receiving permission from the Ministry of Educa-
tion and written consent from the children and their parents,
the children completed the drawing task and the personality
questionnaire during home visits, which were conducted by
trained research assistants. The children were asked to draw
their families on a white piece of paper and were given eight
colored felt-tipped pens. After a child completed the draw-
ing, the research assistants wrote the identity of each fig-
ure depicted in the drawing and any additional information
the child described. The homeroom teachers in the partici-
pating schools completed questionnaires regarding the chil-
dren’s emotional and social adjustment.
Measures
Two trained raters coded all drawings using the sys-
tem developed by Fury et al. (1997), as well as Kaplan and
Main’s (1986) attachment classifications that focus on size,
location, degree of movement, individualized characteris-
tics, completeness of figures, facial expression, context, and
overall impression of vulnerability. A detailed manual was
adapted by further delineating specific examples to dimin-
ish ambiguity, based on the Kaplan and Main study as well
as the research conducted by Fury et al.. The training and
adaptation of the manual, including examination for inter-
rater reliability, was initially constructed based on another
sample. The interrater intraclass reliability for the separate
indicators in this study ranged from .70 to 1.00. Disagree-
ments between the coders were resolved by consensus.
Each drawing was also coded using 8 global scales rang-
ing from 1 (low level) to 7 (high level) for the following di-
mensions: (a) vitality, which refers to emotional investment
in drawing, creativity, and the richness of details in the draw-
ing; (b) pride, which refers to expressions of happiness in
the family group drawn and the degree of connectedness de-
picted among family members; (c) vulnerability as expressed
in the placement of figures and distortions in the size of body
parts; (d) isolation, which refers to a sense of emotional dis-
tance and loneliness expressed in neutral or negative affect
depicted as well as a lack of proximity between the mother
and child figures; (e) tension/anger, referring to restriction
of the figures expressed in careless appearance, lack of col-
ors, and the like; (f ) role reversal as indicated by a dispro-
portion in the size and/or the roles depicted of the figures;
(g) bizarreness, which indicates disorganization reflected in
 14 CHILDREN’S FAMILY DRAWINGS

Table 1 Profile of Drawing Dimensions According to Attachment Classifications

Secure Avoidant Ambivalent Disorganized

n = 92 n = 43 n = 68 n = 19

Dimension M M M M F Attachment η2 Attachment

Vitality 5.17 (1.19) 2.87 (1.04) 3.89 (.83) 3.88 (1.76) 36.32∗∗∗ .36
Pride 4.31 (1.12) 2.77 (.73) 3.13 (.94) 2.76 (1.25) 28.15∗∗∗ .31
Vulnerability 2.58 (1.14) 3.20 (1.42) 4.03 (1.36) 4.29 (1.36) 19.16∗∗∗ .23
Isolation 2.60 (1.30) 4.07 (1.51) 3.59 (1.52) 4.29 (1.49) 13.39∗∗∗ .17
Tension 1.70 (.86) 4.40 (1.35) 3.51 (1.53) 4.00 (2.06) 44.69∗∗∗ .41
Role reversal 1.73 (.26) 2.60 (1.73) 2.46 (1.95) 1.88 (1.76) 3.53∗ .05
Bizarreness 1.24 (.57) 1.73 (1.01) 1.86 (.82) 3.00 (1.12) 25.62∗∗∗ .29
Pathology 1.88 (.75) 3.63 (.89) 3.02 (.66) 3.65 (1.27) 53.65∗∗∗ .46
Note. Standard deviations appear in parentheses after the means.
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012

∗
p < .05, ∗∗∗ p < .001.

unusual signs and fantasy themes; and (h) global pathology,
which refers to negativity, incompleteness of the figures, and
impoverished expression in the drawing, details, and back-
ground. The interrater reliability for the global scales ranged
from .80 to .95.
In addition, drawings were used to classify children into
one of four attachment categories (Kaplan & Main, 1986):
(a) secure, (b) avoidant, (c) ambivalent, and (d) disorga-
nized, as described above. The correspondence between the
coders on the four main classifications, based on 31 sample
drawings, was 77.4%, χ 2(9) = 53.11, p < .001; κ = .68,
p < .001. Disagreements between coders were resolved by
consensus.
Homeroom teachers completed the Strengths and Diffi-
culties Questionnaire (Goodman, 1997) to measure the chil-
dren’s emotional and social adjustment. The questionnaire
consists of five subscales, each with 5 items that are rated
on a 3-point scale (1 = almost never, 2 = sometimes, 3 = al-
most always). The questionnaires generated scores for proso-
cial behaviors (e.g., “Considerate of other people’s feelings”;
α = .82), emotional symptoms (e.g., “Often complains of
headaches, stomach-aches, or sickness”; α = .73), hyperac-
tivity (e.g., “Restless, overactive”; α = .80), social problems
(e.g., “Rather solitary, prefers to be alone”; α = .74), and be-
havioral problems (e.g., “Often loses temper”; α = .70). The
Strengths and Difficulties Questionnaire has been shown to
be a useful tool in diagnosing children’s psychiatric disor-
ders; internal consistency for each of the five subscales has
been shown to be good, with a mean Cronbach alpha of .73
(Goodman, 2001).
Results
We first coded the drawings using the 8 global scales
described above and then used them to assign children to
one of the four attachment classifications. Of the 222 chil-
dren in the sample, 41.4% were classified as secure, 19.4% as
avoidant, 30.6% as ambivalent, and 8.6% as disorganized.
The distribution generally reflects the children’s status as at-
risk for adjustment difficulties.
To examine the different profiles of attachment classi-
fications we conducted a MANOVA, with the attachment
categories serving as the independent variable and the di-
mensions of the 8 global scales serving as the dependent
variable (Table 1). The MANOVA revealed a significant
main effect of attachment, F (8, 185) = 30.58, p < .001,
η2 = .57. As can be seen in Table 1, based on their draw-
ings, secure children demonstrated the highest levels of vi-
tality and pride, and the lowest levels of isolation, ten-
sion, bizarreness, and pathology. They also demonstrated
the lowest levels of vulnerability (together with the avoidant
group) and lower levels of role reversal than the ambiva-
lent group. The disorganized group exhibited the highest
levels of bizarreness and higher levels of vulnerability (to-
gether with the preoccupied group) than the secure and the
avoidant groups. The avoidant group showed higher levels
of tension than the ambivalent group.
Next we examined the relationship between the di-
mensions of the children’s drawings and their personal-
ities by computing Pearson correlations (Table 2). Most
of the correlations emerged between the positive person-
ality traits and the positive dimensions of the 8 scales.
Conscientiousness and agreeableness were positively cor-
related with vitality and were negatively correlated with
bizarreness and pathology; additionally, conscientiousness
was negatively correlated with tension. The other correla-
tions were not significant. A second MANOVA, with at-
tachment categories serving as the independent variable
and personality traits as the dependent variable, was not
significant.
To examine the relationship between children’s draw-
ings and adjustment (as reported by homeroom teachers) we
computed Pearson correlations (Table 3). Prosocial behavior
was negatively correlated with vulnerability. Hyperactivity
and behavioral problems were negatively correlated with vi-
tality and pride, and positively correlated with isolation, ten-
sion, bizarreness, and pathology. Hyperactivity was also pos-
itively correlated with vulnerability. Emotional symptoms
and social problems were not correlated with any of the di-
mensions of the scales.
 GOLDNER / SCHARF 15

Table 2 Pearson Correlations Matrix for Drawing Dimensions and Personality Characteristics

Neuroticism Extroversion Openness Conscientiousness Agreeableness

∗
Vitality −.08 −.02 −.03 .15 .23∗∗
Pride −.10 .02 −.02 .11 .10
Vulnerability .05 .03 .03 −.06 −.00
Isolation .08 .03 .03 −.06 −.00
Tension .13† −.01 .10 −.16∗ −.12†
Role reversal .11 .07 .02 −.01 −.04
Bizarreness .02 .12† −.02 −.15∗ −.18∗∗
Pathology −.02 −.00 .06 −.24∗∗∗ −.25∗∗∗
Note. N = 222 for all analyses.
†
p < .10, ∗ p < .05, ∗∗ p < .01, ∗∗∗ p < .001.
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012

To examine the differences among the attachment
groups, we conducted a final MANOVA with attachment
categories serving as the independent variable and the ad-
justment scales serving as the dependent variable, con-
trolling for gender (Table 4). The MANOVA revealed a
significant main effect of attachment group, F (5, 198) =
3.32, p < .01, η2 = .08. As can be seen in Table 4, secure and
avoidant children exhibited higher levels of prosocial behav-
ior than the ambivalent group, secure children showed lower
levels of conduct problems than the ambivalent and the dis-
organized groups, and avoidant children showed lower levels
of conduct problems than the disorganized group.
Discussion
The study supported the findings of previous research
and demonstrated the usefulness of analyzing children’s fam-
ily drawings using a theory-based global coding system. We
were able to identify children who had secure attachment
and, more importantly, we were able to identify children
who were more vulnerable. Drawings by children with se-
cure attachment revealed the highest levels of positive mark-
ers (vitality and pride) and the lowest levels in dimensions
that reflect potential problems. The ambivalent group was
conspicuous in the higher levels of role reversal depicted in
their drawings as compared to the children with secure at-
tachment; children with ambivalent attachment also drew
drawings with higher levels of vulnerability depicted (to-
gether with the disorganized group). The avoidant group ex-
pressed higher levels of tension than the ambivalent group.
Finally, children with disorganized attachment drew draw-
ings that exhibited the highest levels of bizarreness. These
findings attest to the difficulties of both the ambivalent
group and the disorganized group especially, and empha-
size the positive characteristics of the secure group (Madigan
et al., 2003; Fury et al., 1997).
The findings did not confirm our hypothesis regard-
ing the relationship between attachment and personality.
There were no differences between the four attachment
groups with respect to children’s personality characteris-
tics. However, positive personality traits (conscientiousness
and agreeableness) correlated with vitality, as predicted, and
correlated negatively with bizarreness and pathology; con-
scientiousness was also negatively correlated with tension.
Neuroticism, a personality trait that has been consistently
related to attachment among adults, does not show the same

Table 3 Pearson Correlation Matrix for Drawing Dimensions and Adjustment

Prosocial Emotional Behavioral

Behavior Symptoms Hyperactivity Social Problems Problems
Vitality .04 −.04 −.19∗∗ .06 −.16∗
Pride .07 −.09 −.23∗∗ .00 −.21∗∗
Vulnerability −.15∗ −.02 .16∗ −.01 .08
Isolation −.05 .07 .15∗ −.06 .16∗
Tension −.13† .06 .29∗∗∗ .04 .25∗∗∗
Role reversal .08 −.02 −.01 −.13† .03
Bizarre −.11 −.01 .15∗ .01 .21∗∗
Pathology −.02 .03 .19∗∗ −.07 .15∗
Note. N = 203 for all analyses.
†
p < .10, ∗ p < .05, ∗∗ p < .01, ∗∗∗ p < .001.
 16 CHILDREN’S FAMILY DRAWINGS
Table 4 Reported Adjustment Characteristics by Attachment Classifications
Secure Avoidant Ambivalent
n = 92 n = 43 n = 68
M M M M
Prosocial 2.44 (.51) 2.51 (.49) 2.20 (.53)
behavior
Emotional 1.68 (.55) 1.54 (.51) 1.73 (.49)
symptoms
Hyperactivity 1.82 (.57) 1.95 (.53) 2.05 (.55)
Social problems 1.49 (.49) 1.46 (.46) 1.55 (.48)
Behavioral 1.34 (.44) 1.36 (.38) 1.52 (.52)
problems
Note. Standard deviations appear in parentheses after the means.
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012
∗
p < .05, ∗∗ p < .01.
pattern in our study. It is possible that because our partici-
pants were children, the dimension of neuroticism at this
developmental stage is less problematic than among adults.
It is also possible that the assessment of family drawings is
less finely tuned to uncovering neuroticism.
The findings pertaining to the relationship between at-
tachment representations as reflected in children’s drawings
and the children’s adjustment demonstrated the superior
psychosocial functioning of children who were classified as
having secure attachment and the inferior functioning of the
ambivalent group, as already evidenced in previous stud-
ies (Pianta et al., 1999). The poor adjustment of the chil-
dren classified as disorganized was especially evident. Specifi-
cally, secure and avoidant children exhibited higher levels of
prosocial behavior than the ambivalent group, secure chil-
dren showed lower levels of conduct problems than the am-
bivalent and the disorganized groups, and avoidant children
showed lower levels of conduct problems than the disorga-
nized group.
The finding that children with secure attachment ex-
hibit good psychosocial functioning is not surprising. Their
trust in others, their adaptive ways of dealing with stress,
and their better emotional regulation strategies all promote
greater social–emotional competence and behavioral ad-
justment (Granot & Mayseless, 2001; Kerns et al., 2007;
Ranson & Urichuk, 2008). The inferior functioning of
the children with ambivalent attachment corroborates the
findings of Pianta et al. (1999) in their study of children
in kindergarten with respect to behavioral problems and
less competence. As in our study, Pianta et al. found that
avoidant children functioned relatively well. They suggested
that these children’s task-focused behaviors, as well as their
tendencies to manage independently without demanding
adult help, may result in their teachers perceiving them as
fairly well adjusted.
Also interesting was the lack of correlation between the
8 scales and the emotional symptoms and social problems
as rated by the children’s teachers. It may be that attach-
ment derivatives are more closely related to these disrup-
tive behaviors. Alternately, it may be easier to discern overtly
Disorganized
n = 19
F Attachment η2 Attachment
2.30 (.45) 4.57∗∗ .06
1.91 (.61) 2.09 .03
2.00 (.66) .78 .01
1.61 (.48) .52 .01
1.63 (.44) 2.93∗ .04
disruptive behaviors than to detect more subtle behaviors.
Training teachers to observe and detect less obvious behav-
iors could increase their abilities to identify and help vulner-
able children.
The study demonstrated the effectiveness of children’s
family drawings as a tool for learning about their relational
representations and pinpointing children at risk for having
adjustment problems in the school setting. Thus, children’s
drawings could serve as an important means to reveal the
children’s strengths, vulnerabilities, and weaknesses. Aspects
of children’s drawings were associated with positive person-
ality traits as well as with the more externalized adjustment
problems.
In the current study, homeroom teachers reported
on the participating children’s social adjustment, which
provided an additional and important perspective. Never-
theless, it should be noted that teachers see their pupils
mainly in learning situations, and although they likely are
better at identifying more troublesome behaviors, the teach-
ers may not be as sensitive to less salient behaviors. Reports
from children’s parents and other observers could add impor-
tant information that may provide a richer and more com-
plete picture.
The study was conducted in the Israeli cultural context,
which is characterized by high family values and close family
ties (Lavee & Katz, 2003). Similar studies should be under-
taken in other cultural contexts to allow for generalizations.
It is also important to bear in mind that the study was con-
ducted among a sample of children from families with low
socioeconomic status. The relationship between attachment
and adjustment may be different in children from lower risk
backgrounds.
Finally, it is important for clinicians and educators to
address children’s attachment representations as well as their
maladjusted behavior (London, Downey, Bonica, & Paltin,
2007) in order to improve their adjustment and to help alle-
viate their distress. Children’s drawings enable a glance into
their internal worlds and provide possible starting points
for helping them cope with difficulties and enhance their
adjustment. More systematic exploration and more precise
 GOLDNER / SCHARF
analysis of children’s drawings could improve intervention
efforts and help clinicians focus on specific assets as well as
on certain difficulties.
References
Bäckström, M., & Holmes, B. M. (2001). Measuring adult at-
tachment: A construct validation of two self-report instruments.
Scandinavian Journal of Psychology, 42, 79–86.
Barbaranelli, C., Caprara, G. V., Rabasca, A., & Pastorelli, C.
(2003). A questionnaire for measuring the Big Five in late child-
hood. Personality and Individual Differences, 34, 645–664.
Becker, T. E., Billings, R. S., Eveleth, D. M., & Gilbert, N. W.
(1997). Validity of scores on three attachment style scales: Ex-
ploratory and confirmatory evidence. Educational and Psycholog-
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012
ical Measurement, 57, 477–493.
Belsky, J., & Cassidy, J. (1994). Attachment: Theory and evidence.
In M. Rutter & D. Hay (Eds.), Development through life: A hand-
book for clinicians (pp. 373–402). Oxford, England: Blackwell.
Bowlby, J. (1980). Attachment and loss: Vol. 3. Loss: Sadness and
depression. New York, NY: Basic Books.
Bretherton, I., & Munholland, K. A. (2008). Internal working
models in attachment relationships: Elaborating a central con-
cept in attachment theory. In J. Cassidy & P. R. Shaver (Eds.),
Handbook of attachment: Theory, research and clinical applications
(2nd ed., pp. 102–129). New York, NY: Guilford Press.
Carlson, E. A., Sroufe, L. A., & Egeland, B. (2004). The construc-
tion of experience: A longitudinal study of representation and
behavior. Child Development, 75, 66–83.
Carver, C. S. (1997). Adult attachment and personality: Converg-
ing evidence and a new measure. Personality and Social Psychology
Bulletin, 23, 865–883.
Clarke, L., Ungerer, J., Chahoud, K., Johnson, S., & Stiefel, I.
(2002). Attention deficit hyperactivity disorder is associated
with attachment insecurity. Clinical Child Psychology and Psy-
chiatry, 7 (2), 179–198.
Deaver, S. P. (2009). A normative study of children’s drawings: Pre-
liminary research findings. Art Therapy: Journal of the American
Art Therapy Association, 26 (1), 4–11.
Deklyen, M., & Greenberg, M. T. (2008). Attachment and psy-
chopathology in childhood. In J. Cassidy & P. R. Shaver (Eds.),
Handbook of attachment: Theory, research and clinical applications
(2nd ed., pp. 637–665). New York, NY: Guilford Press.
Early Child Care Research Network. (2006). Infant-mother at-
tachment classification: Risk and protection in relation to
changing maternal caregiving quality. Developmental Psychology,
42, 38–58.
Finzi-Dottan, R., Manor, I., & Tyano, S. (2006). ADHD, tem-
perament, and parental style as predictors of the child’s attach-
ment pattern. Child Psychiatry and Human Development, 37,
103–114.
17
Fury, G., Carlson, E. A., & Sroufe, L. A. (1997). Children’s repre-
sentations of attachment relationships in family drawings. Child
Development, 68, 1154–1164.
Goldwyn, R., Green, J. M., Stanley, C., & Smith, V. (2000). The
Manchester Child Attachment Story Task: Relationship with
parental AAI, SAT and child behavior. Attachment and Human
Development, 2, 65–78.
Goodman, R. (1997). The Strengths and Difficulties Question-
naire: A research note. Journal of Child Psychology and Psychiatry
and Allied Disciplines, 38, 581–586.
Goodman, R. (2001). Psychometric properties of the Strengths
and Difficulties Questionnaire. Journal of American Child and
Adolescent Psychiatry, 40, 1337–1345.
Graham, C. A., & Easterbrooks, M. A. (2000). School-aged chil-
dren’s vulnerability to depressive symptomatology: The role of
attachment security, maternal depressive symptomatology, and
economic risk. Development and Psychopathology, 12, 201–213.
Granot, D., & Mayseless, O. (2001). Attachment security and ad-
justment to school in middle childhood. International Journal of
Behavioral Development, 25, 530–541.
Greenberg, M. T., Speltz, M. L., Deklyen, M., & Endriga, M.
C. (1991). Attachment security in pre-schoolers with and with-
out externalizing problems: A replication. Development and Psy-
chopathology, 3, 413–430.
Hagekull, B., & Bohlin, G. (2003). Early temperament and at-
tachment as predictors of the five factor model of personality.
Attachment and Human Development, 5, 2–18.
Kaplan, N., & Main, M. (1986). Instructions for the classification of
children’s family drawings in terms of representation of attachment.
Berkeley, CA: University of California.
Kerns, K. A., Abraham, M. M., Schlegelmilch, A., & Morgan, T.
A. (2007). Mother-child attachment in later middle childhood:
Assessment approaches and associations with mood and emo-
tion regulation. Attachment and Human Development, 9, 33–
53.
Kerns, K. A., Kelpac, L., & Cole, A. (1996). Peer relationship and
preadolescents’ perceptions of security in child mother relation-
ship. Developmental Psychology, 32, 457–466.
Kerns, K. A., & Stevens, A. C. (1995). Parent-child attachment in
late adolescence: Links to social relations and personality. Jour-
nal of Youth and Adolescence, 25, 323–342.
Lavee, Y., & Katz, R. (2003). The family in Israel: Between tra-
dition and modernity. Marriage and Family Review, 35, 193–
217.
Leon, K., Wallace, T., & Rudy, D. (2007). Representations of
parent-child alliances in children’s family drawings. Social De-
velopment, 16, 440–459.
London, B., Downey, G., Bonica, C., & Paltin, L. (2007). Social
causes and consequences of rejection sensitivity. Journal of Re-
search on Adolescence, 17, 481–506.
 18 CHILDREN’S FAMILY DRAWINGS
Lyons-Ruth, K. (1996). Attachment relationships among children
with aggressive behavior problems: The role of disorganized
early attachment patterns. Journal of Consulting and Clinical Psy-
chology, 64, 64–73.
Madigan, S., Ladd, M., & Goldberg, S. (2003). One picture is
worth a thousand words: Children’s representations of family as
indicators of early attachment. Attachment and Human Develop-
ment, 5, 19–37.
Main, M., & Moran, H. (1996). Disorganization and disorien-
tation in infant strange situation behavior: Phenotypic resem-
blance to dissociative states. In L. K. Michelson & W. J. Ray
(Eds.), Handbook of dissociation: Theoretical, empirical, and clin-
ical perspectives (pp. 103–107). New York, NY: Plenum Press.
Main, M., & Solomon, J. (1990). Procedures for identifying in-
fants as disorganized/disoriented during Ainsworth strange sit-
Downloaded by [Universidad Nacional Colombia] at 20:13 30 September 2012
uations. In M. T. Greenberg, D. Cicchetti, & E. M. Cummings
(Eds.), Attachment in the pre-school years: Theory, research and in-
tervention (pp. 121–160). Chicago, IL: University of Chicago
Press.
McCrae, R. R., Costa, P. T., Jr., Ostendorf, F., Angleitner, A.,
Hřebı́čková, M., Avia, M. D., . . . Smith, P. B. (2000). Nature
over nurture: Temperament, personality, and life span develop-
ment. Journal of Personality and Social Psychology, 78, 173–186.
Measelle, J. R., John, O. P., Ablow, J. C., Cowan, P. A., & Cowan,
P. A. (2005). Can children provide coherent, stable, and valid
self reports on the big-five dimensions? A longitudinal study
from ages 5 to 7. Journal of Personality and Social Psychology, 89,
90–106.
Mikulincer, M., & Shaver, P. R. (2006). Attachment in adulthood:
Structure, dynamics, and change. New York, NY: Guilford Press.
Milne, L. C., Greenway, P., & Best, F. (2005). Children’s behavior
and their graphic representation of parents and self. The Arts in
Psychotherapy, 32, 107–119.
Moss, E., Bureau, J.-F., Cyr, C., Mongeau, C., & St-Laurent, D.
(2004). Correlates of attachment at age 3: Construct validity of
the Preschool Attachment Classification System. Developmental
Psychology, 40, 323–334.
Mostkoff, D. L., & Lazarus, P. J. (1983). The Kinetic Family
Drawing: The reliability of an objective scoring system. Psychol-
ogy in the Schools, 20, 16–20.
Pianta, R. C., Longmaid, K., & Ferguson, K. E. (1999).
Attachment-based classifications of children’s family draw-
ings: Psychometric properties and relations with children’s ad-
justment in kindergarten. Journal of Clinical Psychology, 28,
244–55.
Piperno, F., Di Biasi, S., & Levi, G. (2007). Evaluation of family
drawings of physically and sexually abused children. European
Child and Adolescent Psychiatry, 16, 389–397.
Ranson, K. E., & Urichuk, L. J. (2008). The effect of parent-child
attachment relationship on child biopsychological outcomes: A
review. Early Child Development and Care, 178, 129–152.
Roisman, G. I., Holland, A., Fortuna, K., Fraley, C. R., Clausell,
E., & Clarke, A. (2007). The adult attachment interview and
self-report of attachment style: An empirical rapprochement.
Journal of Personality and Social Psychology, 92, 678–697.
Shaver, P. R., & Brennan, K. S. (1992). Attachment style and the
big five personality traits: Their connection with romantic re-
lationship outcomes. Personality and Social Psychology Bulletin,
18, 536–545.
Shiner, R. L., & Caspi, A. (2003). Personality differences in child-
hood and adolescence: Measurement, development, and conse-
quences. Journal of Child Psychology and Psychiatry, 44, 2–32.
Sroufe, L. A. (1983). Infant–caregiver attachment and adaptation
and competence. In M. Perlmutter (Ed.), Attachment status: Is
frightened and/or frightening parental behavior the linking mecha-
nism? (pp. 41–83). Hillsdale, NJ: Erlbaum.
Walls, M. E. (1998). Multiple predictors of mother-child attachment
and peer competency during the preschool years. (Doctoral disserta-
tion). Retrieved from http://docs.lib.purdue.edu/dissertations/
Weinfield, N. S., Sroufe, A. L., Egeland, B., & Carlson, E. (2008).
Individual differences in infant-caregiver attachment: Concep-
tual and empirical aspects of security. In J. Cassidy & P. R.
Shaver (Eds.), Handbook of attachment: Theory, research and clin-
ical applications (2nd ed., pp. 78–100). New York, NY: Guilford
Press.
Winjngaards-de Meij, L., Stroebe, M., Schut, H., Stroebe, W.,
van den Bout, J., van der Heijden, P., & Dijkstra, I. (2007).
Neuroticism and attachment insecurity as predictors of be-
reavement outcome. Journal of Research in Personality, 41, 498–
505.