evisiting the Bird’s Nest Drawing assessment: Toward a globalpproach
imor Goldnerraduate School of Creative Art Therapies, University of Haifa, Israel Faculty of Education, Oranim College of Education, Haifa, Israel
r t i c l e i n f o
rticle history:eceived 1 February 2014eceived in revised form 24 May 2014ccepted 16 June 2014vailable online 25 June 2014
eywords:ttachment securityird Nest Drawingsamily drawingsrojective art based technique
a b s t r a c t
The Bird’s Nest Drawing (BND) (Kaiser, 1996. The Arts in Psychotherapy, 23, 333–340) is an art-basedtechnique to assess attachment security. In the past 15 years, several studies have tested the validity ofthe BND mainly in adult clinical populations. In an attempt to strengthen the validity of the measure inchildren, the current study examined the associations between the BND and two other frequently usedassessment techniques evaluating attachment security: Kaplan and Main’s (1986. Instructions for theclassification of children’s family drawings in terms of representation of attachment. Berkeley, CA: Uni-versity of California) Family Drawing Coding System and the Attachment Security Questionnaire (Kernset al., 1996. Developmental Psychology, 32, 457–466) on a sample of elementary-school age children(n = 81) in Israel. BNDs were scored using specific indicators as well as global rating scales. The findingspoint to associations between the children’s self-reported security score, and the BND indicators and
global scales. Similarly, scores on both the specific indicators and global scales in family drawings werecorrelated with the BND global scales. Levels of BND scales varied as a function of the children’s attach-ment orientations derived from their family drawings. Associations were also found between attachmentorientations based on family and BND drawings. Results were interpreted as supporting the use of a globalapproach of rating in addition to a sign-based approach.
The Bird’s Nest Drawing (Kaiser, 1996; BND) is an art-basedechnique grounded in Attachment Theory which is used to assessttachment representations models. The current study brieflyeviews Attachment Theory, delineates the significant findingsrom previous studies, and reports on the results of an attempt toalidate the BND in elementary school-age children using specificndicators in addition to a global perspective of rating.
ttachment theory
Attachment Theory posits that beginning in infancy, and con-inuing throughout the lifespan, an individual’s mental healthnd capacity to form close relationships are intimately linked to
revious relationships with attachment figures that provide emo-ional support and protection (Belsky & Cassidy, 1994; Bretherton
Munholland, 2008). According to this theory, children’sctual experiences shape their representational models, which
subsequently serve to guide their behavior in novel circumstances(Bowlby, 1980). Children experiencing sensitive and responsivecare will develop trust in others, comfort with closeness and adap-tive ways of dealing with stress (secure attachment); avoidantattachment is associated with discomfort with closeness and aninclination for self-reliance, whereas anxious (ambivalent) attach-ment is associated with an intense desire for closeness and constantconcern about parental availability (Mikulincer & Shaver, 2006).Children with disorganized attachment are characterized by anapparent lack, or collapse, of a consistent organized strategy fordealing with stress. The particular forms and mixtures of disor-ganized behaviors tend to be idiosyncratic from child to child,but include anxious, helpless, or depressed behaviors, unexpectedfluctuations of approach and avoidance toward the attachment fig-ure, and other conflicted and unpredictable behaviors (see Main &Solomon, 1990).
As children get older, the attachment system develops towardincreased self-reliance on the part of the child in that older childrenare better at coping with stress situations and are less dependent on
parents (Marvin & Britner, 1999). Moreover, there may be a changein the goal of the attachment system, with availability rather thanproximity of the attachment figure becoming the aim of the orga-nization (Bowlby, 1987, cited in Ainsworth, 1990; Kerns, Tomich,
spelmeier, & Contreras, 2000; Lieberman, Doyle, & Markiewicz,999). Nevertheless, despite these changes, children continue toeed and rely on parents as attachment figures (Bowlby, 1979),nd individual differences in attachment security, expressed inmotional regulation and exploration, have important implicationsor personality development as well as for adjustment (Weinfield,roufe, Egeland, & Carlson, 2008).
In general, attachment security facilitates resilient functioningnd serves as a buffer when coping with adversities, whereasnsecure attachment might hamper children’s adjustment. Fornstance, substantial evidence indicates that elementary school agehildren with secure attachment develop better social-emotionalompetence (Granot & Mayseless, 2001; Ranson & Urichuk, 2008).hey have more constructive coping mechanisms, better regula-ion of emotion in the classroom (Kerns, Abraham, Schlegelmilch,
Morgan, 2007), and better behavioral adjustment (Granot &ayseless, 2001). Attachment security is also associated with
ocial competence (NICHD Early Child Care Research Network,006; Kerns, Tomich, & Kim, 2006), and with social support andeer acceptance (Granot & Mayseless, 2001; Kerns, Kelpac, &ole, 1996). Security of attachment was negatively associated withhildren’s loneliness (Kerns & Stevens, 1995; Kerns et al., 1996) andess depressive symptomatology (Graham & Easterbrooks, 2000).n addition, a strong association was found between insecurettachment and early behavioral problems, anti-social behavior,isruptive hyperactive behavior (Lyons-Ruth, 1996), and subse-uent conduct disorder (Greenberg & Speltz, 1988; Greenberg,peltz, Deklyen, & Endriga, 1991).
rojective art based techniques to evaluate attachment security
Given the severity of psychological and social-behavioral symp-oms experienced by insecurely attached children, it seems crucialo better understand the representations of insecurely attachedhildren in their elementary school years (Cicchetti, Toth, & Bush,988). Diverse instruments are available to help healthcare pro-essionals assess attachment orientation in these children usingelf-report measures of attachment security (Kerns et al., 1996) orvoidant and preoccupied coping strategies (Finnegan, Hodges, &erry, 1996). However, completion of these questionnaires requireshe cooperation of the child, who often attempts to conceal hisr her negative experience with his/her caregivers. In these cases,otential deterioration into a clinical situation may be more diffi-ult to detect.
Given the shortcomings of self-report questionnaires, in the lasthirty years two art-based projective assessments have been devel-ped to evaluate children’s attachment security: Kaplan and Main’s1986) Family Drawing Coding system and Kaiser’s (1996) Birdest Drawing. These approaches are grounded on the assumption
hat drawing is a natural mode of expression for children. Longefore children verbalize their feelings and thoughts into words,hey express both conscious and unconscious attitudes, wishes,nd concerns in symbolic methods such as drawing. It thus seemslausible that representations of attachment experiences would beevealed in drawings, and, specifically, that the child’s “inner work-ng models” (Bowlby, 1973) of the self, caregivers, and the self witharegivers would be manifested (Fury, Carelson, & Sraufe, 1997).
lassification of attachment representations through familyrawings
Kaplan and Main’s (1986) sign-based coding system for ana-
yzing children’s family drawings contains a set of indicators thatlassifies children’s family drawings according their attachmentecurity. The researchers studied family drawings by kindergartenhildren as predictors of assignment to the secure or insecure
erapy 41 (2014) 391–399
attachment categories and developed a classification system forcoding a sample of children’s family drawings that matched, with76% accuracy, their Strange Situation classifications (Ainsworth,Blehar, Waters, & Wall, 1978).
The results showed that drawings by children reflecting secureattachment are realistic; figures are complete, grounded and cen-tered, and individuated (figures seem unique and are not drawnexactly alike). There is a natural proximity among family membersand an impression of happiness in the family. Drawings by childrenclassified as avoidant try to convey a positive picture, emphasizinginvulnerability and happiness. Arms may be absent or drawn in away that does not allow holding, there is lack of individuation of thefigures and lack of movement in the picture. Drawings by childrenclassified as ambivalent include figures that are extremely large orsmall, and figures that either overlap or are separated by barriers.Soft body parts and facial features are exaggerated in the drawings.Finally, drawings assigned to the disorganized classification ofteninclude strange marks, threatening and fantasy themes, unfinishedobjects or figures, and sometimes excessive and irrational sweet-ness (Kaplan & Main, 1986).
Based on Kaplan and Main’s (1986) coding system, Furyet al. (1997) developed a global approach for coding chil-dren’s family drawings using eight global rating scales. Thesescales consist of two positive dimensions (vitality/creativity andfamily pride/happiness) and six negative dimensions (vulnera-bility, emotional distance/isolation, tension/anger, role reversal,bizarreness/dissociation, and global pathology).
Although difficult and time-consuming, this system has beenshown to have reliability and validity for determining attachmentcategories (Kaiser & Deaver, 2009). For instance, Carlson, Sroufe,and Egeland (2004) conducted a longitudinal study in which chil-dren’s family drawings at age 8 were found to correlate with theirattachment classifications according to data gathered from pre-vious interviews with the children when they were in preschooland again at age 12 (p < .001). The global rating scales were alsocorrelated with children’s attachment classifications as assessedin infancy (p < .001). The analysis showed that even after con-trolling for IQ, current life stress, and emotional functioning, thechildren’s early attachment history made a significant contributionto the prediction of negative dimensions in their drawings (p < .001)(Fury et al., 1997). In another study (Madigan, Ladd, & Goldberg,2003), children whose family drawings depicted higher levels ofemotional distance, vulnerability, and parent–child role reversalwere found to have an insecure attachment history, whereas fam-ily drawings that scored higher on family pride and lower in globalpathology were drawn by children with a secure attachment his-tory (p < .05–.01).
Recently, researchers have started using Kaplan and Main’s(1986) coding system to assess the associations between children’sattachment classifications as manifested in their family drawingsand children’s adjustment. The findings suggest better adjustmentin the social, academic and behavioral realms among securelyattached children. For example, kindergarten children (n = 200)whose drawings were judged as secure were rated as more sociablewith their peers, more task-oriented and more socially com-petent than insecurely ambivalent children (p < .05–.01) (Pianta,Longmaid, & Ferguson, 1999). The superior functioning of securelyattached children was also evidenced in 9–12 year old Israelis(n = 222) who exhibited higher levels of pro-social behavior andfewer conduct problems than their counterparts (p < .05–.01)(Goldner & Scharf, 2011).
A previous study (Goldner & Scharf, 2012) designed to detect
children’s adjustment found that indicators reflecting attachmentinsecurity such as omitting and adding figure parts, adding bizarremarks, as well as a lack of femininity were correlated with inter-nalizing problems among Israeli elementary school age children
choth
(iaacgii&
B
dldta&tbuaesitafl
aKBtapstwbadcmtd
sg(BPapombaemdcie
d
L. Goldner / The Arts in Psy
n = 222, p < .05–.001). Role reversal in children’s family drawingsn American children (n = 44, mean age = 8.14) with an incarcer-ted or estranged parent was associated with more phone, mail,nd physical contact with this parent. Greater overall insecurity inhildren’s family drawings was associated with more hostile care-iver behavior whereas increased global pathology and bizarrenessn drawings were associated with stressful life events and parent-ng stress in an American sample (p < .05–.001) (Dallaire, Ciccone,
Wilson, 2012).
ird’s nest drawing (BND)
Inspired by the research on family drawings, Kaiser (1996)eveloped the Bird’s Nest Drawing (BND). In an attempt to use a
ess threatening and anxiety-provoking assessment than a familyrawing, the researcher identified personal internal representa-ions of the self and others by eliciting experiences of attachmentlong with metaphorical concepts of safety and protection (Kaiser
Deaver, 2009; Sheller, 2007). BNDs were rated for the presence orhe absence of numerous indicators such as content (eggs or babyirds), a tree, the bottom of the nest able to contain [birds or eggs],se of most of the page, using more than two colors, line quality,nd centered image. Later on, Francis, Kaiser, and Deaver (2003)xpanded this list by adding several indicators such as the per-pective on the nest (the nest as seen from above, the nest drawnn profile, tilted nest), the features of the tree (drawing of a wholeree, depictions of a dead or a dying tree), the existence of feedingctivity, the presence of an entire family of birds, the existence ofying birds and a general faint quality of the drawing.
The classification of BND drawings with existing measures ofttachment security revealed several associations. For instance, inaiser’s original study of the Bird’s Nest Drawing (Kaiser, 1996),NDs were collected from a sample of mothers (n = 41). Usinghe Attachment to Mother (ATM) scale of the Inventory of Parentnd Peer Attachment (IPPA) (Armsden & Greenberg, 1987) partici-ants were divided into a more securely attached group and a lessecurely attached group. The high-ATM group included birds inheir drawings significantly more often than the low-ATM grouphereas the low-ATM group often depicted a tilted nest without
irds. A marginally significant trend was found for the appear-nce of the nest. Participants from the low-ATM group tended toraw nests without bottoms or tilted nests that were unable toontain their contents. Kaiser also noticed that the BNDs of theore securely attached group had a whimsical, happy quality and
hat participants composed titles for their drawings that could beescribed as engaging or gently humorous.
Francis et al. (2003) compared the BNDs of adults with sub-tance abuse (SA) disorders (n = 43) to patients in a comparisonroup from a medical clinic who had no substance abuse disordersn = 27). Participants completed a BND, a story about their BND, andartholomew and Horowitz’s (1991) Relationship Questionnaire.articipants in the SA group were more likely to have an insecurettachment style and to use fewer colors in their BNDs than partici-ants in the control group. The control group, comprised primarilyf participants with a secure attachment style, was significantlyore likely to feature green as the predominant color and include
irds drawn in the nest. The SA group most often depicted the nests tilted or viewed from above. Analysis of the drawings for thentire sample indicated that securely attached participants drewore birds, drew an entire bird family, used four or more colors,
rew the nest in profile (not tilted) and used green as the dominantolor. Content analysis of the stories revealed themes such as fam-
ly, nature and renewal of life, food or hunger, abandonment, andnvironments of personal significance (Francis et al., 2003).
Recently, studies also examined the use of the BND in chil-ren and adolescents (see Kaiser and Deaver, 2009). For example,
erapy 41 (2014) 391–399 393
Sheller (2007) used the BND and a Bird’s Nest Sculpture (BNS) as ametaphoric process to examine children’s internal attachment rep-resentations using a phenomenological approach. The study wasconducted on four school age children identified by the researchfacility as being in the early stage of requiring therapeutic ser-vices due to insecure attachment. The children’s stories about theirartwork revealed themes of danger, lack of protection, and vul-nerability. Moreover, the insecure attachment pattern was furtherportrayed by the absence of the mother bird, distortion of the fatherbird figure, and placing the nest on thin and brittle branches.
In sum, studies have indicated that the BND may be used as anevidence-based tool for the assessment of individuals’ attachmentsecurity using both separate indicators as well as an overall impres-sion. However, these methods also have several shortcomings.Most studies have focused on clinical or maladaptive populationsand were based on relatively small sample sizes. Moreover, theexamination of the validation was based mostly on self-reportmeasures. Finally, attachment classifications were made on thebasis on sign-based approach, as described in Kaplan and Main’s(1986) approach, rather than using general impressions along withgroups of indicators.
To overcome some of these shortcomings, the current study wasdesigned to expand validation of the BND to assess attachmentrepresentations in non-clinical elementary school-aged children.Associations have been examined between the BND indicators andboth explicit (Kerns et al.’s, 1996 Attachment Security self-reportQuestionnaire) and implicit measures (Kaplan & Main’s, 1986 Fam-ily drawing coding system). As in other attachment measures (e.g.,Strange Situation, Adult Attachment Interview), and the use of gen-eral scales by Fury and her Colleagues (1997) validation attemptshave also been made using integrative scales composed of aggre-gates of BND indicators. It is possible that individual signs mightnot discriminate early attachment history, and that integrative rat-ings and general impressions may have more power than discreteindicators (Fury et al., 1997; Kaiser & Deaver, 2009). In particu-lar, it is hypothesized here that indicators and scales indicativeof attachment security in the BND should correlate with indica-tors and scales in children’s Family drawings assessing attachmentsecurity as well as with their attachment security score derivedfrom their self-report questionnaire.
Method
Participants
Eighty-one (81) children participated in the study. The childrenwere recruited from elementary schools in middle-SES neighbor-hoods in Israel. Forty-one percent of the children were boys (n = 33)and 59% were girls (n = 48). The mean age of the children was 10.26years (range 8–12; SD = 1.26); 69% of the children were from two-parent families and 31% were from divorced families. Children hadan average of 2.38 siblings (range 1–11; SD = 1.56).
Procedure
After receiving consent from children and their parents, chil-dren completed the Family drawing, the BND, and the attachmentquestionnaire during home visits, in their natural surroundings,conducted by trained research assistants. Participants were assuredof the confidentiality of their responses.
Measures
Family drawing: Children were asked to draw their families,on white paper, using eight colored felt-tipped pens. After theycompleted the drawings, the research assistants wrote the identity
3 choth
oTa(fseiagbeDpadisidw
(ledffoweatbifiiwerct
aab2D
tiab(uEt(dhr
oTcd
94 L. Goldner / The Arts in Psy
f each figure and any additional information the child provided.wo coders coded all drawings using Kaplan & Main’s (1986)ttachment classifications that focus on the location of the figuresfigures are grounded and centered), natural proximity amongamily members, genuine smile, size of figures (realistic, large ormall), emphasis of soft body and facial parts (belly, lower body,yes and nostrils) missing father or mother, degree of movement,ndividualized characteristics, completeness of figures, missingrms or arms drawn in a way that does not allow holding, exag-erated sweetness or overly bright drawings, addition of strange,izarre, inexplicable or unfinished objects, omission of unexpectedlements from the drawing, scratched out figures and restarts.rawings were coded on a Likert scale ranging from 1–3 (1—notresented, 3—fully presented). A detailed manual, based on Kaplannd Main’s work, was adapted, delineating specific examples toiminish ambiguity. The training and adaptation of the manual,
ncluding reliability examination, was initially based on anotherample. The intra-class reliability between the two coders using anntra-class correlation coefficient (ICC) for 24 drawings (30% of therawings) ranged from .70 to 1.00. Disagreements between codersere resolved by consensus.
Moreover, each drawing was also coded using Fury’s et al.’s1997) eight global scales ranging from 1 (low level) to 7 (highevel) for the following dimensions: (a) vitality, which refers tomotional investment, creativity, and the richness of detail in therawing; (b) pride, which refers to expressions of happiness in theamily group and the degree of connectedness depicted amongamily members; (c) vulnerability, as expressed in the placementf figures and distortions in the size of body parts; (d) isolation,hich refers to a sense of emotional distance and loneliness
xpressed in neutral or negative affects, and is also depicted as lack of proximity between the mother and child figures; (e)ension/anger, referring to a restriction of the figures expressedy careless appearance, lack of colors, etc.; (f) role reversal, as
ndicated by disproportionate size and/or roles depicted in thegures; (g) bizarreness, which indicates disorganization, reflected
n unusual signs and fantasy themes; and (h) global pathology,hich refers to negativity, incompleteness of the figures, and poor
xpression in the drawing, details, and background. The inter-ratereliability for the global scales using intra-class correlation coeffi-ients (ICC) for 24 drawings (30% of the drawings) ranged from .80o .95. Disagreements between coders were resolved by consensus.
In addition, the drawings were classified into one of fourttachment categories (Kaplan & Main, 1986): secure, avoidant,mbivalent and disorganized, as described above. The agreementetween the coders on the four main classifications, based on4 cases, was 84% {�2 (9) = 48.51, p < .001; kappa = .83, p < .001}.isagreements between coders were resolved by consensus.
The Bird Nest Drawing: A list of distinct indicators based onhe list in Kaiser (1996) and Kaiser et al. (2003) was obtained andncluded the depiction of: content in the nest, a tree, a dry tree,
grounded tree, a baby bird, a parent bird, an extremely largeird, an extremely tiny bird, eggs, an adequate caregiving behaviorincubation, feeding, protection), a protected nest, a spiky nest,sing of more than 20% of the page and more than three colors.ach drawing was coded based on the presence or the absence ofhese indicators. All of the 14 markers had sufficiently high Kappas.70 to 1.00) to be considered reliable for 24 drawings (30% of therawings); therefore, the inter-rater reliability was consideredigh. Disagreements between coders regarding the drawings wereesolved by consensus.
In addition, as done in Fury et al. (1997) four global scales based
n Kaiser’s (1996), Francis et al. (2003) lists were constructed.hree scales (Vitality, General pathology, and Bizarreness) wereonstructed on a 7-point Likert scale (1 = low level of the specificimension to 7 = high level) assessing specific dimension in the
erapy 41 (2014) 391–399
drawing, whereas the fourth scale (optimism) was constructed on a10-point Likert scale (1 = low level to 10 = high level) evaluating theoverall feeling of the drawing. Specifically, Vitality refers to emo-tional investment in drawing and creativity manifested in richness,colorfulness, decorativeness, adding details to the drawings, refer-ence to the environment, and interaction between the birds. Generalpathology refers to a general impression of negativity manifestedin sense of emptiness, loneliness and spikiness, unprotected nest(e.g. a hollow nest, a tilted or a nest that is falling apart, spread outor loose), threat to birds, lack of stability, overlap between birds,lack of proportion, and size distortion of birds. Bizarreness refers tothe presence of irrelevant and irrational elements, lack of logic inthe drawing, strange form of the nest, omitting unexpected parts,erasures, over-sweetness, and inadequate caregiving. Finally, Opti-mism refers to an overall positive impression of the drawing. A lowlevel of optimism is expressed by loneliness, emptiness, spikiness,distorted size of the nest or the birds, strangeness, a threat to thebirds, unprotected nest, and a broken line quality. A high level ofoptimism is manifested in a sense of fullness, stability and protec-tion, the existence of adequate caregiving behavior (i.e. gaze, feed-ing or incubation), a spontaneous and a flowing line quality, color-fulness and proportion between objects. Again, a detailed manualwas adapted, delineating specific examples to diminish ambigu-ity. The training and adaptation of the manual, including reliabilityexamination, was initially constructed based on another sam-ple. The intra-class reliability between the two coders regarding24 drawings (30%) ranged from .74 to .86 (for intercorrelationsbetween the sub scales controlling for gender see Appendix 1).
Finally, two classifications of attachment security (secureversus insecure) were drawn from the general scales. SecureBNDs were characterized by high levels of vitality and optimismthat may reflect the prominent place secure individual allocateto intimate relationships and their ability to cope in a flexiblemanner (Cassidy, 2001; Collins & Sroufe, 1999), while Insecuredrawings were represented by low levels of vitality and optimismand high levels of pathology and bizarreness. This kind of picturemay resemble avoidant individuals’ tendency to minimize thevalue of early attachment relationships as well as to deactivateand minimize emotional expression (Dozier & Kobak, 1992) ormirror ambivalent individuals’ constant concern about caregiver’savailability as well as their intense wish for closeness (Mikulincer& Shaver, 2006). The agreement between the coders on the fourmain classifications, based on 24 cases, was 87.5% {�2 (9) = 50.74,p < .001; kappa = .83, p < .001}. Disagreements between coderswere resolved by consensus.
Children’s attachment security: Children’s attachment securitywas evaluated using the Attachment Security Scale (Kerns et al.,1996). The inventory assesses children’s perceptions of securityin mother–child relationships. It is composed of 15 items that arerated on a 4-point scale using Harter’s (1985) format: “Some kids. . .other kids. . .” (e.g., “Some kids find it easy to trust their mom BUTother kids are not sure if they can trust their mom’.’) Children werefirst asked to choose which statement was more characteristic ofthem, and then to indicate whether the statement was really truefor them or sort of true for them. Ratings (on a 4-point scale) weresummed across the 15 items to form an attachment security scoreranging from 15 to 60, with higher scores indicating a more securerelationship. This scale has been associated with general, social,conduct and academic competence (Kerns et al., 1996). Moreover,the scores have been shown to correlate with the degree of emo-tional support received from mother, father and friends, as wellas with mothers’ reports regarding their acceptance and willing-
ness to provide a secure base to their child (Kerns et al., 1996).Test–retest reliability has been reported at .75 and internal consis-tency ranges from .72 to .84 (Bauminger & Kimhi-Kind, 2008; Kerns,Aspelmeier, Gentzler, & Grabill, 2001; Granot & Mayseless, 2001).
chotherapy 41 (2014) 391–399 395
aaahmMs
R
P
ragdsdfddrB
eTtaty
Ti
sQ2taoi�Sd
dslscw
car
Tg
ceb
Table 1Associations between Attachment Security scale, distinct family drawings signs,family drawing global scales and BND global scales, controlling for gender.
Vitality Pathology Bizarreness Optimism
Security scale .29* −.10 .06 .23*
Distinct signs in family drawingsGrounded figure .02 −.22* −.26* .20#
Centered figures .26* −.31** .20# .35**Completed figures .13 .21# .17 .13Omitting unexpected parts −.16 .23* .10 −.17Tiny figures −.07 .04 .09 −.01Exaggerated figures −.07 .24* .11 −.10Emphasized parts .03 .21# .37** .15Genuine smile .23* −.31** −.11 .28*Individuated figures .40*** −.36** −.12 .39***Gender differences .10 −.23# −.24# .16Feminine signs .09 .17 .22 .12Arms do not able to hold −.13 .05 .00 .12Natural closeness .03 −.20# −.34* .16Distance from mother .03 −.04 −.11 −.05Distance from Father .17 .27* −.21 −.21Movement .30** −.29** −.07 .35*Add strange signs −.05 .14 .23* −.10Bizarre and phantasy themes .09 −.03 .19 .09Over-sweetness .12 −.19 −.22# .14Erasures −.11 .27* .35*** −.29*
Global scales in family drawingsVitality .62*** −.59*** −.35** .65***Pride .38** −.44*** −.28* .48***Vulnerability −.12 .27* .16 −.18Isolation −.07 .21# .07 −.18Stress −.48*** .55*** .38** −.59***Role reversal .03 .06 .33* −.13
L. Goldner / The Arts in Psy
As implemented in Kerns et al. (1996), a score of 45 was useds a categorical cut-off point to distinguish secure child–motherttachment (more than 45 points) from insecure child–motherttachment (fewer than 45 points). This categorical cut-off pointas been used in previous studies in the field of attachment iniddle childhood (e.g. Bauminger & Kimhi-Kind, 2008; Granot &ayseless, 2001; Kerns et al., 1996). Alpha Cronbach in the current
tudy .77.
esults
reliminary analyses
For preliminary analyses, a series of independent-sample T-testsevealed that child’s gender was significantly related to the Furynd the BND global scales. Regarding children’s family drawings,irls scored significantly higher than boys for vitality (t = −3.16,f = 78, p < .01) and pride (t = −3.13, df = 78, p < .01), whereas boyscored higher for stress (t = 2.86, df = 78, p < .01). Regarding the chil-ren’s BND global scales, girls scored significantly higher than boysor vitality (t = −2.78, df = 76, p < .001), and optimism (t = −2.95,f = 77, p < .01) whereas boys scored higher in pathology (t = 2.33,f = 77, p < .01). No differences between boys and girls were foundegarding the distinct indicators in children’s family drawings orNDs.
The two-way, secure/insecure distribution based on Kern’st al.’s questionnaire was 61 secure (75%) and 20 (25%) insecure.he four-way secure/avoidant/ambivalent/disorganized distribu-ion based on the children’s family drawings was 31 secure (38%), 8voidant (10%), 30 ambivalent (37%) and 12 disorganized (15%). Thewo-way secure/insecure distribution based on children’s BNDsielded 27 secure (34%), 52 insecure (66%).
he association between attachment security score and the BNDndicators and global scales
To examine the associations between attachment securitycores as was obtained from the children’s Attachment Securityuestionnaire and the distinct markers in the BND, a series of 2X
Chi-Square tests for the entire sample was conducted betweenhe two attachment groups (secure/insecure) as derived from thettachment security questionnaire using Kern et al.’s (1996) cut-ff point and the distinct indicators of the BND. Differences in BNDndicators were found for two indicators: depiction of a baby bird2 (1, n = 79) = 4.66, p < .05 and a tiny bird �2 (1, n = 56) = 5.97, p < .05.ecure children tended to draw a baby bird, whereas insecure chil-ren tended to draw a tiny bird.
To examine the relations between the continuous scale of chil-ren’s attachment security (Kerns et al., 1996) and the BND generalcales, after controlling for gender, a series of partial Pearson corre-ations between scores on the continuous security scale and generalcores on the BND scales was conducted. As can be seen in Table 1hildren’s security score correlated with vitality (r = .29, p < .05) andith optimism (r = .23, p < .05).
Finally, to examine the associations between attachmentlassifications (secure versus insecure) according to children’sttachment security scale and the BNDs a 2X 2 Chi-Square test wasun, however no correlation was found between the classifications.
he associations between children’s family drawings and the BNDlobal scales
To examine the relations between the distinct indicators of thehildren’s family drawings (Kaplan & Main, 1986) and the BND gen-ral scales, a series of partial Pearson correlations was computedetween the scores on the distinct indicators of the family drawing
Note: # p < .10, p < .05, ** p < .01, ***p < .001.
and the BND global scales, controlling for gender. As can be seen inTable 1, levels of grounded figures in the family drawings were neg-atively correlated with pathology and bizarreness in BNDs. Levelsof centered figures, genuine smile, individuated figures, and move-ment in the family drawings were positively correlated with vitalityand optimism and negatively with pathology in the BNDs. Levels oferasures in the family drawings were positively correlated withpathology and bizarreness and negatively with optimism in BNDs.Omitting unexpected parts, exaggerated figures, and distance fromfather in the family drawings was positively correlated with pathol-ogy in the BNDs. Emphasized parts and add strange signs werepositively correlated with bizarreness, while natural closeness wasnegatively correlated.
To examine the associations between the general scales of thechildren’s family drawings (Fury et al., 1997) and the BND gen-eral scales, a series of partial Pearson correlations was conductedbetween scores on the general scales of the family drawing and theBND scales, controlling for gender. As can be seen in Table 1, vitalityand pride in family drawings were positively correlated with vital-ity and optimism and negatively with pathology and bizarreness inthe BNDs. Stress and pathology in family drawings were negativelycorrelated with vitality and optimism and positively with pathol-ogy and bizarreness in the BNDs. Vulnerability in family drawingswas positively correlated with pathology in the BNDs. Role rever-sal in family drawings was positively correlated with bizarrenessin BNDs.
To examine the associations between the distinct markers of theBND drawings and attachment classification as was derived fromthe children’s family drawings, a series of 2X 2 Chi-Square tests
were examined. Due to the small number of participants in thedisorganized and the avoidant groups, the three insecure classifi-cations were combined into one insecure group. As can be seen inTable 2, differences in BND indicators were found for six indicators:
396 L. Goldner / The Arts in Psychotherapy 41 (2014) 391–399
Table 2Chi-Square analyses: Attachment security according to children’s family drawingsby BND indicators.
Indicator Secure Insecure Chi-Squaren = 31 n = 48
Content No 4 36 1.71Yes 27 12
Depiction of a tree No 7 19 2.47yes 24 29
Grounded tree No 3 15 5.71*Yes 24 25
Protected nest No 11 34 9.60**Yes 20 14
Use of more than 20percent of the format
No 1 11 5.67*Yes 30 37
Depiction of young birds No 21 37 .84Yes 10 11
Depiction of a parent bird No 8 20 2.28Yes 23 27
Depiction of eggs No 15 27 .47Yes 16 21
Depiction of a spiky nest No 24 27 3.69#
Yes 7 21
Depiction of a tiny bird No 20 20 1.63Yes 5 11
Depiction of a large bird No 25 23 4.38*Yes 1 7
Depiction of a dry tree No 21 17 4.77*yes 4 13
N
dtbmn�gggoaopttpb
Table 4Chi-Square analyses: Association between attachment classifications for the entiresample and for boys and girls separately.
BND classification
Family Secure Insecure Total Chi-Square KappaDrawings
epiction of a grounded tree �2 (1, n = 67) = 5.71, p < .05, depic-ion of a dry tree �2 (1, n = 55) = 4. 77, p < .05, depiction of a largeird �2 (1, n = 56) = 4.38, p < .05, use of more than 20% of the for-at �2 (1, n = 79) = 5.67, p < .05, depiction of a protected nest �2 (1,
= 79) = 9.60, p < .01, and depiction of adequate caregiving behavior2 (1, n = 79) = 5.08, p < .05. Children from the insecure attachmentroup tended to draw less protected nests and less adequate care-iving behavior. By contrast, children from the secure attachmentroup tended to use more than 20% of the page, to draw less driedut trees and large birds. Examination of these associations for boysnd girls separately showed gender differences regarding the usef these indicators. Specifically, insecure boys tended to draw lessrotected nests �2 (1, n = 33) = 14.06, p < .001. Secure girls tended
2
o use more than 20% of the page � (1, n = 46) = 4.71, p < .05 ando draw more adequate caregiving behavior �2 (1, n = 46) = 6.88,
< .01. On the other hand, insecure girls tended to include largeirds in their drawings �2 (1, n = 34) = 4.53, p < .05.
able 3ifferences in BND global scales across attachment groups according to family drawings,
Secure Avoidant Ambivalent
n = 30 n = 7 n = 29
Vitality 4.85 2.43 4.05
1.25 1.17 .99
Pathology 2.15 3.57 2.85
.73 .53 .72
Bizarreness 1.73 2.50 2.24
.69 .96 76
Optimism 6.01 3.07 4.87
1.59 1.13 .90
ote: ** p < .01, ***p < .001.
Note: ***p < .001.
To examine the differences between children’ scores on theirBND global scales according to their attachment groups as wereobtained from their family drawings, a MANCOVA analysis control-ling for gender was conducted with attachment patterns servingas the independent variable and the BND general scales servingas the dependent variable. The MANCOVA revealed a significantmain effect for groups F(4,71) = 7.99, p < .001, �2 = .31 but not forgender. As can be seen in Table 3 the Anova analyses showed dif-ferences for all four variables. Post-hoc Duncan tests demonstratedhigher levels of vitality and optimism among securely attachedchildren compared to children with disorganized and ambivalentattachment security. These in turn had a higher level of vitalitycompared to avoidant children. In addition, securely attached chil-dren and children with ambivalent attachment security exhibited alower level of bizarreness compared to their counterparts. Finally,securely attached children exhibited a lower level of pathologycompared to children with ambivalent and disorganized attach-ment security orientations. These in turn exhibited a lower level ofpathology than children with an avoidant attachment classification(see Table 3).
Finally, to examine the associations between attachment clas-sifications (secure versus insecure) according to the children’sfamily drawings and the BNDs a 2X 2 Chi-Square test was runand revealed a significant association between the classifications�2 (1, n = 79) = 48.97, p < .001, Kappa = 78, p < .001 (see Table 4).As can be seen in Table 4, a similar picture was obtained forboys �2 (1, n = 33) = 20.11, p < .001, Kappa = 78, p < .001, and girlsseparately �2 (1, n = 44) = 25.56, p < .01, Kappa = 78, p < .001. Girlstended to produce secure drawings while boys tended to make
insecure drawings. Pictures 1 to 6 illustrate pairs of drawings (i.e.family and bird nest) by the same child. Pictures 1 to 2 show asecure classification. Pictures 3 to 6 show an insecure classification(Figs. 1–6).
controlling for gender.
Disorganized F �2 Contrastn = 12
3.70 8.47*** .26 2<3,4<1.65
3.37 10.73*** .31 1<3,4<2.80
2.75 4.80** .17 1,3<2,41.30
3.87 17.76*** .34 2<3,4<1.80
L. Goldner / The Arts in Psychotherapy 41 (2014) 391–399 397
Fig. 1. An example for a secure family drawing.
Fig. 2. An example of a secure BND.
Fig. 3. An example for an insecure family drawing.
Fig. 4. An example of an insecure BND.
Fig. 5. An example for an insecure family drawing.
Fig. 6. An example of an insecure BND.
Discussion
The results of the current study lend additional weight to usingthe BND as a way to tap attachment representations in elemen-tary school age children. The current study provides an additionalvalidation of certain distinct indicators such as the depiction of aparent bird or a tree among secure children. Moreover, it suggeststhere should be a broader approach to classifying drawings andhighlights the need for a more integrative perspective rather thansole reliance on individual signs. The use of aggregations of signsand a global rating scale as a powerful tool to identity attachmentclassification is consistent with the findings reported in Fury et al.(1997) regarding children’s family drawings, and emphasizes therole of organization of the drawings similar to other attachmentmeasures (e.g., the AAI or the Strange Situation) which are basedchiefly on coherence (Main, 1996). Specifically, the most robustpattern of results was obtained using the optimism scale, whichcontains aspects of caregiving behaviors, and positive interactionsbetween birds and the child’s emotional investment in the drawing.This pattern may signify the centrality of caregiving relationshipsin human development (Bowlby, 1969, 1973) that afford emotionalsupport and protection (Belsky & Cassidy, 1994). This optimisticdepiction may also indicate children’s positive internal workingmodels regarding their social network as available, benevolent andtrustworthy as well as their exploratory competence and ability tocreate in a relatively flexible manner.
This study also revealed gender differences in children’s fam-ily drawings and BNDs. Girls exhibited high scores in the positiveglobal scales indicative of attachment security both in their familyand the bird nest drawings, whereas boys scored higher in the neg-ative scales indicative of attachment insecurity. The data for girlsis consistent with findings in previous studies that demonstratedAmerican girls’ superiority compared to boys in drawing human fig-ures, especially female figures (Koppitz, 1968; Scott, 1981), as wellas Japanese (Behrens & Kaplan, 2011) and Israeli girls’ higher lev-els of vitality and pride in family drawings (Goldner & Levi, 2014).However, they contradict Deavers’ findings (2009) regarding theimpact of gender on Human Figure Drawings (HFD) among second
and fourth American graders, which showed a significant differ-ence between boys and girls only with regard to color fit with girlsusing color more realistically than boys.
3 choth
edwdattM
rrpiitce&2Wtaie
C
fatratdst1
ccditefisotis(Ttdm
L
tho
98 L. Goldner / The Arts in Psy
These differences may reflect girls’ more extensive drawingxperience, their extra attention to detail and richness in theirrawings, as well as their tendency to organize drawings in a frontalay. In contrast, the lower levels of maturity and investment inrawing among boys, their avoidance of detail in their drawings,nd their tendency to include fantasy and imaginary themes ando use a perspective from above may explain their higher scores onhe negative global scales (Lijima, Arisaka, Minamoto, & Arai, 2001;
ilne & Greenway, 1999; Richards & Ross, 1967; Turgeon, 2008).Finally, counter to the prevalent assumption in attachment
esearch regarding the standard distribution of attachment secu-ity, the sample in the current study revealed many ambivalentarticipants, according to their family drawings. This finding is
n line with previous studies, mostly conducted elsewhere thann European and North American countries, suggesting a pat-ern of cross-cultural differences, in which the C (ambivalent)lassification is relatively more frequent in collectivistic soci-ties such as Israel (Leyendecker, Lamb, Fracasso, Schoelmerich,
Larson, 1997; Takahashi, 1990; Tomlinson, Cooper, & Murray,005; van IJzendoorn & Kroonenberg, 1988; Zevalkink, Riksen-alraven, & Van Lieshout 1999). Researchers have attributed
his unique distribution to interdependent societal values, whichre often associated with more maternal intrusiveness and over-nvolvement, resulting in resistant infant behaviors (Leyendeckert al., 1997; Zevalkink et al., 1999).
linical implications
The BND holds promise for providing valuable informationor therapists involved in assessing and treating children withttachment problems to improve their adjustment and alleviateheir distress. This is especially important given the prominentole of attachment security in promoting or hampering children’sdjustment (Weinfield et al., 2008), and in buffering the nega-ive consequences of adverse experiences due to deficits in theevelopment of children’s self-system processes that include highelf-esteem and a sense of agency, optimism and perceived compe-ence (Kim & Cicchetti, 2006; Masten, 2001; Masten & Coatsworth,998; Mikulincer & Shaver, 2007).
Furthermore, a growing body of evidence in the field of psy-hotherapy suggests that the quality of attachment bonds betweenhildren and their attachment figures may be generalized to otheryadic relationships such as psychotherapy and may play a crit-
cal role in psychotherapy processes and outcomes. According tohis perspective, the therapeutic relationship is believed to mirrorssential features of an attachment relationship, with the therapistunctioning as the secure base from which clients explore mean-ngful personal material (Mikulincer & Shaver, 2007). In addition,imilar to early attachment figures, therapists may serve as sec-ndary attachment figures as they are perceived by their patientso be wiser and stronger, to provide help and emotional regulationn times of distress, and to be dependably available and respon-ive to needs, and thus functioning as a secure base for explorationBowlby, 1988; Dozier & Tyrrell, 1998; Mikulincer & Shaver, 2007).herefore, combined with other clinical data, attachment orien-ation may shed light on how the therapeutic relationship mightevelop and how best to establish a secure relationship in treat-ent (Byng-Hall, 2001; Slade, 1999).
imitations and future directions
Several limitations of this study should be acknowledged. First,he attachment inventory examined security in general, and mayave been biased toward positive presentation. Specifying the typef insecurity might help clarify the associations between the BND
erapy 41 (2014) 391–399
indicators and scales and type of attachment. Second, the samplesize was relatively small; therefore generalizing from this studyrequires future corroboration. Moreover due to the small size noclassification into the four types of attachment security was madeon the basis of the children’s BNDs. Third, the correlations in thecurrent study were relatively moderate; hence the effects sizeswere rather small. The different assessment methods (projectivemeasures versus self-questionnaire) might account for the mag-nitude of associations. Furthermore, the sample includes rathernormative elementary school children. A clinical sample might leadto larger and more clear-cut effect sizes. Additionally, other aspectssuch as children’s temperament, cognitive abilities and generalfunctioning may moderate these associations. Future researchshould take these characteristics in account when testing theassociations. Furthermore, the study was conducted in the Israelicultural context, which is characterized by high family values andclose family ties (Lavee & Katz, 2003). The relationships betweenthe indicators and especially the large number of ambivalent chil-dren as seen in the family drawings may be different in childrenfrom different cultures. Thus, as suggested by Betts (2013), it mightbe the task of future studies to explore whether these findings alsoapply to other cultures and contexts. Finally, most of the securedrawings were made by girls. Future studies should reexamine thefindings and classifications to enable further generalizations.
Appendix 1.
Intercorrelations between BND scales, controlling for gender
Ainsworth, M. D. S. (1990). Epilogue: Some considerations regarding theory andassessment relevant to attachments beyond infancy. In M. T. Greenberg, D. Cic-chetti, & E. M. Cummings (Eds.), Attachment in the preschool years (pp. 463–488).Chicago, IL: University of Chicago Press.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment.Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Armsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and peer attach-ment: Individual differences and their relationship to psychological well-beingin adolescence. Journal of Youth and Adolescence, 16, 427–454.
Bauminger, N., & Kimhi-Kind, I. (2008). Social information processing, security ofattachment, and emotion regulation in children with learning disabilities. Jour-nal of Learning Disabilities, 41, 315–332.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults:A test of a four-category model. Journal of Personality and Social Psychology, 61,221–244.
Behrens, K., & Kaplan, N. (2011). Japanese children’s family drawings and their linkto attachment. Attachment & Human Development, 13, 437–450.
Belsky, J., & Cassidy, J. (1994). Attachment: Theory and evidence. In M. Rutter, &D. Hay (Eds.), Development through life: A handbook for clinicians (pp. 373–402).Oxford: Blackwell.
Betts, D. J. (2013). A Review of the principles for culturally appropriate art therapyassessment tools. Art Therapy, 30, 98–106 (Journal of the American Art TherapyAssociation).
Bretherton, I., & Munholland, K. A. (2008). Internal working models in attachmentrelationships. In J. Cassidy, & P. Shaver (Eds.), Handbook of attachment: Theory,research and clinical applications (pp. 89–111). New York, NY: Guilford Press.
Bowlby, J. (1969/1982). Attachment and loss. Attachment (vol. 1) New York, NY: BasicBooks.
Bowlby, J. (1973). Attachment and loss. Separation: Anxiety and anger (vol. 2) NewYork, NY: Basic Books.
Bowlby, J. (1979). The making and breaking of affectional bonds. London: Tavistock.Byng-Hall, J. (2001). Attachment as a base for family and couple therapy. Child and
Adolescent Mental Health, 6, 31–36.Bowlby, J. (1980). Attachment and loss. Loss, sadness and depression (vol. 3) New York,
NY: Basic Books.Bowlby, J. (1988). A secure base: Parent–child attachment and healthy human devel-
arlson, E. A., Sroufe, L. A., & Egeland, B. (2004). The construction of experience: Alongitudinal study of representation and behavior. Child Development, 75, 66–83.
assidy, J. (2001). Truth, lies, and intimacy: An attachment perspective. Attachment& Human Development, 3, 121–155.
icchetti, D., Toth, S., & Bush, M. (1988). Developmental psychopathology and incom-petence in childhood: Suggestions for intervention. In B. B. Lahey, & A. E. Kazdin(Eds.), Advances in clinical child psychology (vol. 11) (pp. 1–71). New York, NY,US: Plenum Press.
ollins, W. A., & Sroufe, L. A. (1999). Capacity for intimate relationships. The devel-opment of romantic relationships in adolescence. In W. Furman, B. B. Brown,& C. Feiring (Eds.), The development of romantic relationships in adolescence (pp.125–147). New York, NY: Cambridge University Press.
allaire, D. H., Ciccone, A., & Wilson, L. C. (2012). The family drawings of at-riskchildren: Concurrent relations with contact with incarcerated parents, caregiverbehavior, and stress. Attachment & Human Development, 14, 161–183.
eaver, S. P. (2009). A normative study of children’s drawings: Preliminary researchfindings. Art Therapy, 26, 4–11 (Journal of the American Art Therapy Association).
ozier, M., & Kobak, R. R. (1992). Psychophysiology in attachment interviews: Con-verging evidence for deactivating strategies. Child Development, 63, 1473–1480.
ozier, M. A., & Tyrrell, C. (1998). The role of attachment in therapeutic relationships.In J. A. Simpson, & W. A. Rholes (Eds.), Attachment theory and close relationships(pp. 221–248). New York, NY: Guilford Press.
innegan, R. A., Hodges, E. V., & Perry, D. G. (1996). Preoccupied and avoidant copingduring middle childhood. Child Development, 67, 1318–1328.
rancis, D., Kaiser, D., & Deaver, S. (2003). Representations of attachment security inthe Bird’s Next Drawings of clients with substance abuse disorders. Art Therapy,20, 124–137 (Journal of the American Art Therapy Association).
ury, G., Carlson, E. A., & Sroufe, L. A. (1997). Children’s representations of attachmentrelationships in family drawings. Child Development, 68, 1154–1164.
oldner, L., & Levi, M. (2014). Children’s family drawings, body perceptions, andeating attitudes: The moderating role of gender. The Arts in Psychotherapy, 41,79–88.
oldner, L., & Scharf, M. (2011). Children’s family drawings: A study of attachment,personality, and adjustment. Art Therapy, 28, 11–18.
oldner, L., & Scharf, M. (2012). Children’s family drawings and internalizing prob-lems. The Arts in Psychotherapy, 39, 262–271.
raham, C. A., & Easterbrooks, M. A. (2000). School-aged children’s vulnerability todepressive symptomatology: The role of attachment security, maternal depres-sive symptomatology, and economic risk. Development and Psychopathology, 12,201–213.
reenberg, M. T., & Speltz, M. L. (1988). Contributions of attachment theory to theunderstanding of conduct problems during the preschool years. In J. Belsky, &T. Nezworski (Eds.), Clinical implications of attachment (pp. 177–218). Hillsdale,NJ: Lawrence Erlbaum Associates, Inc.
reenberg, M. T., Speltz, M. L., Deklyen, M., & Endriga, M. C. (1991). Attachmentsecurity in preschoolers with and without externalizing problems: A replication.Development and Psychopathology, 3, 413–430.
ranot, D., & Mayseless, O. (2001). Attachment security and adjustment toschool in middle childhood. International Journal of Behavioral Development, 25,530–541.
arter, S. (1985). Supplementary description of the self-perception for children revisionof the perceived competence scale for children. University of Denver (Unpublishedmanuscript).
aiser, D. H. (1996). Indications of attachment security in a drawing task. The Arts inPsychotherapy, 23, 333–340.
aiser, D. H., & Deaver, S. (2009). Assessing attachment with the Bird’s Nest Drawing:A review of the research. Art Therapy, 26, 26–33.
aplan, N., & Main, M. (1986). Instructions for the classification of children’s familydrawings in terms of representation of attachment. Berkeley, CA: University ofCalifornia.
erns, K. A., Abraham, M. M., Schlegelmilch, A., & Morgan, T. A. (2007). Mother–childattachment in later middle childhood: Assessment approaches and associationswith mood and emotion regulation. Attachment and Human Development, 9,33–53.
erns, K. A., Aspelmeier, J. E., Gentzler, A. L., & Grabill, C. M. (2001). Parent–childattachment and monitoring in middle childhood. Journal of Family Psychology,15, 69–81.
erns, K. A., Kelpac, L., & Cole, A. (1996). Peer relationship and preadolescents’ per-ceptions of security in child mother relationship. Developmental Psychology, 32,457–466.
erns, K. A., & Stevens, A. C. (1995). Parent–child attachment in late adolescence:Links to social relations and personality. Journal of Youth and Adolescence, 25,323–342.
erns, K. A., Tomich, P. L., Aspelmeier, J. E., & Contreras, J. M. (2000).
Attachment-based assessments of parent–child relationships in middle child-hood. Developmental Psychology, 36, 614–626.
erns, K. A., Tomich, P. L., & Kim, P. (2006). Normative trends in children’s perceptionsof availability and utilization of attachment figures in middle childhood. SocialDevelopment, 15, 1–22.
erapy 41 (2014) 391–399 399
Kim, J., & Cicchetti, D. (2006). Longitudinal trajectories of self-system processesand depressive symptoms among maltreated and nonmaltreated children. ChildDevelopment, 77, 624–639.
Koppitz, E. M. (1968). Psychological evaluation of children’s human figure drawing.New York, NY: Grune & Stratton.
Lavee, Y., & Katz, R. (2003). The family in Israel: Between tradition and modernity.Marriage & Family Review, 35, 193–217.
Leyendecker, B., Lamb, M. E., Fracasso, M. P., Schoelmerich, A., & Larson, C. (1997).Playful interaction and the antecedents of attachment: A longitudinal study ofCentral American and Euro-American mothers and infants. Merrill-Palmer Quar-terly, 43, 24–47.
Lieberman, M., Doyle, A., & Markiewicz, D. (1999). Developmental patterns in secu-rity of attachment to mother and father in late childhood and early adolescence:Associations with peer relations. Child Development, 70, 202–213.
Lijima, M., Arisaka, O., Minamoto, F., & Arai, Y. (2001). Sex differences in children’sfree drawings: A study on females with congenital adrenal hyperplasia. Hor-mones and Behavior, 40, 99–104.
Lyons-Ruth, K. (1996). Attachment relationships among children with aggressivebehaviour problems: The role of disorganised early attachment patterns. Journalof Consulting and Clinical Psychology, 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. Attach-ment and Human Development, 5, 19–37.
Main, M. (1996). Introduction to the special section on attachment and psy-chopathology: 2. Overview of the field of attachment. Journal of Consulting andClinical Psychology, 64, 237.
Main, M., & Solomon, I. (1990). Procedures for identifying infants as disorga-nized/disorientated during the Ainsworh Strange Situation. In M. Greenberg,D. Cicchetti, & E. M. Cummings (Eds.), Attachment in the preschool years: Theory,research and intervention (pp. 121–160). Chicago: University of Chicago Press.
Marvin, R. S., & Britner, P. A. (1999). Normative development: The ontogeny ofattachment. In J. Cassidy, & P. R. Shaver (Eds.), Handbook of attachment: Theory,research, and clinical applications (pp. 44–67). New York, NY: Guilford Press.
Masten, A. S. (2001). Ordinary magic: Resilience processes in development. AmericanPsychologist, 56, 227–238.
Masten, A. S., & Coatsworth, J. D. (1998). The development of competence in favorableand unfavorable environments: Lessons from research on successful children.American Psychologist, 53, 205–220.
Mikulincer, M., & Shaver, P. R. (2006). Attachment in adulthood: Structure, dynamics,and change. New York, NY: Guilford Press.
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamicsand change. New York, NY: Guilford Press.
Milne, L. C., & Greenway, P. (1999). Color in children’s drawings: The influence ofage and gender. The Arts in Psychotherapy, 26, 261–263.
NICHD Early Child Care Research Network. (2006). Infant–mother attachment classi-fication: Risk and protection in relation to changing maternal caregiving quality.Developmental Psychology, 42, 38–58.
Pianta, R. C., Longmaid, K., & Ferguson, K. E. (1999). Attachment-based classificationsof children’s family drawings: Psychometric properties and relations with chil-dren’s adjustment in kindergarten. Journal of Clinical Psychology, 28, 244–255.
Scott, L. H. (1981). Measuring intelligence with the Goodenough–Harris Drawingtest. Psychological Bulletin, 89, 483–505.
Sheller, S. (2007). Understanding insecure attachment: A study using children’s birdnest imagery. Art Therapy, 24, 119–127 (Journal of the American Art TherapyAssociation).
Slade, A. (1999). Attachment theory and research: Implications for the theory andpractice of individual psychotherapy with adults. In J. Cassidy, & P. R. Shaver(Eds.), Handbook of attachment: Theory, research, and clinical applications (pp.575–594). New York, NY, US: Guilford Press.
Ranson, K. E., & Urichuk, L. J. (2008). The effect of parent–child attachment relation-ship on child biopsychological outcomes: A review. Early Child Development andCare, 178, 129–152.
Richards, M. P., & Ross, H. E. (1967). Developmental changes in children’s drawings.British Journal of Educational Psychology, 37, 73–80.
Takahashi, K. (1990). Are the key assumptions of the ‘strange situation’ procedureuniversal? A view from Japanese research. Human Development, 33, 23–30.
Tomlinson, M., Cooper, P., & Murray, L. (2005). The mother–infant relationship andinfant attachment in a South African peri-urban settlement. Child Development,76, 1044–1054.
Turgeon, S. M. (2008). Sex differences in children’s free drawings and their relation-ship to 2D:4D ratio. Personality and Individual Differences, 45, 527–532.
van IJzendoorn, M. H., & Kroonenberg, P. M. (1988). Cross-cultural patterns of attach-ment: A meta-analysis of the strange situation. Child Development, 59, 147–156.
Weinfield, N. S., Sroufe, A. L., Egeland, B., & Carlson, E. (2008). Individual differences
in infant–caregiver attachment: Conceptual and empirical aspects of security.In J. Cassidy, & P. R. Shaver (Eds.), Handbook of attachment: Theory, research andclinical applications (2nd ed., vol. 11, pp. 78–100). New York, NY: Guilford Press.
Zevalkink, J., Riksen-Walraven, J. M., & Van Lieshout, C. F. M. (1999). Attachment inthe Indonesian care-giving context. Social Development, 8, 21–40.
