PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [Eastern Michigan University]On: 10 June 2011Access details: Access Details: [subscription number 932521369]Publisher RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Clinical Child & Adolescent PsychologyPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t775648094

The Effects of Domestic Violence on the Stability of Attachment fromInfancy to PreschoolAlytia A. Levendoskya; G. Anne Bogata; Alissa C. Huth-Bocksb; Katherine Rosenblumc; Alexander vonEyea

a Department of Psychology, Michigan State University, b Department of Psychology, EasternMichigan University, c Center for Human Growth and Development, University of Michigan,

Online publication date: 29 April 2011

To cite this Article Levendosky, Alytia A. , Bogat, G. Anne , Huth-Bocks, Alissa C. , Rosenblum, Katherine and von Eye,Alexander(2011) 'The Effects of Domestic Violence on the Stability of Attachment from Infancy to Preschool', Journal ofClinical Child & Adolescent Psychology, 40: 3, 398 — 410To link to this Article: DOI: 10.1080/15374416.2011.563460URL: http://dx.doi.org/10.1080/15374416.2011.563460

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

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

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould 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 directlyor indirectly in connection with or arising out of the use of this material.

The Effects of Domestic Violence on the Stability ofAttachment from Infancy to Preschool

Alytia A. Levendosky and G. Anne Bogat

Department of Psychology, Michigan State University

Alissa C. Huth-Bocks

Department of Psychology, Eastern Michigan University

Katherine Rosenblum

Center for Human Growth and Development, University of Michigan

Alexander von Eye

Department of Psychology, Michigan State University

We hypothesized that trajectories of domestic violence (DV), maternal depression, andhousehold income (from pregnancy to age 4) would be differentially associated withinstability and stability of attachment, as measured by the Strange Situation at ages 1and 4. Participants were 150 women and children. Women were first assessed duringpregnancy and then yearly when the children were 1 to 4 years old. Overall, attachmentwas unstable for 56% of the sample from age 1 to age 4. Trajectories of DV and incomeboth predicted attachment patterns. Positive outcomes (secure-secure and insecure-secure) were related to initially low levels of DV that stayed constant or became loweras well as initially high or low levels of income that increased over time.

The focus of the current research is to better understanddevelopmental pathways that result in stability or insta-bility in child–mother attachment in a heterogeneoushigh-risk sample selected to overrepresent domestic viol-ence (DV), defined as male violence against his femaleromantic partner. Attachment theory predicts stabilityof attachment in individuals over time while allowing thatenvironmental factors may, on occasion, influence achange in attachment quality (Bowlby, 1969=1982).

Infant attachment is usually assessed in early childhoodby administering the Strange Situation Procedure(Ainsworth, Blehar, Waters, & Wall, 1978), a 22-minparadigm during which the mother and child experiencetwo separations and reunions. Briefly, child attachmentis classified into secure and insecure categories, with afurther subdivision of insecure into avoidant, ambivalent,and disorganized categories. Securely attached babies aredistressed at the mother’s departure but are easilysoothed upon her return (Carlson & Sroufe, 1995). Avoi-dant babies typically do not show a reaction to themother’s departure and appear indifferent to her return.Ambivalent babies are highly distressed at the mother’sdeparture and are very difficult to soothe when shereturns. Children are classified as disorganized when theydo not show a consistent strategy in the reunions withthe mother. They also frequently show odd behaviors,such as freezing and disorientation (Van IJzendoorn,Schuengel, & Bakermans-Kranenburg, 1999).

This study was supported by grants from the National Institute of

Justice (8–7958-MI-IJ) and the Centers for Disease Control (RO1=

CCR518519–01) to the first, second, and fifth authors. We thank the

Mother–Infant Study research assistants for the considerable time

and energy they devoted to the challenging data collection, as well as

Susan Paris and Bonnie Conley for coding the Strange Situation

videotapes.Correspondence should be addressed to Alytia A. Levendosky,

107C Psychology Building, Michigan State University, East Lansing,

MI 48824. E-mail: levendo1@msu.edu

Journal of Clinical Child & Adolescent Psychology, 40(3), 398–410, 2011

Copyright # Taylor & Francis Group, LLC

ISSN: 1537-4416 print=1537-4424 online

DOI: 10.1080/15374416.2011.563460

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

In young children, Bowlby considered attachment tobe sensitive to environmental changes, but later in child-hood, attachment quality may become more resistant tochange. Empirical support for this supposition has beenfound (Fraley, 2002). However, longitudinal researchhas not yet determined the interplay of individual andcontextual factors that might influence the stabilityand instability of early child–mother attachment.Understanding the factors that can affect attachmentstability is critical to designing interventions that caneither increase the likelihood that children with insecureattachment move toward security or help children tomaintain stability of secure attachment.

To date, longitudinal research on attachment hasexamined only two time periods in a child’s life. Studiesof attachment in low risk populations have reported1-year stability rates ranging from 53% to 78% (NICHDEarly Child Care Research Network [ECCRN], 2001;Owen, Easterbrooks, Chase-Lansdale, & Goldberg,1984; Thompson, Lamb, & Estes, 1982). In thesestudies, the stability is highest for the secure group.Similarly, studies of attachment in high-risk andheterogeneous-for-risk populations have also reportedstability rates that exceed chance, ranging from 50to 60%, particularly when the two assessments arewithin 1 year (Barnett, Ganiban, & Cicchetti, 1999;Easterbrooks, 1989; Egeland & Farber, 1984; Vondra,Hommerding, & Shaw, 1999).

Individual characteristics of the mother, the mother’srelationship with her partner, and the demographiccharacteristics of the family and the child are key predic-tors of stability or instability of attachment (Egeland &Farber, 1984; Moss, Cyr, Bureau, Tarabulsy, &Dubois-Comtois, 2005; NICHD ECCRN, 2001; Vondraet al., 1999). These findings are consistent with Bowlby’stheory, as well as a developmental psychopathology per-spective, which suggests that children’s behavior resultsfrom the influence of numerous factors existing atmultiple levels of the ecology and not from one specificfactor (Finkelhor & Kendall-Tacket, 1997; Masten,Best, & Garmezy, 1990).

However, extant studies of stability=instability ofattachment assess contextual predictors and attachmentstatus concurrently. This methodology fails to measurethese same contextual predictors prior to the first assess-ment of attachment (typically at about age 1). As wedescribe later, research indicates that prior maternal his-tory, particularly during pregnancy, can have a stronginfluence on the mother–child relationship. In addition,the concurrent measurement of criterion and predictorsat two periods (e.g., ages 1 and 4) can only assesswhether a linear change in predictors is related toattachment stability=instability. Neither the trajectoryof child development nor the trajectories of contextualfactors that occur with it are likely to be linear. The

inclusion of data from predictors at five time points(during pregnancy and throughout early childhood)allowed us to assess, in addition to linear change, otherpatterns of change. We also determined whether theselinear or nonlinear patterns of change were related tolinear or nonlinear patterns of change in the predictorsof attachment stability=instability. We examined threecontextual predictors (DV, household income, andmaternal depressive symptoms) that have been pre-viously associated with instability of attachment as wellas with problematic child outcomes associated with chil-dren’s exposure to DV. As reviewed next, these risk fac-tors have been associated with attachment concurrentlyor have predicted attachment at a later period. How-ever, there is no research examining trajectories ofattachment and their relationship to trajectories of theserisk factors.

DV AND ATTACHMENT

Bowlby noted that the attachment system becomes acti-vated as the child experiences fear when confronted withtwo specific threats: a threat to the caregiver or a threatin the child’s environment. Kobak, Cassidy, and Ziv(2004) identified four specific trauma-related threats toattachment, one of which is ‘‘physical or sexual abuseof a parent.’’ Thus, the child’s exposure to DV is a stressthat can have consequences for attachment.

Recent research suggests that stress experienced bythe pregnant mother might influence the child in utero.While the mother is pregnant, the trauma itself, as wellas the biological correlates of maternal stress, may affectbrain development and thus permanently influence thechild’s responses to stress after birth (see Schore, 2003).

Research finds that women who experienced DVduring their pregnancies had significantly more negativecaregiving schemas of their unborn babies (Huth-Bocks,Levendosky, Bogat, & von Eye, 2004) and were morelikely to be classified as nonbalanced (similar to insecure)than those who did not experience pregnancy DV(Huth-Bocks, Levendosky, Theran, & Bogat, 2004). Inaddition, prenatal representations predict observedmaternal parenting at age 1 (Dayton, Levendosky,Davidson, & Bogat, 2010).

Postpartum, DV continues to play a role in theorganization of child attachment. When women experi-enced DV prenatally and postpartum or experienced DVfor the first time during the 1st year of their children’slives, their children were likely to either maintaininsecure attachments or develop them by age 4. How-ever, women who left the abusive relationship after thechild was born were more likely to demonstratebalanced postpartum maternal representations, and thechild was then more likely to become securely attached

DOMESTIC VIOLENCE ON STABILITY OF ATTACHMENT 399

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

by age 4 (Levendosky, Bogat, Huth-Bocks, &Rosenblum, 2007).

It is important to note than DV is generally an epi-sodic experience. Women often move in and out of viol-ent relationships (e.g., Bogat, Levendosky, Theran, vonEye, & Davidson, 2003; Fleury, Sullivan, Bybee, &Davidson, 1998), and violence within relationships ebbsand flows (Feld & Straus, 1989; Follingstad, Hause,Rutledge, & Polek, 1992; Wofford, Mihalic, & Menard,1994). This could affect children in two ways: (a)Women’s parenting sensitivity may fluctuate dependingon whether DV is present or absent at any given time,and (b) children’s exposure to DV may be consistentor intermittent. In the latter case, Martinez-Torteya,Bogat, von Eye, and Levendosky (2009) found that con-sistent DV exposure, rather than intermittent exposure,predicted internalizing and externalizing problems inchildren.

MATERNAL DEPRESSIVE SYMPTOMS ANDATTACHMENT

Parental depression has deleterious effects on parentingand children’s functioning (e.g., Lyons-Ruth, Lyubchik,Wolfe, & Bronfman, 2002; Rogosch, Mowbray, &Bogat, 1992; Sameroff, Seifer, & Zax, 1982). However,not all children are negatively affected (see NICHDECCRN, 2006). Depression may contribute to less opti-mal caregiving behaviors by reducing mothers’ attentionand interest in their children (e.g., Gelfand & Teti, 1990;Zahn-Waxler Iannotti, Cummings, & Denham, 1990),thus decreasing maternal availability to assist the childwith emotion regulation (e.g., Cummings & Cicchetti,1990; Zahn-Waxler et al., 1990). Maternal depressionmay have stronger effects on young children, comparedto older children, because of their nearly total depen-dence on the caretaker (Beardslee, Bemporad, Keller,& Klerman, 1983; Cummings & Cicchetti, 1990).

Further, maternal depression is associated withinsecure infant–mother attachment (Lovejoy, Graczyk,O’Hare, & Neuman, 2000; Lyons-Ruth et al., 2002),the stability of insecure attachment among infants(Edwards, Eiden, & Leonard, 2004), and movementtoward insecurity from infancy to early adulthood(Weinfield, Whaley, & Egeland, 2004). DV is also asso-ciated with elevated depressive symptoms in women(e.g., Bogat et al., 2003; Kessler, Molnar, Feurer, &Appelbaum, 2001), but the symptoms are related to tim-ing and duration of DV. Bogat, Levendosky, DeJonghe,Davidson, and von Eye (2004) found elevated depressivesymptoms for both abused and nonabused women dur-ing pregnancy, which, for most women, then diminishand plateau in the first 2 years of the child’s life. How-ever, for women who consistently experience abuse

during this time, symptoms again worsened by theirchildren’s 2nd year of life.

HOUSEHOLD INCOME AND ATTACHMENT

Poverty, which often co-occurs with DV (Tolman &Raphael, 2000), has been linked to increased risk forinsecure attachment (e.g., Bakermans-Kranenburg, vanIJzendoorn, & Kroonenberg, 2004; Egeland & Sroufe,1981; NICHD ECCRN, 2001; Van IJzendoorn et al.,1999). The ‘‘family stress model’’ posits that socioeco-nomic disadvantage affects parenting practices, whichthen influence children’s development (Bakermans-Kranenburg et al., 2004; Conger, Conger, Elder, &Lorenz, 1992, 1993). Poverty has been associated withless consistent, attentive caregiving and less emphasison infant needs compared to personal or familial needs(Aber, Jones, & Cohen, 2000; Cain & Combs-Orme,2005). Recent results generally support this conclusion,but they also suggest that the effects of socioeconomicstatus (SES) on children’s attachment may not beentirely mediated through diminished parenting sensi-tivity (NICHD ECCRN, 2001).

CURRENT STUDY

In the current study, we examined whether maternaldepressive symptoms, household income, and DV pre-dicted patterns of stability and instability of attachmentfrom age 1 to 4 in a heterogeneous-for-risk sample, cho-sen to overrepresent the incidence of DV among preg-nant women. We hypothesized that trajectories relatedto the timing and duration of several contextual vari-ables (DV, maternal depressive symptoms, and lowincome) would be related to patterns of stability andinstability of attachment classifications. Based on find-ings from prior literature examining two periods, weexpected that increasing levels of risk factors would beassociated with stability of insecure attachment ormovement toward insecure attachment. We alsoexpected that decreasing levels of risk factors would beassociated with stability of secure attachment or move-ment toward secure attachment. However, becausewomen move in and out of DV relationships, andbecause of the different amounts of time that theyexperience (and their children are exposed to) DV, itwas important to run models that allowed for more thandifferential linear slopes. Thus, we accommodated fluc-tuations over time in the independent variables byallowing for nonlinear curvatures. However, becausethis is a first exploration of attachment pattern trajec-tories, there was no current theory or empirical research

400 LEVENDOSKY ET AL.

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

to guide a priori predictions about the exact shape ofthese trajectories.

METHODS

Participants

Participants were 150 women and their target childrenenrolled in a larger longitudinal study (N¼ 206) examin-ing risk and resilience factors for women and childrenexperiencing DV. Women and children were includedin this research only if they had completed the StrangeSituation at both ages 1 and 4. The 56 women whodid not complete the Strange Situation included 18who either refused participation or could not be locatedin the age 1 data collection wave, 15 who either refusedparticipation or could not be located in the age 4 datacollection wave, and 23 who were not able to be inter-viewed in project offices at either the age 1 or age 4 datacollection wave. The participants did not differ from thenonparticipants in marital status, education, race=ethnicity, receipt of public assistance, or level of prenatalDV. Participants had significantly higher income duringpregnancy than those who did not participate,t(202)¼ 2.30, p< .05.

The participating women were 62% Caucasian, 25%African American, 5% Latina, 5% biracial, and 3% ofother ethnic=racial backgrounds. During pregnancy,47% were single, 43% were married, and 10% wereseparated=divorced=widowed. Almost half (42%) hadcompleted high school or less, 44% had completed somecollege or trade school, and 14% had a college or gradu-ate degree. The participants were comparable on eth-nicity and education level to the population statisticsfrom the two counties where 93% of the sample wasrecruited (U.S. Census Bureau, 2000). However, the dis-tribution of participants’ income levels was different,such that 57% of the participants had incomes less than$20,000 annually, whereas only 31% of the women inthese two counties had incomes in that range. Therewere 72 participating boys and 78 participating girls.Forty-seven percent of children were Caucasian; 25%African American; 24% biracial; and 4% NativeAmerican, Asian, or Latino=a.

Procedures

The women for the larger study were recruited byplacing flyers announcing the study in various sites inthe community that served women (e.g., OB=GYNclinics). There were two types of flyers—one recruitingspecifically for participants experiencing DV and theother recruiting for a mother–baby study. To partici-pate, women had to be between the ages of 18 and 40,

pregnant at the time of the first interview, involved ina romantic relationship of at least 6 weeks duration dur-ing the pregnancy, and fluent in English. Women whowere interested in participating telephoned projectoffices and were then screened for inclusion criteria.After approximately half of the sample was recruited(n¼ 96), we began exclusively using the flyer recruitingwomen who had experienced DV as well as conductingtelephone screening with the Conflict Tactics Scale(Straus, 1979) to increase the number of pregnantwomen experiencing DV in our sample.

Women were first interviewed during their third tri-mester of pregnancy; subsequent interviews occurredyearly at the time of the children’s birthdays. Informedconsent was obtained for each woman and her targetchild at each wave of data collection, whereas assentto participate was obtained from each child at age 4.Institutional Review Board approval for all study proce-dures was obtained from the relevant institutions priorto recruitment. For the 150 participants in the presentstudy, 87 women experienced DV during their pregnan-cies; 63 women did not.

Measures

Demographics. Demographic information, includ-ing race=ethnicity, maternal age, maternal years of edu-cation, household income, and gender of the child, wereobtained. We analyzed household income because ofthe important relationship between it and children’soutcomes (e.g., NICHD ECCRN, 2001).

Beck Depression Inventory (BDI; Beck, Ward,Mendelson, Mock, & Erbaugh, 1961). This measureincludes items about a variety of symptoms such asmood, sleep, and appetite disturbances. Each item israted on a 0-to-3 scale, and all items are then summed.Higher scores indicate greater symptom severity. Thereis well-established reliability and validity of the BDI(see review by Beck, Steer, & Garbin, 1988). In the cur-rent study, the coefficient alphas were .85, .85, .87, .90,and .90, respectively, at the five waves. Scores of 12and above are considered indicators of mild to moderatedepression (Beck et al., 1961). At each wave of datacollection, 20 to 30% of the women reported scoresabove 12; the mean scores at each wave varied from5.74 to 10.57.

Strange Situation Procedure (SS; Ainsworth et al.,1978). The SS was administered when the childrenwere age 1 and age 4. The SS is a 22-min laboratory pro-cedure in which the child and mother participate in eightepisodes of play, separation, and reunion. The seriesof episodes is considered to be mildly to moderately

DOMESTIC VIOLENCE ON STABILITY OF ATTACHMENT 401

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

stressful for the young child and, thus, induces parti-cular attachment behaviors, such as crying, clinging,and withdrawal. At age 1, child behaviors are codedon four 7-point scales: proximity seeking, contact main-tenance, avoidance, and resistance (Ainsworth et al.,1978) and one 9-point scale for disorganization (Main& Solomon, 1986). For the age 4 SS, child behavior iscoded on three scales (a 9-point scale of security, a7-point scale of avoidance, and a 9-point scale ofdisorganization; Cassidy, Marvin, & the MacArthurWorking Group, 1992) based on the following behavior:proximity=contact seeking, body orientation, speech,gaze, and affect. Children at both ages were classifiedinto one of the following attachment categories: secure,insecure-avoidant, insecure-ambivalent, or disorganized=controlling. Disorganized children were grouped withthe two insecure categories for analyses.

Some researchers have suggested that the duration ofseparation in the SS may not be long enough to activatethe attachment system in preschoolers (Solomon &George, 1999). However, the SS has been widely usedwith this age group, with interrater reliability rangingfrom 75 to 92% (Bar-Haim, Sutton, Fox, & Marvin,2000; Bretherton, Ridgeway, & Cassidy, 1990; Marvin& Pianta, 1996; NICHD ECCRN, 2001; Shouldice &Stevenson-Hinde, 1992). Secure and insecure classifica-tions at age 4 are related to classifications in other pre-school paradigms assessing attachment representations(Bretherton et al., 1990: 75% concordant, p< .01) as wellas showing expected relationships with maternal beha-vior, child behavior, mother–child interaction, and his-tory of maltreatment (Cicchetti & Barnett, 1991; Moss,Bureau, Cyr, St-Laurent, & Mongeau, 2004; NICHDECCRN, 2001; Stevenson-Hinde & Shouldice, 1995).Indeed, later assessments of attachment may providestronger information regarding child risk for subsequentproblematic outcomes; a recent meta-analysis of 69studies indicated a larger effect size for preschool (vs.infant) SS attachment classifications in relation tochildhood externalizing behavior problems (Fearon,Bakermans-Kranenburg, Van IJzendoorn, Lapsley, &Roisman, in press).

SS videotapes were coded at the University ofWashington. At age 1, there was a 90% agreement rate(k¼ .84, p< .001), based on 11% double-coding. Forage 4, the agreement rate was 76% (k¼ .53, p< .001),based on 35% double-coding. Differences in classifica-tions were settled through conferencing.

Severity of Violence Against Women Scales(SVAWS; Marshall, 1992). The SVAWS is a 46-itemquestionnaire that measures threats of violence, actualphysical violence, and sexual violence ranging from mildto severe during the past year. Threats of violence areincluded in this measure because they are a form of

psychological abuse. Men are often able to control theirpartners through threats without the need to enactphysical violence. Both physical and psychologicalabuse negatively affect women’s mental health and func-tioning (Baldry, 2003; Follingstad, 2009; Levendosky &Graham-Bermann, 2000; Morland, Leskin, Block,Campbell, & Friedman, 2008). Participants indicatedhow often they experienced these events based on a4-point scale ranging from never to many times, and allitems are then summed. Women completed the ques-tionnaire during pregnancy and on the children’s birth-days (ages 1, 2, 3, and 4). Dutton, Pianta, and Marvin(2001) found that women’s reports of abuse on theSVAWS correlated with their partner’s reports. In oursample, the coefficient alphas were .95, .95, .95, .99,and .94, respectively at the five waves.

RESULTS

Missing Data

Missing data were imputed using the ‘‘hot-deckmethod’’ (Lisrel 8.8; Joreskog & Sorbom, 2006). Thismethod was used because it substitutes real values formissing values, obtaining the real values from anotherparticipant whose responses on specified variables mostclosely match those of the participant with missingdata. One advantage of this procedure is that thecorrelations among the predictor variables across timewill not be artificially increased (see Table 1). Fourwomen were missing SVAWS data at one period, andfive women were missing household income data atone period.

Attachment Classifications

At age 1, 17.3% of the children were classified as avoi-dant, 56% as secure, 16% as ambivalent, and 10.7% asdisorganized. At age 4, 18.7% of the children were classi-fied as avoidant, 64% as secure, 3.3% as ambivalent, and14% as disorganized. Overall, 44.3% of the childrendemonstrated stability of attachment with their mothersfrom ages 1 to 4. This agreement was not significantlygreater than chance (see Table 2). When collapsed intogroups of secure and insecure classifications (childrenwith disorganized attachment were collapsed into theinsecure category), 57.4% stayed concordant. This wasstill not better than chance (p¼ .07). Due to sample sizeconstraints, the collapsed groups of secure and insecurewere used for the rest of the analyses. Based on classifi-cations of attachment from age 1 and age 4: secure-secure (n¼ 58), secure-insecure (n¼ 26), insecure-secure(n¼ 38), and insecure-insecure (n¼ 28).

402 LEVENDOSKY ET AL.

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

TABLE1

CorrelationsofthePredictorVariablesOverTim

e

Income

Preg.

Income

Age1

Income

Age2

Income

Age3

Income

Age4

DV

Preg.

DV

Age1

DV

Age2

DV

Age3

DV

Age3

Mat.Dep.

Preg.

Mat.Dep.

Age1

Mat.Dep.

Age2

Mat.Dep.

Age3

Mat.Dep.

Age4

IncomePregnancy

1

IncomeAge1

.61�

1

IncomeAge2

.51�

.56�

1

IncomeAge3

.59�

.70�

.61�

1

IncomeAge4

.59��

.62�

.58��

.75�

1

DV

Pregnancy

�.18�

�.28�

�.11

�.33�

�.27�

1

DV

Age1

.00

�.20�

�.20�

�.24�

�.21�

.34�

1

DV

Age2

�.03

�.19�

�.18�

�.24�

�.14

.40�

.38�

1

DV

Age3

�.15

�.11

�.10

�.20�

�.13

.29�

.28�

.23�

1

DV

Age4

.05

�.13

�.15

�.16

�.01

.19�

.28�

.31�

.29�

1

MaternalDepressionPregnancy

�.27�

�.21�

�.25��

�.40�

�.35�

.38�

.06

.12

.38�

.15

1

MaternalDepressionAge1

�.14

�.09

�.07

�.23�

�.16

.16

.12

.07

.27�

.09

.45�

1

MaternalDepressionAge2

�.20�

�.17�

�.16

�.33�

�.33�

.29�

.36�

.27�

.31�

.17�

.46�

.68�

1

MaternalDepressionAge3

�.13

�.11

�.12

�.24�

�.24�

.19�

.30�

.10

.30�

.20�

.43�

.67�

.69�

1

MaternalDepressionAge4

�.20�

�.15

�.05

�.25�

�.21�

.24�

.13

.15

.33�

.23�

.44�

.56�

.58�

.74�

1

Note:Preg.¼

pregnancy;DV¼domesticviolence;Mat.Dep.¼

maternaldepression.

� p<.05.

403

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

Factors Related to Stability and Instability ofAttachment

We chose to conduct repeated measures analyses ofvariance (ANOVAs) rather than structural equation

modeling because of (a) the constraints of our samplesize, and (b) the advantage of examining the differentialpredictability of change in attachment from the longi-tudinal trajectories of the predictors (i.e., DV, income,and maternal depressive symptoms). The trajectoriescould be linear as well as accelerating, decelerating,and changing in acceleration or deceleration. The depen-dent variables were regressed on the relevant covariatesat each time point, and the unstandardized residualsfrom these regressions were then used in the repeatedmeasures analyses.

Prenatal DV was covaried in the analyses that did notinvolve DV as a predictor. In addition, because thewomen were heterogeneous on SES, indices of SES(i.e., maternal level of education and income duringpregnancy) were also covaried in the analyses. However,the analyses that involved income as a predictoronly covaried education and DV. Maternal level of edu-cation and family income were significantly correlated(p¼ .38). However, we included both because they share

TABLE 2

Distribution of Strange Situation Classifications from Age 1 to Age 4

Age 4

Secure Avoidant Ambivalent Disorganized

Total

Age 4

Age 1

Secure 58 14 2 10 84

Avoidant 13 5 1 7 26

Ambivalent 16 5 2 1 24

Disorganized 9 4 0 3 16

Total Age 1 96 28 5 21 150

Note: 44.3% agreement, j¼ .07; v2(9, N¼ 150)¼ 9.76, p¼ .37.

When collapsed into secure=insecure groups, there was 57% agree-

ment, j¼ .12; v2(1, N¼ 150)¼ 2.11, p< .15.

FIGURE 1 Four attachment pattern groups and the trajectory of domestic violence from pregnancy through age 4.

404 LEVENDOSKY ET AL.

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

only 14% of their variances and together they are solidproxy variables of SES. In addition, correlated predic-tors’ contributions are not always diminished in analyseswhen they are simultaneously included in regressionmodels. The pattern of security=insecurity of attachmentwas not related to the demographic variables ofmaternal age, maternal or child ethnicity, or maternalmarital status. However, the pattern was significantlyrelated to gender; thus, gender was included as a factorin each repeated measures ANOVA.

Domestic violence. A repeated measures ANOVAwith polynomial decomposition examined the relation-ship between trajectory of DV and the pattern of attach-ment over time and gender (the pregnancy interviewvariables of maternal education and family incomewere covariates). The covariates were not significantlyrelated to the outcomes. The multivariate tests indicateda significant Time�Pattern interaction, F(12, 150)¼2.01, p< .05, g2¼ .05. The within-subjects tests indi-cated a significant interaction of Time�Pattern,

F(12, 150)¼ 1.79, p< .05, g2¼ .04. There was a signifi-cant quadratic effect of Time�Pattern, F(3, 150)¼5.49, p< .05, g2¼ .09, indicating differential curvaturein the trajectories of these groups. See Figure 1. Thesepatterns show the following: (a) The secure-secure groupbegan with relatively low rates of DV that remained lowover time; (b) the secure-insecure group had low rates ofDV initially and showed a convex and then concavetrajectory with a higher mean at the last period; (c) theinsecure-secure group had low rates of prenatal DVand also showed a slight decrease over time, rising toinitial levels at the last time point; and (d) theinsecure-insecure group had high rates of prenatal DVwith a concave trajectory. Finally, there was a signifi-cant between-subjects main effect of gender, F(1, 143)¼5.36, p< .05, g2¼ .04, such that mothers of girls hadhigher levels of DV at all periods. There were no othersignificant effects.

Income. A repeated measures ANOVA was con-ducted. The covariates, DV during pregnancy and

FIGURE 2 Four attachment pattern groups and the trajectory of income from pregnancy through age 4.

DOMESTIC VIOLENCE ON STABILITY OF ATTACHMENT 405

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

maternal education, were not significantly related to theoutcomes. Multivariate tests revealed a significantTime�Pattern interaction, F(12, 150)¼ 1.87, p< .05,g2¼ .05, with a significant within-subjects interactionof Time�Pattern, F(12, 150)¼ 1.96, p< .05, g2¼ .05.There was a significant cubic effect, indicating changein curvature at different points between the groups,F(3, 150)¼ 5.57, p< .05, g2¼ .10. See Figure 2.

The secure-secure group had a fairly flat trajectory ofmoderate income. The secure-insecure group also beganwith a relatively higher income which had a slight con-cave trajectory. The insecure-secure group began witha relatively low income with a strong concave trajectory,ending slightly higher than they began at age 4. Finally,the insecure-insecure group had the lowest prenatalincome, with a fairly flat trajectory. There were no othersignificant main or interaction effects.

Maternal depressive symptoms. A repeated mea-sures ANOVA was conducted. Pregnancy DV, maternaleducation, and family income were used as covariates;they were not significantly related to the outcomes.The multivariate tests revealed a significant Time�Gender interaction, F(4, 150)¼ 2.63, p< .05, g2¼ .07.The within-subjects Time�Gender interaction was sig-nificant, F(4, 150)¼ 2.85, p< .05, g2¼ .02, with a quad-ratic effect, F(1, 141)¼ 6.67, p< .05, g2¼ .05. Mothersof girls had a U-shaped curve for depression; they weremost depressed during pregnancy and when their daugh-ters were age 4. Mothers of the boys had an invertedU-shape for depression; they were most depressed whentheir sons were age 2. There were no other significantmain or interaction effects.

DISCUSSION

This study investigated developmental trajectories ofrisk factors that predict attachment stability in a samplecharacterized by an overrepresentation of DV. Theimportance of examining more than two periods of theindependent variables is borne out by our findings; therelationship between stability=instability of attachmentand the predictors is complicated, and not linear.

In our sample, the stability rates of attachment fromage 1 to age 4 were between 44% and 57%, depending onwhether a two-group or four-group classification wasused. These are similar to prior studies of high-riskand heterogeneous samples (e.g., Barnett et al., 1999;NICHD ECCRN, 2001). Our study adds to the litera-ture by extending the assessment of attachment stabilityfrom 12 to 48 months in a heterogeneous sample; theonly other stability assessments greater than 36 monthswere in low-risk samples (e.g., Bar-Haim et al., 2000;Gloger-Tippelt, Gomille, Koenig, & Vetter, 2002).

Only about half of the children in our sample mani-fest stable attachment from infancy to preschool. Thesehigh rates of instability are consistent with other studiesfinding that the stability of the caregiving environment isa better predictor of later child outcomes than infantattachment (Belsky & Fearon, 2002; NICHD ECCRN,2006; Sroufe, Egeland, Carlson, & Collins, 2005;Tarabulsy et al., 2005). The child’s representationaladvances in cognitive and emotional development atage 4 (e.g., Thompson, 2000) may make these childrenmore sensitive to changes in their home environments.

Linear slopes in the trajectories of DV and householdincome explained little variance; it was the curvatures(cubic for income and quadratic for DV) of the trajec-tories that predicted the patterns of stability and insta-bility. For example, for the secure-insecure group, whostarted with low levels of DV during pregnancy, whichincreased, decreased, and then increased again, highlevels of DV at age 4 appeared to disrupt the high qual-ity of the mother–infant relationship that began duringthe 1st year of the child’s life. Perhaps the mother, inresponse to DV, became less involved with her child,or more distracted in her caregiving, as she coped withthe conflict and violence in her relationship with herpartner. Alternatively, the child’s developing emotionregulation, which is part of the attachment system,may have become dysregulated by the trauma of witnes-sing DV or its aftereffects. Our findings are consistentwith prior research indicating that contextual risk fac-tors during pregnancy affect children’s functioningthrough the 1st year of life (Benoit, Parker, & Zeanah,1997; Huth-Bocks, Levendosky, Bogat, & von Eye,2004; Lieberman, 1999). However, our findings also sug-gest that high levels of DV during pregnancy appearedto disrupt the mother–infant relationship throughoutthe first 4 years of the child’s life. This may be relatedto the mother’s working models of caregiving that aredisrupted by pregnancy DV (Huth-Bocks, Levendosky,Theran, et al., 2004). However, our analyses do notallow us to distinguish fully the effects of pregnancyDV from DV occurring during the postpartum period.

The cubic trajectory of income, controlling for theeffects of prenatal DV, was also related to the changeand stability of attachment. Socioeconomic disadvan-tage has been associated with negative parenting beha-viors, but it also may affect aspects of the homeenvironment that contribute to attachment security,including consistency of caregiving, stability of housing,and parental involvement. Our results for the effects ofincome during pregnancy, as well as stability of incomefrom pregnancy to age 4, are similar to recent findingsdemonstrating that lower levels of economic well-being,and the corollary of elevated perceptions of economicpressure, can indirectly affect parenting behavior andchild well-being through an adverse impact on parental

406 LEVENDOSKY ET AL.

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

psychological well-being (Mistry, Vandewater, Huston,& McLoyd, 2002). However, instability of attachmentwas not related in predictable ways to income. Thewomen in this study did not have particularly highhousehold incomes; thus, our results might not be gener-alizable to all women. Also, factors associated withincome (e.g., life stress) might be more predictive ofstability and instability of attachment. We did not havelife stress=life events data for all periods; thus, futureresearch is needed to explore this possibility.

Surprisingly, the trajectory of maternal depressivesymptoms was not associated with attachment stabilityor instability when prenatal DV was controlled. Priorresearch indicates that, among infants, maternaldepression is associated with the stability of insecureattachment (Edwards et al., 2004) as well as movementfrom security to insecurity from infancy to early adult-hood (Weinfield et al., 2004). Our research measuredmaternal depressive symptoms beginning in pregnancy,whereas the other studies did not. Thus, direct compar-isons between our research and other studies are notpossible. However, the lack of a relationship betweendepressive symptoms and attachment stability=insta-bility may reflect that most individuals in our sampledid not have clinical levels of depression. That is, themean scores for all attachment groups at each perioddid not reach the suggested clinical cutoff for the BDI.Alternatively, other research did not measure women’sexperiences of DV during pregnancy; these experiencesmay be better predictors of attachment than depressivesymptoms.

However, there was a Time�Gender finding suchthat the trajectory of depressive symptoms was reversedfor mothers of boys and girls. Prenatally, the mothers ofgirls had higher depression scores. However, this may bean artifact related to the fact that they were also morelikely to be in abusive relationships (see next). Postna-tally, the findings, which do not correlate with the DVfindings for gender, may instead reflect other difficultiesin the mothers’ lives, such as perceptions of differentialdifficulty of parenting boys and girls, which are agerelated or amount of social support given to mothersof boys and girls at different ages.

Mothers of girls had higher levels of DV at all periodscompared with mothers of boys. This is a surprisingfinding and may suggest that mothers of girls are atparticular risk for DV. One possible interpretation isthat mothers of boys may be more likely to leave theDV relationship, fearing that boys may model theaggression of the adult male in the family. This fearhas been previously reported in the literature but notspecifically linked to leaving the partner (e.g., Leven-dosky, Lynch & Graham-Bermann, 2000). There is noresearch examining whether gender of the child predictswhether a woman stays with or leaves her partner.

However, our results suggest a fruitful area for futureresearch.

Our findings should be interpreted cautiously, basedon several limitations in our methodology. Specifically,all contextual measures were obtained from a singlereporter—the mother. Another measurement limitationwas the lower reliability of the Strange Situation cod-ing at age 4. However, our interrater reliability is con-sistent with the MacArthur group guidelines (Solomon& George, 2008). In addition, we used continuousmeasures of all predictors. This allowed us to deter-mine that higher or lower levels of the predictors wereassociated with risk. However, we could not determinewhether a risk factor was only predictive when itreached a particular threshold or was under a parti-cular baseline. For example, maternal depressive symp-toms may be associated only with attachment stabilitywhen the mother fits diagnostic criteria for a mooddisorder. Finally, because of the size of our sampleand the need to analyze data from multiple timepoints, we could not test all of the predictors in onemodel. The covariance matrix of 1,770 entries wouldnot yield a model that would converge with 150 part-icipants. Future research with larger samples may beable to do so.

Implications for Research, Policy, and Practice

Despite these limitations, our results support thebroader theory that attachment classifications canchange over time in response to life experiences thatdirectly affect the caregiving environment. In particular,for children with unstable life circumstances, there maybe many opportunities to move toward or away fromsecurity. Further, our results demonstrate the impor-tance of multiple assessments of risk factors over timelest simple linear changes be presumed to be associatedwith changes in attachment stability. Rather, the pictureis more complex and suggests that certain periods of thechild’s early development may be more or less vulner-able to the influences of contextual risk factors, suchas DV or income. Although this study did not assessthe possibility of sensitive periods, this would be animportant direction for future research.

Especially important was our finding that DV duringpregnancy was a risk factor for insecure attachment atage 1. Social policies sensitive to these findings wouldensure adequate resources, including financial assistanceand mental health services for pregnant women as well asthe frequent monitoring of medical and psychologicalwell-being of women who are experiencing or at riskfor DV. Our findings emphasize the importance of train-ing physicians and other health care professionals work-ing with pregnant women to conduct adequateassessments for DV.

DOMESTIC VIOLENCE ON STABILITY OF ATTACHMENT 407

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

In addition, trajectories of stable low income and ris-ing DV were also associated with maintenance ofinsecurity or movement toward insecurity in the pre-school children. This suggests that mild or moderatelevels of DV should not be ignored. For example,pediatricians could administer a brief DV screen to themother during every well-child visit to evaluate this,and recommend appropriate treatment. A recentmeta-analysis showed that moderate-length inter-ventions targeting the mother–child relationship are suc-cessful in increasing sensitivity and changing attachmentquality (Bakermans-Kranenburg, van IJzendoorn, &Juffer, 2003) in generally low-risk populations. Oneexample of this type of intervention applied to familieswith DV is the child–parent psychotherapy for preschoolchildren developed by Lieberman and colleagues(Lieberman, van Horn, & Ghosh Ippen, 2005). For thosemothers who do not leave the abusive relationship, per-haps such interventions might counteract the influenceof increasing levels of DV.

REFERENCES

Aber, J. L., Jones, S., & Cohen, J. (2000). The impact of poverty on the

mental health and development of very young children. In C. H.

Zeanah (Ed.), Handbook of infant mental health 2nd ed. (pp. 113–

128). New York: Guilford Press.

Ainsworth, M. D. S., Blehar, M., Waters, E., & Wall, S. (1978). Pat-

terns of attachment. Hillsdale, NJ: Erlbaum.

Bakermans-Kranenburg, M. J., Van IJzendoorn, M. H., & Juffer, F.

(2003). Less is more: Meta-analyses of sensitivity and attachment

interventions in early childhood. Psychological Bulletin, 129,

195–215.

Bakersman-Kranenburg, M. J., Van IJzendoorn, M. H., &

Kroonenberg, P. M. (2004). Differences in attachment security

between African-American and white children: Ethnicity or socio-

economic status. Infant Behavior and Development, 27, 417–433.

Baldry, A. C. (2003). ‘‘Sticks and stones hurt my bones but his glance

and words hurt more’’: The impact of physiological abuse and

physical violence by current and former partners on battered

women in Italy. The International Journal of Forensic Mental

Health, 2, 47–57.

Bar-Haim, Y., Sutton, D. B., Fox, N. A., & Marvin, R. S. (2000). Stab-

ility and change of attachment at 14, 24, and 58 months of age:

Behavior, representation, and life events. Journal of Child Psy-

chology and Psychiatry, 41, 381–388.

Barnett, D., Ganiban, J., & Cicchetti, D. (1999). Maltreatment, nega-

tive expressivity, and the development of type D attachments from

12 to 24 months of age. In J. I. Vondra & D. Barnett (Eds.), Atypical

attachment in infancy and early childhood among children at develop-

mental risk. Monographs of the Society for Research in Child Devel-

opment, 64 Serial No. 258, (pp. 97–118). Hoboken, NJ: Wiley.

Beardslee, W. R., Bemporad, J., Keller, M. B., & Klerman, G. L.

(1983). Children of parents with major affective disorder: A review.

The American Journal of Psychiatry, 140, 825–832.

Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric

properties of the Beck Depression Inventory: Twenty-five years of

evaluation. Clinical Psychology Review, 8, 77–100.

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J.

(1961). An inventory for measuring depression. Archives of General

Psychiatry, 4, 561–571.

Belsky, J., & Fearon, R. M. P. (2002). Early attachment security, sub-

sequent maternal sensitivity, and later child development: Does con-

tinuity in development depend upon continuity of caregiving?

Attachment & Human Development, 4, 361–387.

Benoit, D., Parker, K. C. H., & Zeanah, C. H. (1997). Mothers’ repre-

sentations of their infants assessed prenatally: Stability and associ-

ation with infants’ attachment classifications. Journal of Child

Psychology and Psychiatry and Allied Disciplines, 38, 307–313.

Bogat, G. A., Levendosky, A. A., DeJonghe, E., Davidson, W. S., &

von Eye, A. (2004). Pathways of suffering: The temporal effects of

domestic violence on women’s mental health. Maltrattamento e

abuso all’infanzia, 6, 97–112.

Bogat, G. A., Levendosky, A. A., Theran, S., von Eye, A., & Davidson,

W. S., II. (2003). Predicting the psychosocial effects of interpersonal

partner violence (IPV): How much does a woman’s history of IPV

matter? Journal of Interpersonal Violence, 18, 1271–1291.

Bowlby, J. (1982). Attachment and loss: Volume 1. Attachment (2nd

ed.). New York: Basic Books. (Original work published 1969)

Bretherton, I., Ridgeway, D., & Cassidy, J. (1990). Assessing internal

working models of the attachment relationship: An attachment

story completion tasks for 3 year olds. In M. Greenberg, D.

Cicchetti & E. M. Cummings (Eds.), Attachment in the preschool

years (pp. 273–308). Chicago: The Chicago University Press.

Cain, D. S., & Combs-Orme, T. (2005). Family structure effects on

parenting stress and practices in the African American family. Jour-

nal of Sociology and Social Welfare, 32, 19–40.

Carlson, E. A., & Sroufe, L. A. (1995). Contributions of attachment

theory to developmental psychopathology. In D. Cicchetti & D. J.

Cohen (Eds.), Developmental psychopathology (pp. 581–617). New

York: Wiley.

Cassidy, J., Marvin, R. S., & the MacArthur Working Group.

(1992). Attachment organization in preschool children: Procedures

and coding manual. Unpublished manuscript, University of Virginia,

Charlottesville.

Cicchetti, D., & Barnett, D. (1991). Attachment organization in mal-

treated preschoolers. Development and Psychopathology, 3, 397–411.

Conger, R. D., Conger, K. J., Elder, G. H., & Lorenz, F. O. (1992). A

family process model of economic hardship and adjustment of early

adolescent boys. Child Development, 63, 526–541.

Conger, R. D., Conger, K. J., Elder, G. H., & Lorenz, F. O. (1993).

Family economic stress and adjustment of early adolescent girls.

Developmental Psychology, 29, 206–219.

Cummings, E. M., & Cicchetti, D. (1990). Toward a transactional

model of relations between attachment and depression. In M. T.

Greenberg, D. Cicchetti & E. M. Cummings (Eds.), Attachment in

the preschool years: Theory, research, and intervention (pp. 339–

372). Chicago: University of Chicago Press.

Dayton, C. J., Levendosky, A. A., Davidson, W. S., & Bogat, G. A.

(2010). The child as held in the mind of the mother: The influence

of prenatal maternal representations on parenting behaviors. Infant

Mental Health Journal, 31, 220–241.

Dutton, S., Pianta, R. C., & Marvin, R. S. (2001). Mothers’ represen-

tations of relationships with their children: Relations with parenting

behavior, mother characteristics, and child disability status. Social

Development, 10, 455–472.

Easterbrooks, M. A. (1989). Quality of attachment to mother and

to father: Effects of perinatal risk status. Child Development, 60,

825–830.

Edwards, E. P., Eiden, R. D., & Leonard, K. E. (2004). Impact of

fathers’ alcoholism and associated risk factors on parent–infant

attachment stability from 12 to 18 months. Infant Mental Health

Journal, 25, 556–579.

408 LEVENDOSKY ET AL.

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

Egeland, B., & Farber, E. A. (1984). Infant–mother attachment:

Factors related to its development and changes over time. Child

Development, 55, 753–771.

Egeland, B., & Sroufe, L. A. (1981). Attachment and early maltreat-

ment. Child Development, 52, 44–52.

Fearon, R. P., Bakermans-Kranenburg, M. J., Van Ijzendoorn, M. H.,

Lapsley, A., & Roisman, G. I. (in press). The significance of insecure

attachment and disorganization in the development of children’s

externalizing behavior: A meta-analytic study. Child Development.

Feld, S. L., & Straus, M. A. (1989). Escalation and desistance of wife

assault in marriage. Criminology, 1, 141–161.

Finkelhor, D., & Kendall-Tackett, K. A. (1997). A developmental

perspective on the childhood impact of crime, abuse, and violent

victimization. In D. Cicchetti & S. L. Toth (Eds.), Developmental

perspectives on trauma: Theory, research, and intervention (pp. 1–

32). Rochester, NY: University of Rochester Press.

Fleury, R. E., Sullivan, C. M., Bybee, D. I., & Davidson, W. S., II.

(1998). ‘‘Why don’t they just call the cops?’’—Reasons for differen-

tial police contact among women with abusive partners. Violence

and Victims, 13, 333–346.

Follingstad, D. R. (2009). The impact of psychological aggression on

women’s mental health and behavior: The status of the field.

Trauma, Violence, & Abuse, 10, 271–289.

Follingstad, D. R., Hause, E. S., Rutledge, L. L., & Polek, D. S.

(1992). Effects of battered women’s early responses on later abuse

patterns. Violence and Victims, 7, 109–128.

Fraley, R. C. (2002). Attachment stability from infancy to adulthood:

Meta-analysis and dynamic modeling of developmental mechan-

isms. Personality and Social Psychology Review, 6, 123–151.

Gelfand, D., & Teti, D. M. (1990). The effects of maternal depression

on children. Clinical Psychology Review, 10, 329–353.

Gloger-Tippelt, G., Gomille, B., Koenig, L., & Vetter, J. (2002).

Attachment representations in 6-year-olds: Related longitudinally

to the quality of attachment in infancy and mothers’ attachment

representations. Attachment & Human Development, 4, 318–339.

Huth-Bocks, A. C., Levendosky, A. A., Bogat, G. A., & von Eye, A.

(2004). The impact of maternal characteristics and contextual vari-

ables on infant-mother attachment. Child Development, 75, 480–496.

Huth-Bocks, A. C., Levendosky, A. A., Theran, S. A., & Bogat, G. A.

(2004). The impact of domestic violence on mothers’ prenatal repre-

sentations of their infants. Infant Mental Health Journal, 25, 79–98.

Joreskog, K. G., & Sorbom, D. (2006). LISREL 8.8 for Windows

[Computer software]. Lincolnwood, IL: Scientific Software

International.

Kessler, R. C., Molnar, B. E., Feurer, I. D., & Appelbaum, M. (2001).

Patterns and mental health predictors of domestic violence in the

United States: Results from the National Comorbidity Survey.

International Journal of Law and Psychiatry, 24, 487–508.

Kobak, R., Cassidy, J., & Ziv, Y. (2004). Attachment related trauma

and posttraumatic stress disorder: Implications for adult adap-

tation. In S. W. Rholes & J. A. Simpson (Eds.), Adult attachment:

Theory, research, and clinical implications (pp. 388–407). NY:

Guilford.

K. L. (2007, April). Attachment from infancy to preschool: Domestic

violence and other predictors of stability and instability. Paper

presented at the Society for Research in Child Development.

Boston, MA.

Levendosky, A. A., & Graham-Bermann, S. A. (2000). Behavioral

observations of parenting in battered women. Journal of Family

Psychology, 14, 80–94.

Levendosky, A. A., Leahy, K., Bogat, G. A., Davidson, W. S., & von

Eye, A. (2006). Domestic violence, maternal parenting, maternal

mental health, and infant externalizing behavior. Journal of Family

Psychology, 20, 544–552.

Levendosky, A. A., Lynch, S., & Graham-Bermann, S. A. (2000).

Maternal perceptions of the impact of violence on their parenting

strategies. Violence Against Women, 6, 247–271.

Lieberman, A. F. (1999). Negative maternal attributions: Effects on

toddlers’ sense of self. Psychoanalytic Inquiry, 19, 737–756.

Lieberman, A. F., van Horn, P., & Ghosh Ippen, C. (2005). Toward

evidence-based treatment: Child–parent psychotherapy with pre-

schoolers exposed to marital violence. Journal of the American Acad-

emy of Child and Adolescent Psychiatry, 44, 1241–1248.

Lovejoy, M. C., Graczyk, P. A., O’Hare, E., & Neuman, G. (2000).

Maternal depression and parenting behavior: A meta-analytic

review. Clinical Psychology Review, 20, 561–592.

Lyons-Ruth, K., Lyubchik, A., Wolfe, R., & Bronfman, E. (2002).

Parental depression and child attachment: Hostile and helpless pro-

files of parent and child behavior among families at risk. In S. H.

Goodman & I. H. Gotlib (Eds.), Children of depressed parents:

Mechanisms of risk and implications for treatment (pp. 89–120).

Washington, DC: American Psychological Association.

Main, M., & Solomon, J. (1986). Discovery of an insecure-

disorganized=disoriented attachment pattern. In T. B. Brazelton &

M. Yogman (Eds.), Affective development in infancy (pp. 95–124).

Norwood, NJ: Ablex.

Marshall, L. L. (1992). Development of the severity of violence against

women scales. Journal of Family Violence, 7, 103–121.

Martinez-Torteya, C., Bogat, G. A., von Eye, A., & Levendosky, A. A.

(2009). Resilience among children exposed to domestic violence: The

role of protective and vulnerability factors. Child Development, 80,

562–577.

Marvin, R. S., & Pianta, R. C. (1996). Mothers’ reactions to their

child’s diagnosis: Relations with security of attachment. Journal of

Clinical Child Psychology, 25, 436–445.

Masten, A. S., Best, K. M., & Garmezy, N. (1990). Resilience and

development: Contributions from the study of children who over-

come adversity. Development and Psychopathology, 2, 425–444.

Mistry, R. S., Vandewater, E. A., Huston, A. C., & McLoyd, V. C.

(2002). Economic well-being and children’s social adjustment: The

role of family process in an ethnically diverse low-income sample.

Child Development, 73, 935–951.

Morland, L.A., Leskin,G.A., Block,C.R., Campbell, J.C., &Friedman,

M. J. (2008). Intimate partner violence and miscarriage: Examination

of the role of physical and psychological abuse and posttraumatic stress

disorder. Journal of Interpersonal Violence, 23, 652–669.

Moss, E., Bureau, J. F., Cyr, C., St-Laurent, D., & Mongeau, C.

(2004). Correlates of attachment at age 3: Construct validity of the

preschool attachment classification system. Developmental Psy-

chology, 40, 323–334.

Moss, E., Cyr, C., Bureau, J. F., Tarabulsy, G. M., &

Dubois-Comtois, K. (2005). Stability of attachment during the pre-

school period. Developmental Psychology, 41, 773–783.

NICHD Early Child Care Research Network. (2001). Child-care and

family predictors of preschool attachment and stability from

infancy. Developmental Psychology, 37, 847–862.

NICHD Early Child Care Research Network. (2006). Infant–mother

attachment classification: Risk and protection in relation to changing

maternal caregiving quality. Developmental Psychology, 42, 38–58.

Owen, M. T., Easterbrooks, M. A., Chase-Lansdale, L., & Goldberg,

W. A. (1984). The relation between maternal employment status and

the stability of attachments to mother and father. Child Develop-

ment, 55, 1894–1901.

Rogosch, F. A., Mowbray, C. T., & Bogat, G. A. (1992). Determinants

of parenting attitudes in mothers with severe psychopathology.

Development and Psychopathology, 4, 469–487.

Sameroff, A. J., Seifer, R., & Zax, M. (1982). Early development of

children at risk for emotional disorder. Monographs of the Society

for Research in Child Development, 47, 1–82.

DOMESTIC VIOLENCE ON STABILITY OF ATTACHMENT 409

Levendosky, A. A., Bogat, G. A., Huth-Bocks, A. C., & Rosenblum,

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011

Schore, A. N. (2003). Affect dysregulation and the disorders of the self.

New York: Norton.

Shouldice, A., & Stevenson-Hinde, J. (1992). Coping with security dis-

tress: The Separation Anxiety Test and attachment classifications at

4.5 years. Journal of Child Psychology and Psychiatry, 33, 331–348.

Solomon, J., & George, C. (1999). The measurement of attachment

security in infancy and childhood. In J. Cassidy & P. Shaver

(Eds.), Handbook of attachment: theory, research, and clinical appli-

cations (pp. 287–316). New York: Guilford.

Solomon, J., & George, C. (2008). The measurement of attachment

security and related constructs in infancy and early childhood. In

J. Cassidy & P. Shaver (Eds.), Handbook of attachment, 2nd ed.

(pp. 383–416). New York: Guilford.

Sroufe, L. A., Egeland, B., Carlson, E. A., & Collins, W. A. (2005).

The development of the person: The Minnesota study of risk and adap-

tation from birth to adulthood. New York: Guilford.

Stevenson-Hinde, J., & Shouldice, A. (1995). Maternal interactions

and self-reports related to attachment classifications at 4.5 years.

Child Development, 66, 583–596.

Straus, M. (1979). Measuring intrafamily conflict and violence: The

conflict tactics scales. Journal of Marriage and the Family, 41,

75–88.

Tarabulsy, G. M., Bernier, A., Provost, M. A., Maranda, J., Larose, S.,

Moss, E., et al. (2005). Another look inside the gap: Ecological con-

tributions to the transmission of attachment in a sample of adolescent

mother–infant dyads. Developmental Psychology, 41, 212–224.

Thompson, R. A. (2000). The legacy of early attachments. Child Devel-

opment, 71, 145–152.

Thompson, R. A., Lamb, M. E., & Estes, D. (1982). Stability of

infant–mother attachment and its relationship to changing life

circumstances in an unselected middle-class sample. Child Develop-

ment, 53, 144–148.

Tolman, R. M., & Raphael, J. (2000). A review of research on welfare

and domestic violence. Journal of Social Issues, 56, 654–682.

U.S. Census Bureau. (2000). Census 2000 data for the state of

Michigan. Retrieved from http://www.census.gov/census2000/

states/mi.html

Van IJzendoorn, M. H., Schuengel, C., & Bakermans-Kranenburg,

M. J. (1999). Disorganized attachment in early childhood: Meta-

analysis of precursors, concomitants, and sequelae. Development

and Psychopathology, 11, 225–249.

Vondra, J. I., Hommerding, K. D., & Shaw, D. S. (1999). Stability and

change in infant attachment in a low income sample. In J. I. Vondra

& D. Barnett (Eds.), Atypical attachment in infancy and early child-

hood among children at developmental risk. Monographs of the

Society for Research in Child Development, 64 Serial No. 258 (pp.

119–144). New York: Wiley.

Weinfield, N. S., Whaley, G. J. L, & Egeland, B. (2004). Continuity,

discontinuity, and coherence in attachment from infancy to late ado-

lescence: Sequelae of organization and disorganization. Attachment

& Human Development, 6, 73–97.

Wofford, S., Mihalic, D. E., & Menard, S. (1994). Continuities in mari-

tal violence. Journal of Family Violence, 9, 195–225.

Zahn-Waxler, C., Iannotti, R. J., Cummings, E. M., & Denham, S.

(1990). Antecedents of problem behaviors in children of depressed

mothers. Development and Psychopathology, 2, 271–291.

410 LEVENDOSKY ET AL.

Downloaded By: [Eastern Michigan University] At: 14:04 10 June 2011