Assessing Children's Exposure to Intimate Partner Violence ...

Clin Child Fam Psychol Rev (2009) 12:157–173 DOI 10.1007/s10567-009-0048-1

Assessing Children’s Exposure to Intimate Partner Violence John F. Knutson Æ Erika Lawrence Æ Sarah M. Taber Æ Lew Bank Æ David S. DeGarmo

Published online: 13 May 2009 Springer Science+Business Media, LLC 2009

Abstract Child exposure to intimate partner violence (IPV) is widely acknowledged as a threat to the psychosocial and academic well-being of children. Unfortunately, as reflected in the literature, the specific link between such exposure and childhood outcomes is ambiguous. Based on a review of the literature, this article suggests that this state of affairs is due, in part, to the manner with which exposure to IPV is operationally defined. After reviewing the dominant strategies for operationally defining exposure to IPV and the problems associated with those strategies, this article reports original data contrasting three measures derived from maternal reports, three measures derived from child reports, and the limited concordance among those different indices of exposure to IPV. The implications of these findings for research on child outcomes and the clinical assessment of children who might have been exposed to IPV are discussed. Keywords Intimate partner violence Child exposure Child report Parent report

Intimate partner violence (IPV) is a highly prevalent and severe problem in the United States and other societies where it has been studied. Prevalence rates based on nationally representative surveys have found that 21–34% of women will be physically victimized by an intimate

J. F. Knutson (&) E. Lawrence S. M. Taber University of Iowa, E11 Seashore Hall, Iowa City, IA 52242, USA e-mail: [email protected] L. Bank D. S. DeGarmo Oregon Social Learning Center, Eugene, OR, USA

male during adulthood (e.g., pushing, slapping, or more severe acts of aggression; Straus and Gelles 1986, 1990; Straus et al. 1980). Moreover, approximately 1.8 million women are severely assaulted by male partners each year in the United States (e.g., kicking, punching, hitting with an object, using a weapon; Straus and Gelles 1990; Straus et al. 1980). Such prevalence rates are even higher during the early stages of relationships (i.e., among dating, cohabiting, engaged, or newlywed couples), with rates ranging from one-quarter to one-half of couples in the year prior to assessment (Elliot et al. 1986; Langer et al. 2008; Lawrence and Bradbury 2007; Leonard and Roberts 1998; O’Leary et al. 1989). The problem of IPV becomes even more alarming when one considers the number of children residing in these homes. Some estimates suggest that between 3 and 18 million children are exposed to interparental violence each year (Carlson 1984; Straus and Gelles 1990). Other researchers, estimating prevalence rates, have concluded that approximately one-third of children are exposed to adult-relational violence during childhood (Silvern et al. 1995; Spaccarelli et al. 1994). Although these estimated rates are high, they might actually underestimate prevalence rates of childhood exposure, as many of these studies were limited to twoparent households and/or economically advantaged homes. Such limitations, along with the tremendous variability among published prevalence rates, have led to a call for more refined research into the incidence and prevalence rates of children exposed to IPV and their consequences (e.g., Holden et al. 1998; Lewis-O’Connor et al. 2006). The purpose of this article is to review prior efforts to operationalize and measure children’s exposure to IPV, to present new approaches to assessing children’s exposure, and to offer recommendations for establishing a

123

158

standardized assessment protocol among researchers and clinicians assessing this important problem.

An Introduction to Research on Child Exposure to Intimate Partner Violence Although it seems obvious that exposure to IPV would have adverse effects on children’s psycho-social development, the links between exposure to IPV and specific adverse outcomes, and the mechanisms through which such links emerge, have yet to be fully explicated (cf. Kitzmann et al. 2003). Although exposure to IPV clearly has an adverse impact on the psychosocial status of children, it is impossible to draw definitive conclusions regarding the causal mechanisms between these variables. A key reason for this dearth of knowledge is that IPV can be embedded in other circumstances that are disadvantageous to child development (i.e., poverty, alcohol abuse, relational instability). Moreover, caretakers engaging in IPV may also demonstrate ineffective or dysfunctional parenting styles (e.g., harsh discipline, neglect), which can also impact children’s development; however, prior research has generally not disentangled these phenomena. In sum, to understand the role of IPV in child outcomes, and to develop effective interventions for children, it is critical to assess the impact of the exposure in a manner that captures its true complexity. One interesting aspect of the existing research on childhood outcomes of exposure to IPV is the fact that both internalizing and externalizing outcomes are reported. Although the former are typically considered to be reflections of the traumatic aspects of the exposure, the latter are considered indicative of the training in coercive behavior provided by a violent household. Importantly, in a recent meta-analysis (Kitzmann et al. 2003), effect sizes did not differ for internalizing versus externalizing symptoms, and both outcomes were prominent. This state of affairs could be created, in part, by the covariation between internalizing and externalizing behaviors when the Child Behavior Checklist (Achenbach 1991) is used as the index of behavior problems, but it is also possible that these findings reflect underlying process variables in the child’s home. For example, Henning et al. (1997) noted that adverse outcomes consistent with the traumatic influence of exposure to IPV were mediated by decreased parental warmth and caring. Thus, in order to understand the consequences of exposure to IPV, it may be important to assess the presence of negative and positive parenting styles in these homes. Contextual factors are incorporated into contemporary models of developmental psychopathology to explain how childhood experiences influence short- and/or long-term

123

Clin Child Fam Psychol Rev (2009) 12:157–173

child outcomes. For example, issues of multifinality (specific stressors associated with different childhood outcomes) and of equifinality (different environmental stressors associated with the same outcome) are prominent (cf. Cicchetti and Rogosch 1996, 2002). Thus, it should not be surprising to developmentalists that exposure to IPV could be associated with many different outcomes. If exposure to IPV were conceptualized as a contextual stressor along the lines discussed by Steinberg and Avenevoli (2000), then that exposure would be viewed as something that puts stress on a vulnerable child and as a circumstance that might lead to a mediation of deviant or maladaptive behavior. Within such a framework, understanding the impact of IPV on children’s social, emotional, and academic behavior requires a comprehensive assessment of the child, of the contextual aspects of exposure to IPV, and of the degree to which that context influences particular child behaviors. In all such researches, however, a pivotal problem has to do with how researchers actually assess a child’s direct and/or indirect exposure to IPV. Often, child exposure is inferred from secondary sources of information (e.g., administrative data); at other times, it is assumed from other information about the child’s family. In all of these cases, exposure is operationally assumed rather than actually assessed. We now turn to a review of prior sampling and methodological strategies used to assess children’s exposure to IPV.

Sampling Strategies for Assessing Children’s Exposure to Intimate Partner Violence Various sampling strategies have been used to identify children who have been exposed to IPV. The two most widely used strategies are to sample children whose mothers reside in domestic violence shelters and to sample children identified via administrative data [e.g., court documents, Department of Human Services (DHS) reports]. Both strategies are limited with respect to accurately establishing that a child has been exposed to IPV and each strategy occasions some unique potential problems for research. Sampling from Domestic Violence Shelters Because domestic violence (DV) shelters provide the most ready access to mothers who have experienced IPV, and to their children, shelter samples have provided a major vehicle for conducting research on the outcome of children exposed to IPV. When research participants are sampled from a shelter, the child’s exposure to IPV is assumed on the basis of shelter residence. Shelter samples are useful for


studying homes in which severe physical violence has occurred, and for clarifying victims’ and children’s perceptions of events. However, identifying children as having been exposed to IPV because they are (or their mother is) living in a shelter has several methodological limitations. For one thing, conclusions drawn from children (and their mothers) living in DV shelters likely do not represent the full range of IPV to which children are exposed. For example, women living in battered women’s shelters typically have experienced very severe physical violence and physical injuries, and they fear for their physical safety and the safety of their children. Thus, conclusions drawn from data collected from shelter samples may reflect the extreme end of the continuum of IPV and its consequences, but might not be applicable to the full spectrum of IPV (e.g., low levels of physical aggression that do not result in physical injuries or fear for safety). Research conducted with shelter samples also limits our understanding of how family structure and/or disruptions in family structure contribute—directly or indirectly—to child functioning. First, women living in DV shelters are, by definition, separated from their partners. However, couples experiencing IPV often do not separate and, even when they do separate, they often reconcile and/or continue to co-parent (Ackerman et al. 1999). More generally, it is important to study children from a broad spectrum of family structures: intact couples, couples whose relationships dissolve, and couples who establish new relationships during the child’s early years. Second, children living in DV shelters are residing under atypical conditions that likely do not generalize to the population of children exposed to IPV. These children have been exposed to the stress of being uprooted and living in a quasi-communal context, and are often in a context that can compromise effective parenting. In addition, changes in family structure (i.e., absence of the father) resulting from shelter residence may contribute to child functioning, directly or indirectly, in important ways (cf. Ackerman et al. 2002). Relatedly, mothers who have been in relationships characterized by IPV are often involved in sequential relationships. Relationship instability itself is known to adversely affect children’s academic and social functioning beyond that which can be accounted for by consideration of changes in family structure and SES (e.g., Ackerman et al. 1999; Ackerman et al. 2002). In sum, efforts to control for the influences of other disruptions that come with shelter housing—or to incorporate those factors into our explanatory frameworks—are important to understanding the impact of exposure to IPV on child outcomes. Third, using shelter samples necessarily precludes collecting data from fathers and cohabiting males involved in the IPV. Although some key variables might be assessed through indirect means (e.g., asking the mother about the

159

father’s alcohol and drug use, which carries with it its own methodological limitations), other factors such as assessment of personality traits (e.g., antisocial traits) and direct observations of father-child interactions cannot be obtained. Relatedly, the presence of sequential relationships also suggests that the child may potentially interact with, or be exposed to, multiple adult partners of his or her mother. Notably, the limited data collected from and about fathers in general, and the lack of data regarding paternal contributions to child development specifically (Phares et al. 2005), is a critical problem in many areas of developmental work, including research on parenting and child maltreatment (cf. Cabrera et al. 2000; Guterman and Lee 2005; Haskett et al. 1996; Knutson and Schartz 1997). Unfortunately, working with shelter samples largely precludes even the opportunity to collect those data and fully incorporate paternal factors into explanatory models. In sum, unpacking the multiple adult influences on a child’s development necessitates understanding how the child interacts with all adults playing a role in his or her development, a circumstance that is typically precluded with shelter samples. Sampling from Administrative Data In many areas of child maltreatment research, administrative data [e.g., Child Protective Service (CPS) records, law enforcement records] have been used to establish the abused or neglected status of the child (e.g., Sullivan and Knutson 1998, 2000a, b). The use of such data can have important advantages for researchers, and such an approach has provided important information with respect to both epidemiology and consequences of maltreatment. Because of these advantages, researchers often use administrative data to establish children’s exposure to IPV as well. Although administrative data can be useful (cf. Runyan and English 2006), using such data to define child maltreatment in general has occasioned a number of methodological concerns (cf. Knutson and Schartz 1997; National Academy of Science 1993; Sternberg et al. 2004; Widom 1988). The same above cited concerns can apply to the use of administrative data to define the occurrence of exposure to IPV. There are three primary sources for administrative data on children’s exposure to IPV. Because exposure to IPV is specified in the child protective statutes of most states, children who have been identified as residing in a household characterized by IPV often are included in central registries of that state. Typically, such children are identified as having been neglected by virtue of the fact that the custodial parent ‘‘permitted’’ the child to be exposed to such circumstances. Importantly, like obtaining samples from IPV shelters, state central registries do not necessarily include documentation of direct exposure to IPV.

123

160

A second source of administrative data comprises arrest records for domestic assaults or domestic disturbances. Because of the desire of law enforcement agencies to successfully prosecute criminal acts, the law enforcement records tend to include rather comprehensive narratives of the event that results in an arrest. Unfortunately, because children are not likely to be important prosecutorial witnesses in court, the specific involvement of children in the events are often not well documented. Thus, like other administrative data regarding exposure to IPV, law enforcement records often do not provide the sort of detailed information that actually establishes the circumstances of the child’s exposure to the event. A third source of administrative data is the office of the County Attorney, District Attorney, or other office that prosecutes domestic assault. Depending on the criminal codes of the state, the offices of the prosecutor may only have information regarding events that reach the standard of an indictable offense, or even felony assault. Again, to the extent that information about the child’s presence during a violent altercation is germane to the prosecution, the prosecutor’s office will review such records. Indeed, records in the prosecutor’s office are often silent with respect to whether children resided in a household at the time the violence occurred. Studies of the prevalence of physical aggression in young couples also point to the limitations of administrative data. Langer et al. (2008) found that 62% of newly married couples (M = 3 months of marriage) reported the use of physically aggressive tactics during dyadic conflict in the year prior to assessment. These rates are similar to those found in other samples of newlywed couples (Lawrence and Bradbury 2001; Leonard and Roberts 1998; O’Leary et al. 1989, 1994). Moreover, fully half of the physically aggressive couples reported the use of more severe tactics (punching, kicking, hitting with objects), but none of this aggression was reported to any agency. Thus, there are many couples that engage in physical aggression but do not come to the attention of treatment programs or governmental agencies and their acts are not represented in administrative data. In addition, in the Lawrence and Bradbury (2001) study, 56% of the physically aggressive marriages were still intact by the end of the 4-year study. Given the proportion of couples that have children in the early years of marriage, children reared in physically aggressive homes are likely to experience repeated episodes of aggression between their parents, especially during their early development. To the extent that a clinical sample and a community control sample are similar with respect to potential exposure, the greater is the probability that the differences between groups in child outcomes will be minimized. Importantly, these data also point to the potential problem when

123


community controls are compared with shelter-based samples, which may represent qualitatively different phenomena (e.g., Johnson and Ferraro 2000). Finally, samples derived from administrative data in CPS or law enforcement records tend to be largely economically disadvantaged, yet few studies of child exposure to IPV control for the influence of SES (Kitzmann et al. 2003). Clearly, children from unstable relationships and children from families residing in IPV shelters tend to be also confronted with problems of poverty and reduced financial resources. Household poverty and residing in neighborhoods of economic disadvantage contribute toward a considerable risk to children’s development (e.g., Jencks and Mayer 1990; Klebanov et al. 1998; Kohen et al. 2002; Tienda 1991). Thus, studies of the impact of IPV on children need to consider the contribution of economic disadvantage. Factors of social and economic disadvantage are also related to indices of community violence (Duncan and Brooks-Gunn 2000; Selner-O’Hagan et al. 1998), and can moderate the influence of exposure to community violence on children’s social, academic, and physical health outcomes (Ceballo and McLoyd 2002; Ewart and Suchday 2002; Kohen et al. 2002). In sum, children’s exposure to community violence, and other neighborhood influences associated with community dysfunction and poverty, could sharpen or blunt the exposure to IPV on children’s outcomes (cf. Coulton et al. 1999), although these factors are not readily available in administrative data. Another fundamental problem with using administrative data is that there are important jurisdictional differences as to the events that are entered into a record (see Flango 1988). Thus, when administrative data are the sole source of information regarding possible exposure to IPV, there may be idiosyncratic and nonreplicable differences among jurisdictions as to what events constitute IPV, and as to the (im)probability that the child’s direct exposure is documented. Notably, the Interagency Task Force on Research Definitions of Maltreatment (see Sternberg et al. 2004) concluded that the use of administrative data, such as that typically provided by CPS records to establish parenting and child maltreatment, is often insufficient. We extend that conclusion to the use of administrative data to identify children’s exposure to IPV.

Study #1: An Example of the Potential Problems with Administrative Data Differences among jurisdictions can also result in different circumstances of referral that leads to a family’s presence in IPV administrative data (e.g., Flango 1988). In developing this article, we examined information regarding maternal


reports of IPV in a secondary analysis of a high risk sample enrolled in a study of deficient parenting and child outcomes (see Knutson et al. 2005). The samples in that research were disadvantaged mothers with children 4– 8 years old who were recruited from either rural North Central Wisconsin (N = 100 mothers) or small urban and suburban communities in Southeastern Iowa (N = 102 mothers). The overall rates of IPV, as reported by the mothers during structured interviews, did not differ between the two sites. However, two interesting differences did emerge with respect to involvement of law enforcement agencies in response to IPV. First, the police were significantly less likely to be called in response to IPV in the more rural Wisconsin sample. Second, and perhaps more importantly, there were significant differences in how law enforcement agencies were contacted when an episode of IPV occurred and was reported to the police. Although half of the police calls were initiated by neighbors in the IA sample, none of the police calls in the more rural WI sample were initiated by a neighbor or a party outside the home. Thus, the different patterns of notifying law enforcement agencies may be related to the degree of urbanization of the family. With contact from law enforcement not coming from neighbors in rural areas, and considerable evidence that many women do not request intervention when a physically violent event occurs, using law enforcement records to identify families with potential exposure to IPV in rural samples would be much less likely to be accurate or representative than urban records. In addition, when police were contacted from within the home, often it was a child who called the police. This raises an interesting problem in the measurement of ‘‘exposure.’’ Although the actual IPV could have been comparable in two situations, it is possible that the exposure to violence experienced by a child who calls the police is qualitatively different from that of a child whose neighbor calls the police. To the degree that rural residency increases child participation in the act of notifying a law enforcement agency of IPV in his or her home, rural residency could alter the circumstances of exposure to children. Moreover, the circumstances of child exposure reflected in those records might be different. At the least, these data suggest that administrative data from rural areas might systematically underestimate households with IPV. Notably, these families were all socially and economically disadvantaged. Although the families could have been identified as physically abusive or neglectful, they had not been identified as having experienced IPV in the recruitment process. Moreover, 50% of the mothers were in intimate relationships (married or cohabiting) at the time of assessment. Two sources of information were available regarding the relationships of these women with their partners. A structured interview provided information about

161

conflict and IPV, and the Dyadic Adjustment Scale (DAS; Spanier 1976) provided broader information about that current relationship. On the basis of the DAS, approximately 45% of the couples fell below the range typically used to classify a couple as happily married. More importantly, 28% described themselves in the seriously distressed range. Thus, approximately half of the children in this sample who were largely identified and recruited on the basis of economic disadvantage were being reared in a household with parents who characterized their relationships as distressed. When the interview data were examined, the quality of the relationship between the adults would seem to be even more problematic for the children. Within the sample from Iowa, physical encounters, although less frequent than verbal conflict, were described by close to 20% of the coupled sample. Although less severe forms of physical aggression (i.e., pushing, grabbing, pulling, restraining) were most common, 16% reported throwing objects, and approximately 12% reported hitting, with half of those subjects reporting hitting with objects. Approximately 14% of the coupled Iowa sample indicated that the police had been called at least once because of their conflicts. Within the WI sample, physical encounters were described by close to 35% of the mothers who were currently in a coupled relationship, with 79% of them reporting that the throwing of objects was involved in some physical encounters. Of the WI sample that acknowledged physical aggression, 25% reported that injuries occurred. Approximately 11% of the sample in a marital or cohabiting relationship indicated that the police had been called at least once because of conflicts. In short, this sample, unselected with respect to marital discord and IPV provide data that suggests considerable IPV to which a child could be exposed would not be found in administrative data from both jurisdictions. The conflict and intimate partner data described above were based on the prevailing relationships of the mothers at the time they were recruited into the study. The structured interview also provided information regarding any prior relationships of the mother during the child’s lifetime. Approximately 32% of the women reported experiencing physical aggression in a prior relationship during the child’s lifetime. When a test of association between prior relationships and current relationships was conducted, it was clear that the mothers who reported experiencing IPV in a prior relationship were no more or less likely to be in a currently aggressive relationship than those who had not had such prior experiences. Thus, studies of child exposure to IPV in a current household tells us nothing about potential exposure in the mother’s home when she was in a different romantic relationship. Thus, in studies of the consequences of exposure to IPV, it is critical to determine child exposure in the mother’s current relationship and in

123

162

any prior relationships. To the extent that administrative data do not capture the full range of a mother’s relationship experiences, those data will not provide access to the full range of IPV to which a child could have been exposed. These secondary data analyses from the IA and WI communities indicate that when current and prior relationships are considered, it is not unreasonable to estimate that disadvantaged samples (unselected with respect to IPV), could reflect a population in which 35–40% of the children were being reared in a household characterized by physical aggression between the mother and her partner. Importantly, according to maternal report, considerable physical conflict between the mother and her partner occurred that was not reported to law enforcement agencies. In some research contexts, children from such samples could be used as a control group that was presumably not exposed to IPV. Thus, these interview data emphasize the need to directly assess IPV exposure rather than using administrative data to establish a child’s exposure.

Measurement Strategies for Assessing Children’s Exposure to Intimate Partner Violence Establishing prevalence rates of aggression, and the severity and frequency of that aggression, has largely been accomplished using the Conflict Tactics Scales (CTS, Straus 1979; CTS-2, Straus et al. 1996), the most widely used self-report measure of physical aggression in intimate relationships. The CTS also often serve as a proxy measure for child exposure to IPV. That is, if a parent endorses specific items on the CTS, thereby reporting the occurrence of specific violent acts within an adult relationship, the children are classified as having been exposed to IPV. However, there are no items on the CTS regarding the presence of children during aggressive encounters. Thus, CTS data provide more an index of potential exposure to IPV rather than to measuring actual exposure. There are only two sources of information regarding child exposure to IPV—the adults engaged in the IPV and the child who could have been exposed. Historically, researchers have used administrative data, inferential evidence, or maternal reports to identify children who have been exposed to IPV. In contrast, children have not been used as sources of information. Given that research has been conducted on child reports regarding parenting (e.g., Gaylord et al. 2003; Locke and Prinz 2003), and given the call to interview children in clinical and forensic contexts regarding IPV (e.g., Faller 2003), it is somewhat surprising that child reports have not been used in research on the consequences of exposure to IPV. Although obtaining reports from children can occasion some challenges, when questions of actual direct exposure to IPV emerge, the

123


possibly exposed children are one potential source of data, and those children might provide critically important information.

Study #2: Method and Source Contributions to Assessing Children’s Exposure to Intimate Partner Violence Based on the material described in the forgoing sections of this article, we elected to determine whether different methods for assessing exposure to IPV using maternal report and child report would yield comparable patterns of results. The data we present in this section are part of a larger longitudinal project investigating the role of deficient parenting on children’s behavioral problems. Participants and Sampling Strategies Participants were recruited from suburban communities, small towns, and rural agricultural areas in eastern Iowa, and from small towns and rural communities in northcentral Wisconsin. The Wisconsin county from which the sample was recruited is supported by logging, light industry and tourism, though many residents live in prototypical isolated rural poverty. Sentinel agencies (Iowa Department of Human Services and Oneida County Department of Social Services) provided lists of eligible families (recipients of economic services and at least one child between the ages of 4 and 8). In addition, the data of women who had been involved in injurious (indictable) domestic assaults were obtained from the offices of the County Attorney (CA) and the District Attoney (DA) in IA and WI, respectively. Families were deemed ineligible if the child was known to be the victim of sexual abuse; child victims of physical abuse or neglect were still eligible. In addition, families were not eligible if the child was currently in an out-of-home placement or the family was receiving intensive in-home services. The final sample (N = 252) was diverse with respect to degree of urbanization and with respect to racial and ethnic composition (68% Caucasian, 24% African American; 2.4% Native American, 4% Hispanic, and 1.6% Asian/Pacific Islander). Mean age of the enrolled child was 5.7 years. In brief, recruitment of eligible subjects involved sending an initial letter to the parent of the eligible child subject outlining the research project, informing the parents that they would be compensated for each session (including informed consent) in which they participated, that their child would be compensated for each session, and how to contact the laboratory, if they were interested. If a parent did not respond within 3–4 weeks of being sent the original letter, a follow-up letter reiterating much of the initial letter


is sent together with an indication that someone will telephone soon to discuss the project. Within a 2-week period after the second letter, a person telephoned the home to answer questions and schedule an initial appointment conducted in the potential participant’s home. Because of the mobility of these families, loss of services (i.e., disconnected phones), and a tendency not to check mail reliably, we estimate that we could actually reach (by mail or phone) approximately 50% of the potential participants who have been sent letters. Of those actually contacted, based on over 7 years of experience with this recruitment strategy and the recruitment of over 850 families in several projects, we estimate that approximately 55% of the families contacted schedule an initial appointment to learn about the project and participate in the informed consent process. Less than 1% of those who participated in the informed consent process declined to participate. Because the records from the DA and CA do not necessarily indicate whether the woman involved in the domestic assault actually had children, it is impossible to determine a response rate for those potential participants. All of the procedures were conducted under the aegis of The University of Iowa Institutional Review Board (IRB-2) and with a Certificate of Confidentiality issued by NICHD. A structured interview with the mother was administered in the family home. All other relevant data were collected during laboratory sessions typically scheduled within 30 days of the initial session in the home. Laboratory sessions included multiple measures of children’s exposure to IPV as well as a series of measures and procedures beyond the scope of this article. Each measure of exposure to IPV was obtained during a different laboratory appointment separated by at least 1 week. Research assistants who administered the measures were blind to the recruitment source from which the participants were obtained, to any other data on IPV in the family, and to any child maltreatment in the family. Maternal Measures of Children’s Exposure to IPV Structured Interview The structured interview addresses a wide range of topics regarding family function, parenting, and child behavior. In particular, it includes questions regarding the mother’s experiences with IPV in her current relationship as well as in any prior relationships that occurred during the child’s lifetime. If the mother described any encounters that involved actual physical force between the mother and a partner, she was also asked about the child’s location during the dispute. In addition, she was asked whether a law enforcement agency had been contacted and to identify

163

the person, if she knew, who had contacted the law enforcement agency. Conflict Tactics Scales—2nd Edition (CTS-2, Straus et al. 1996) The CTS-2 is a 78-item self-report scale of aggression that has occurred in the context of conflicts with a partner during the previous 6 months, detaiIing physical, psychological, and sexual tactics. It has moderate internal consistency and yields significant interpartner agreement (Straus et al. 1996). In this article, only the Physical Assault scale was used. Examples of items : throwing something at partner, pushing, grabbing or shoving partner, and punching or hitting partner. The CTS-2 was included as an index of potential child exposure to IPV. For the purposes of the present analyses, we simply computed prevalence rates of male-to-female and female-to-male physical aggression. The Context of Intimate Partner Violence Interview (CIPVI, Lawrence et al. 2008) The CIPVI is a structured, individual interview that provides a detailed summary of specific couple conflicts characterized by IPV. With regard to their current relationship, mothers report on the most recent argument in which IPV occurred and on the ‘‘worst argument ever’’ in their current relationship (i.e., the argument with the most severe physical violence). A broad spectrum of conflict behaviors are assessed (e.g., verbal aggression, withdrawal, psychological aggression, physical threats, moderate and severe physical aggression) as well as the proximal context and consequences of those events (e.g., alcohol consumption, injuries, help-seeking behaviors). Frequency and typicality of aggressive acts are also determined. The CIPVI is then administered a second time to gather data on past relationships in which IPV occurred. Mothers are screened with regard to IPV across all of their relationships during the child’s lifetime. Decision-making criteria are then used to prioritize which past relationships to target with the CIPVI (i.e., the past relationship marked by the most severe physical violence, the most significant past relationship). Across all of the arguments assessed via the CIPVI, the degree of actual child exposure to those arguments is investigated. Child exposure is measured on an ordinal scale and includes various degrees of direct exposure (e.g., child was in the room and saw the IPV, child was upstairs but awake and heard everything) and indirect exposure (e.g., child heard the mother telling the police about it, child inferred that IPV occurred because the father was taken away by the police).

123

164

Child Measures of Children’s Exposure to IPV Because the mean age of the children enrolled in the project was under 6 years of age at the time of enrollment (M = 5.7 years of age), most of the children did not have the literacy to respond effectively to questionnaires. Thus, all of these measures were administered in interview format. Interviewing young children about events such as exposure to IPV requires a combination of structure and flexibility to match the developmental status of the child and to minimize interviewer influence or intervening misinformational distortion. In order to address these challenges, child interviews were structured by guidelines recommended for forensic interviewing (e.g., Lamb et al. 1999; Poole and Lamb 1998), research on autobiographical memories of victim and witness children (e.g., Orbach et al. 2001), and developmental research assessing children’s recall of specific events (e.g., Brainerd and Reyna 2002; Geddie et al. 2001; Holliday 2003; Marche and Howe 1995; Pipe et al. 1999), including traumatic events (e.g., Peterson 1999; Peterson and Rideout 1998).


selected response, so the child can change his or her answer. A research assistant sits next to the child during the administration to facilitate the handling of the equipment and to encourage responding but does not provide any feedback to the child regarding his or her responses. The CACI includes 10 items pertaining to disputatious encounters between the child’s parents, four of which specifically relate to physical encounters. These four items result in an IPV exposure score ranging from 0 to 8. ‘‘My Exposure to Violence’’ interview (METV, Selner-O’Hagan et al. 1998)

The structured child interview is designed to assess a number of areas of social adjustment, peer relations, and direct and indirect exposure to IPV. With respect to IPV, according to the recommendations of Faller (2003), children were asked to describe conflicts and arguments between their parents (or between their mother and her current partner). Descriptions were sought to distinguish between directly observing the event versus knowing about it from the reports of others.

The ‘‘My Exposure to Violence’’ interview (METV, Selner-O’Hagan et al. 1998) was developed to assess the full range of violent experiences of children (i.e., both family and community violence). The instrument includes 20 descriptors ranging from relatively minor but potentially injurious acts to the use of weapons and homicide. In addition, there are four items pertaining to being told about people they knew being shot, killed, committing suicide, or sudden death or injury. For each item, children can report the event occurring at home, at school, in the current neighborhood, or in a former neighborhood. For each positive response, the child is given an opportunity to describe the event. In this article, we only examined the children’s responses to the nine questions that pertained to violent events within the child’s home, which were not directed at the child, yielding potential scores of 0–9. In this sample, scores ranged from 0 to 3. Because the METV does not require children to specify the perpetrators and recipients of the events reported, it serves as an index of potential exposure to IPV.

Computer-Assisted Child Interview—2nd Edition (CACI-2, Bank et al. 2000)

Mothers’ Data on Their Children’s Exposure to Intimate Partner Violence

The Computer-Assisted Child Interview—2nd Edition (CACI-2, Bank et al. 2000) is a structured interview for children of ages 4–9 based on Ci3 software (Sawtooth Technologies, 1999, Northbrook, IL, USA), and an effective strategy for obtaining information from children (e.g., Bank et al. 2000; J. Snyder, 2002, personal communication). The CACI includes items regarding verbal and physical aggression between the child’s parents (e.g., ‘‘My mom yells at my dad,’’ ‘‘My mom and dad get into fights,’’ ‘‘When Mom gets mad she throws things at my dad’’). Questions are presented via audio statements and clip art. Response choices are represented with visual images of three jars of gumballs to represent the three response choices (an empty jar = ‘‘not at all;’’ a half-full jar = ‘‘sometimes;’’ a full jar = ‘‘a lot’’). Finally, audio feedback (‘‘You answered…’’) is given regarding the

At the time of enrollment, 31% of the households were single-parent families. Of the mothers who reported residing with a partner, 60% were with the child’s biological father, 16% were married to the child’s stepfather, 19% were residing with a boyfriend unrelated to the child, and 5% were in other partnered relationships (e.g., a same-sex romantic relationship; a heterosexual, ‘‘purely sexual’’ relationship). Thus, risk for exposure to IPV represented a current risk for 69% of the sample and risk for exposure from one or more of their mother’s prior relationships included about 68% of the sample, with risk being based on the mother being in any romantic partnership during the child’s lifetime. On the basis of the structured interview with the child’s mother, 31.3% indicated that altercations with her current partner involved some physical acts. When the mothers described the location of the child during those physical

Structured Child Interview

123


altercations, 76% indicated the child was in the same room for at least part of the encounter or in an adjacent room and aware of the acts. Mothers who denied that the child had been exposed to the altercation reported that the child was either away from the home with another party, outside during the encounter, or that the child had been sleeping and not awakened by the fight. Of those mothers who had been in prior romantic relationships during the child’s lifetime, 37% indicated that the relationship had involved physical altercations with their partner. Approximately 50% of those reporting a physical encounter with the partner in a prior relationship indicated that the subject child had been exposed to that conflict. When the reports of IPV in the current and past relationships were combined, 52% of the total sample of children could have been exposed to IPV during their lifetime because intimate violence had occurred. According to the interview reports of the mothers, 34% of the total sample of children was actually exposed to such acts during one or more of the mother’s relationships. Information regarding IPV exposure from the CTS is confined to the current partnered relationship of the mother at the time of enrollment. In order to classify the relationship as potentially exposing a child to IPV, the mother’s report of any of the moderate and severe physical acts by the mother, her partner, or both was used. Based on the CTS classification, 28.4% of the mothers indicated that their current relationship had involved at least one moderate level physical altercation. Thus, estimating potential exposure from the CTS-2 yielded a prevalence rate that was approximately 90% of that estimated from the structured interview. In order to determine potential exposure on the basis of the CIPVI interview, descriptions of physically aggressive encounters in the current relationship and past relationships were aggregated across both the most recent encounter and the worst encounter. For the current relationship, 30.4% of the mothers described at least one physically aggressive encounter with their partner. Of those who had been in a past relationship during their child’s lifetime, 64.8% reported at least one physically aggressive encounter with that partner. Combining past relationships and the mother’s current relationship, 39.9% of the children had the potential of direct exposure to IPV because their mothers reported one or more episodes of such physical aggression during the child’s lifetime, a prevalence rate approximately 80% of the rate based on the structured interview. Based on the mother’s report of the location of the child during the physically aggressive encounter, children could be distinguished as not having any direct exposure (outside, at a remote location), having direct exposure (e.g., present, in an adjacent room), and unknown exposure (e.g., sleeping, outside in the yard). Aggregating exposure from past and present relationships, mothers reported direct exposure for 35% of the children

165

residing in homes in which there had been IPV, no exposure for 34%, and ambiguity in exposure for 8%. In addition, approximately 24% of the reported incidents were described as having occurred when the child was an infant, and therefore, not likely to have been aware of the event. Children’s Data on Their Own Exposure to Intimate Partner Violence In the structured interview, the child was asked to describe up to three arguments or fights that their mother had with her partner. The child’s descriptions were classified as reporting physical or nonphysical acts. The physical acts could be attributed to the mother, her male partner, or both. Of the total sample, 9.1% provided a description of an encounter that could be classified as having been a physically aggressive act by one or both of the adults. Thus, the structured interview yielded a rate of exposure to IPV of less than one-third of that based on maternal reports. The interview administration of METV provided up to nine opportunities for children to endorse specific items related to acts of physical coercion within their homes, which could involve their mothers as victims or perpetrators. Scores ranged from 0 to 3, with 74.6% of the subjects reporting no physical violence on the METV. Of the children who reported some physically aggressive acts in their homes, two-thirds reported a single event. If the data are merely dichotomized, approximately 25% reported exposure to some violence in their homes. The rate is more than double the rate based on the child interview directed at the topic of interparental conflict, suggesting the children experienced considerable violence in the home, but that violence did not necessarily involve their parent(s). There were four items on the CACI that pertained to specific acts of physical aggression in encounters between the child’s parents. Children could indicate that the events never happened, happened sometimes, or happened a lot. Sixty percent of the children indicated that none of the acts occurred, with 40% of the children endorsing at least one physically aggressive act by a parent or parent equivalent. Approximately 19% of those reporting some physical aggression indicated only one act occurring ‘‘some.’’ The remaining children endorsed either multiple events occurring some of the time or they endorsed a single event occurring ‘‘a lot.’’ Thus, the CACI interview resulted in a prevalence estimate from the child that was appreciably higher than either METV or the structured interview. Concordance Between Respondents and Among Measures In order to determine whether there is any concordance among these measures of exposure to IPV, the first analysis

123

166

entailed pairwise comparisons across measures as to whether the child was classified as exposed to IPV by the mother. Because the CTS-2 only provides information about potential exposure, the classification of participants with the CTS-2 was cross-tabulated with the potential score based on the structured maternal interview. Although the estimate from the CTS-2 was lower, the association between the two measures was statistically significant (v2 = 13.7; p \ .0001). When the CTS-2 measure of potential exposure was cross-tabulated with the estimate of potential IPV from the CIPVI based only on the mother’s current relationship, the result was v2 = 58.7 (p \ .0001). When the potential score based on the structured interview was crosstabulated with the CIPVI estimate of potential IPV exposure based on past and current relationships of the mother, the result was v2 = 23.44 (p \ .001). In short, although the reports of IPV were not identical across measures, there was a high degree of concordance among measures of potential exposure to IPV based on three different measures of maternal report. When the two indices of whether or not the child had been exposed to the IPV based on maternal report (Structured Interview and CIPVI) were crosstabulated, the test of association was statistically significant (v2 = 10.8; p = .001). Although there was a reliable association between the two measures, 36% of the subjects were classified as inconsistent. Three crosstabular comparisons of the classification of the children on the basis of the three indices of child report yielded only a single statistically significant test of association (v2 = 7.18; p = .007) between the reported exposure during the structured interview and the CACI. The tests of association between the interview and METV and between METV and the CACI did not approach statistical significance. This could reflect the fact that the METV does not specify interparental conflict, whereas the interview and the CACI explicitly focused on interparental events. Although the association between the CACI and the interview was statistically significant, with the relatively low base rate of the children reporting exposure during the interview, the high rate of inconsistent responses between measures (34%) underscores the degree to which the measures are not congruent enough to be treated as comparable indices. The markedly different rates of ‘‘exposure’’ to IPV reported by mothers and their child suggest there would be limited congruence between child reports and parent reports. Indeed, none of the tests of association between the mother reports and the child reports even approached statistical significance. The analyses of these different indices of actual and potential exposure to IPV strongly suggest that indices of young child exposure to IPV are influenced by both the source of information (mother versus child) and the method used to obtain the information. Although indices from

123


mothers were congruent, the absolute magnitude of differences between measures and the inconsistencies in classification based on just the maternal measures suggest that different methodologies are sufficiently disparate such that they could yield different findings in tests of the consequences of exposure to IPV. Focusing just on the maternal measures, the findings do suggest that, regardless of measure used, the children in this disadvantaged sample were likely to have been exposed to a large amount of IPV. The vast majority of that intimate partner aggression did not, according to the mother’s report, result in any report to a law enforcement agency. Conclusions About Child Versus Mother Reports The inconsistencies among child reports and between the child reports and the maternal reports are sobering, especially given that child reports are thought to be useful in verifying parent reports (see Sternberg et al. 1997), and given that child reports are recommended for use in forensic assessments as well as in research (see Faller 2003). Notably, the conduct of the interviews was structured to minimize untoward influences in the interview process (e.g., using different interviewers for each interview, keeping interviewers blind to other available information regarding IPV), and the interviews were designed to be optimal for obtaining reliable and accurate information from young children (e.g.,. using structured interviews with concrete questions). Of course, one of the more replicated findings in clinical research is that different informants often disagree with one another (De Los Reyes and Kazdin 2005; Edelbrock et al. 1986; Weissman et al. 1980). When pairs of informants fail to agree, it has been argued that the two informants have different experiences with respect to the domain of interest or experience the target in different circumstances (cf. Achenbach et al. 1987; Renk 2005). In the current context, however, both the mothers and children have presumably had the same experiences. Furthermore, much of the research to date examining informant discrepancies has primarily focused on discrepancies in reports of child behavior (e.g., externalizing behaviors in children). For example, De Los Reyes and Kazdin (2005) emphasize the implications of discrepancies in reports of child psychopathology for assessment and diagnostic outcomes in children. In contrast, with few exceptions (e.g., Gaylord et al. 2003; Schwarz et al. 1985; Sessa et al. 2001; Tein et al. 1994), there is a relative absence of research specifically examining parent-child discrepancies in reports of parental behavior (e.g., parenting practices, IPV). Moreover, although it is important to pursue implications of report discrepancies in determining associated child outcomes (as in the De Los Reyes and Kazdin 2005 work) in the context of this article,


establishing the veridicality of children’s reports of parental behavior is specifically germane to the assessment of exposure to IPV. In general, when there is a mismatch between child reports and parent reports of parental behaviors, the inconsistency is often attributed to dysfunctional family circumstances or adjustment problems of the child (e.g., Gaylord et al. 2003; Sessa et al. 2001). However, if one is trying to determine the impact of exposure to IPV to a poor child outcome, one cannot invoke a poor child outcome as a basis for discrepant reports of exposure to that IPV. In short, if we hope to understand the contribution of exposure to IPV on child outcomes, we cannot merely assume that poor correspondence between maternal reports and child reports is due to child psychopathology or child behavior problems. Although discrepancies between respondents are typically greater when the respondents are child reporters (e.g., Phares et al. 1989; Renk and Phares 2004), that does not suggest that accurate information cannot be obtained from child informants. There is a sizeable literature that indicates that children can provide reliable and accurate information (e.g., Chambers and Johnston 2002; Lamb et al. 2003; Quas et al. 2007; Schwartz-Kenney and Goodman 1999), and that young children can provide useful and accurate information about traumatic events (e.g., Berliner et al. 2003). Because some research has suggested that age could be a factor in accurate reporting of information, due to influences of memory, language development, and susceptibility to misattributional errors (e.g., Ceci and Bruck 1993; Ceci and Huffman 1997), we attempted a number of analyses to relate age of the children to log-transformed child reports of exposure to IPV and log-transformed concordance rates: age did not significantly influence those indices. Admittedly, the sample was relatively young and, therefore, had a somewhat truncated age range. However, studies have demonstrated accurate reporting of traumatic and nontraumatic events by children as young as 3 years of age. Therefore, the discrepancies between child and mother reports in this study cannot simply be attributed to the age of the child respondents. Within contemporary research, where complex models of family function and child outcomes are tested, it is desirable to use multiple informants and multiple measures for construct development. In this context, although the three indices provided by the mothers provided some consistency, the lack of correspondence between the children and the mothers suggests that multi-informant construct development with respect to exposure to IPV could be a difficult proposition. However, researchers have begun to examine potential methods for combining discrepant informant ratings (e.g., De Los Reyes and Kazdin 2004; Offord et al. 1996). One alternative strategy would be to classify children in a three category system which takes into

167

account two respondents indicating exposure, either respondent indicating exposure, and neither respondent indicating exposure. In the existing research assessing accuracy of child reports of traumatic events, or in analog tests of accuracy of child reports, there is a known standard against which the child report can be compared. In the current context, there is no known standard. In other words, we attempted to assess the concordance between two fallible sources using measures that, although designed to minimize interviewer influences and enhance accuracy of reporting, clearly have measurement error. Lacking an objective ‘‘gold standard,’’ researchers and clinicians attempting to assess exposure to IPV are forced to use such fallible measures. Recognizing that both mother reports and child reports are fallible, and that there is no basis for concluding that either report is superior to the other, we believe that efforts to obtain reports from both children and parents should continue, and that systematic efforts to integrate those different sources and methodologies are needed.

Moving to a Contextual, Continuous, Dynamic Conceptualization of Children’s Exposure to Intimate Partner Violence In addition to issues of sampling and measurement, other factors are relevant to truly understanding the nature of children’s exposure to IPV. In this section, we discuss two such factors: whether to measure children’s exposure to IPV as a continuous and dynamic phenomena, and the potential utility of considering the proximal, dyadic, and familial contexts in which IPV occurs. Measurement of Intimate Partner Violence Although many childhood experiences are modeled continuously (e.g., Greenwald et al. 1997; Knutson et al. 2004), there is a tendency to treat children’s exposure to IPV as a static and binary category, rather than as a continuous, dynamic process. This approach likely limits our understanding of both quantitative and qualitative differences among circumstances of IPV exposure. For example, children are routinely classified as having been exposed to IPV without clarification of the frequency or severity of the episodes (Kitzmann et al. 2003). It is rare that an episode of IPV would occur in isolation in a long-term relationship. Moreover, the more severe a given episode of IPV, the more likely it will occur again. Women who report severe violence at the first assessment are markedly more likely to experience aggression at later time points (Feld and Straus 1990; Lawrence and Bradbury 2007; Quigley and Leonard 1996; Woffordt et al. 1994).

123

168

In addition, the longitudinal courses of IPV differ substantially across relationships. Cross-sectional data suggest a tendency for interpartner aggression to decrease markedly as couples get older and the duration of the relationship increases (cf. Pan et al. 1994; Straus and Gelles 1986). However, when specific couples are tracked across time, a significant proportion of them do remain aggressive, and 25–30% of individuals who report aggression in their dating and newly married relationships report an escalation in the severity of the aggression over time (O’Leary 1988; O’Leary et al. 1994; Roscoe and Benaske 1985; Stith et al. 2000; Woffordt et al. 1994). The essential model of escalation has been important in laboratory models of aggression (e.g., Knutson et al. 1980; Knutson and Viken 1984), in models of children’s aggression (e.g., Patterson 1982; Patterson and Reid 1970), and in efforts to understand physically abusive parenting (e.g., Greenwald et al. 1997; Knutson and Bower 1994). In addition, the more severe initial aggression is associated with a greater likelihood of a continuation of the aggression. Women who report severe violence at the first assessment are markedly more likely to experience aggression at later time points (Feld and Straus 1990; Quigley and Leonard 1996; Woffordt et al. 1994). Still other studies have found no systematic change over time (e.g., Langer et al. 2008; Lawrence and Bradbury 2007). Such data underscore the fact that children in households characterized by couple aggression could experience markedly different patterns of exposure across time. Therefore, research on children’s exposure to that violence needs to reflect the child’s exposure to a dynamic process rather than to the mere presence or absence of a discrete event. Indeed, the nature of the IPV to which the child is exposed over time may be more influential on child development than the impact of a given episode. Proximal, Relationship, and Familial Contexts in Which IPV Occurs Episodes of IPV occur within several contexts, including proximal contexts (e.g., a sequence of behaviors), and dyadic and familial contexts. With regard to the proximal context of IPV, factors—such as the temporal antecedents and provoking events triggering the aggression (e.g., childfocused discussions; Papp et al. 2002), how aggressive conflicts are resolved, and the attributions and/or emotional reactions of couples before, during, and after the violent episodes—are all likely to be relevant. For example, such contextual factors could play a role in subsequent parenting, influence a child’s immediate response to the violence, and/ or mediate a child’s longitudinal emotional and social adjustment. In addition, the actual manner in which children are exposed to IPV, and how parents interact with their children during the proximal and distal periods following

123


such conflicts, might also be important in determining the impact of the exposure to IPV on child outcomes. In terms of the relationship context, one factor to be considered is the relationship satisfaction of the couple. Findings from longitudinal studies of physical aggression in young marriages, and the findings from our WI sample described above, suggest the need to assess relationship satisfaction as well as IPV, and to consider the relations that these two variables may have with each other. In young couples, physical aggression typically precedes relationship decline (e.g., Lawrence and Bradbury 2007; Langer et al. 2008). Relatedly, physical aggression often occurs among satisfied couples (e.g., Lawrence and Bradbury 2001, 2007). Thus, studies of childhood exposure to aggression need to detail the full range of both relational adjustment and physical aggression between the child’s caretakers, rather than make assumptions about exposure to IPV as a function of other indices of relationship adjustment or satisfaction. Another factor to be considered is that many children are reared in households in which there are frequent changes in the parent’s romantic partner. With an increased risk for couple aggression in the early stages of a relationship, children reared in homes with relationship instability may be at elevated risk for exposure to IPV, as well as exposure to multiple perpetrators of that violence. Research needs to consider the persistent effects of IPV from earlier parenting contexts as well as conflict in the current relationship that could influence the development of the child. For example, Kim et al. (1999) reported that step-father negativity was related to the development of externalizing behaviors and deviant peer associations. For maltreatment and abuse, families with mother and father figures present, fathers are alleged perpetrators in an estimated 71% of physical abuse cases and 69% emotional maltreatment cases (Crooks et al. 2006). A small number studies also show that the presence of an unrelated male adult increases risk for child maltreatment and injuries (U.S. DHHS 2006). It must be noted, however, that although such findings suggest the step-family and father figure contexts could exacerbate the influence of prior exposure to IPV, Bradley et al. (1994) noted that the presence of a male in the household can also be a buffering influence when assessing the impact of poverty on the outcome of high risk children. Perhaps, a positive male in the household could buffer a child from a previously violent relationship. In short, when studying children exposed to IPV, there is a need to determine the circumstances of IPV in the current relationship as well as in past relationships to which the child was exposed. Another factor that needs to be considered when evaluating the impact of exposure to IPV on child outcomes is the broader familial context to which the child is exposed.


There is a large and growing literature examining the emotional and behavioral impact of familial and parental conflict (e.g., Davies et al. 2002), as well as evidence that parenting, and the child’s perceptions of parental support, can buffer the child from the adverse impact of interparental conflict (e.g., Davies and Windle 2001). Other research has suggested that parenting mediates the impact of interparent conflict on child psychosocial adjustment (e.g., Jones et al. 2003). More specifically, IPV covaries with harsh punitive discipline and neglect. Although some researchers have attempted to assess the impact of both IPV and child maltreatment (e.g., Maughan and Cicchetti 2002), when the actual exposure is ambiguous, it becomes impossible for researchers to disentangle the impact of the exposure to IPV from the impact of being reared in a disadvantageous circumstance characterized by deficient parenting. To the extent that evidence suggests that parenting practices can mediate the relation between disadvantage and poor child outcomes (e.g., Linver et al. 2002; Grant et al. 2003; Knutson et al. 2005), it is crucial for researchers to be able to distinguish the specific impact of IPV from the impact of deficient parenting and economic disadvantage. Direct Versus Indirect Exposure As noted throughout the material reviewed,mostly exposure to IPV by the child has been assumed on the basis of parental membership in a specific sampling group. One of the fundamental problems with measuring exposure to IPV is that there is rarely an opportunity for a neutral and objective observer to make a determination of whether a child was directly or indirectly exposed to IPV. As noted throughout the material reviewed, often exposure to IPV by the child has been assumed on the basis of parental membership in a specific sampling group. For example, when administrative data are used to establish exposure to IPV, or the CTS is used to establish exposure, the degree to which the exposure is direct or indirect is ambiguous. Similarly, when shelter samples are studied, there is an assumption that the child was directly exposed to the violence that led the family members to seek shelter. Such children typically know of the events that occasioned moving to a shelter, but that knowledge could be also be indirect (e.g., the child was not home at the time but the next day, the child overheard his or her mother telling someone else about the violence that occurred). Indeed, Sternberg et al. (1997, 2004) have raised questions about whether the assumption of direct exposure is justified for samples derived from shelters without additional information. There are examples in the literature, however, in which the degree to which the child is directly or indirectly

169

exposed to the IPV has been a focus (see El-Sheikh and Harger 2001; Jouriles et al. 1998, 2001a, b; Kolko et al. 1996). However, even when exposure is clear (i.e., the child’s presence in the home is documented), the degree of child involvement in the event is not necessarily detailed. There might be differential impact in circumstances where the child is actively involved (e.g., calls 911, opens door for police, runs for help) versus those circumstances where the child is in the room and visually witnesses the violence but is not actively involved, versus those situations when the child listens to the encounter from an adjacent room. When we examined the raw interview data from the WI and IA samples in the secondary analyses described above (Study #1), the mothers’ responses were often highly ambiguous with respect to whether exposure was direct or indirect. When mothers described a episode of IPV, they were asked about the location of the child when that encounter occurred. Approximately 11% indicated that the child was elsewhere (outside or at a remote location) during the encounter. For the others, the mothers indicated that the child was in the home during the altercation, but only 33% of the mothers reported the child was directly present during the encounter. With large numbers of children in the home but not necessarily present during the fight, the maternal reports of ‘‘exposure’’ are ambiguous. Moreover, the interviews also indicated that mothers often told the child about the event, providing indirect exposure to the child. With regard to differential exposure, three important points are worth noting. First, as noted in the data described above, there can be considerable disagreement between parent and child reports regarding that exposure (cf. Sternberg et al. 1997). This disagreement may be particularly relevant when parents assert that a child was not actually exposed to the conflict or was not aware of the event. Second, there is some evidence that more severe aggression occurs during arguments that might directly involve the child. For example, based on conflict diaries maintained by parents, Papp et al. (2002) reported that the more severe and destructive parental conflict occurred in the presence of the child. Such work strongly suggests that researchers attempting to understand the impact of IPV and related events need to obtain more comprehensive information based on multiple informants to truly understand direct and indirect child exposure and the types of conflict in which that exposure is embedded. Third, even if the child is not present during the aversive encounter, interparental conflict has been found to impact parenting. Data from a recent laboratory analog study indicated that marital conflict that occurs immediately before a parent-child interaction can affect those parent-child interactions even when the marital conflict does not occur in the presence of the child (Kitzmann 2000). Such work also underscores the

123

170

call from Kitzmann et al. (2003) to complete detailed assessments of interparental conflict, the physical aggression that occurs during interparental conflict, and the degree to which the child is exposed to that conflict.

Conclusion The material reviewed above and the original data in this report underscore the complex process that is involved in efforts to measure child exposure to IPV and the putative outcomes. In view of the fact that influential models advanced to account for the impact of IPV on children (e.g., Gottman and Katz 1989; Grych and Fincham 1990; Grych et al. 2000) are based on the assumption that children are directly exposed to the IPV, it is clear that efforts to establish the veridicality of child exposure are needed. Tests of theoretical propositions and understanding the effects of IPV on children, requires effective measures of that exposure. Although developing such measures will be a complex process, the prevalence of IPV and the high rate of child exposure suggest that new efforts directed at the measurement of the exposure are sorely needed. Acknowledgments This research was supported in part, by Grant RO1 MH 61731 funded by NIMH and ACYF (John F. Knutson, PI), and, in part, by Grant R01 HD-46789 funded by NICHD (John F. Knutson, PI), in part, by Centers for Disease Control and Prevention Grants R49 CCR721682 and R49 CE721682 and, in part, by Grant P30 DA 023920, Division of Epidemiology, Services and Prevention Branch, NIDA. The facilitation of the research accomplished by Barry Bennett, Cheryl Whitney, Marc Baty, Mark Schmidt, Trisha Barto, Mindy Norwood, and Wayne McCracken (Iowa Department of Human Services), Paul Spencer (Oneida County Department of Social Services), and the assistance of Ashley Anderson, Lisa Bauer, Robin Barry, Allyson Bone, Beth Boyer, Kristy DePalma, Aubra Hoffman, Esther Hoffman, Kathryn Holman, Kyla Kinnick, Gina Koeppl, Robert Latzman, Katie Meyer, Bethany Murphy, Amanda Murray, Laureen Ann Rapier, Eunyoe Ro, Nicole Shay, and Sarah Taber is gratefully acknowledged.

References Achenbach, T. M. (1991). Child behavior checklist. Burlington, VT: University of Vermont, Department of Psychology. Achenbach, T. M., McConaughy, S. H., & Howell, C. T. (1987). Child/adolescent behavioral and emotional problems: Implications of cross-informant correlations for situational specificity. Psychological Bulletin, 101, 213–232. Ackerman, B. P., Brown, E. D., D’Eramo, K. S., & Izard, C. (2002). Maternal relationship instability and the school behavior of children from disadvantaged families. Developmental Psychology, 38, 694–704. Ackerman, B. P., Kogos, J., Youngstrom, E., Schoff, K., & Izard, C. (1999). Family instability and the problem behaviors of children from economically disadvantaged families. Developmental Psychology, 35, 258–268.

123

Clin Child Fam Psychol Rev (2009) 12:157–173 Bank, L., Mustoe, J., Johnson-Shelton, D., Ford, D., Smith, E., Prescott, A., & Burraston, B. (2000, November). The OSLC computer-assisted child interview for children aged 4–9. Paper presented at the annual meeting of the association for the advancement of behavior therapy, New Orleans, LA. Berliner, L., Hyman, I., Thomas, A., & Fitgerald, M. (2003). Children’s memory for trauma and positive experiences. Journal of Traumatic Stress, 16, 229–236. Bradley, R. H., Whiteside, L., Mundfrom, D. J., Casey, P. H., Kelleher, K. J., & Pope, S. K. (1994). Early indications of resilience and their relation to experiences in the home environments of low birthweight, premature children living in poverty. Child Development, 65, 346–360. Brainerd, C. J., & Reyna, V. F. (2002). Recollection rejection: How children edit their false memories. Developmental Psychology, 38, 156–172. Cabrera, H. J., Tamis-LeMonda, C. S., Bradley, R. H., Hofferth, S., & Lamb, M. E. (2000). Fatherhood in the twenty-first century. Child Development, 71, 127–136. Carlson, B. E. (1984). Children’s observations of interparental violence. In A. R. Roberts (Ed.), Battered women and their families (pp. 147–167). New York, NY: Springer. Ceballo, R., & McLoyd, V. C. (2002). Social support and parenting in poor, dangerous neighborhoods. Child Development, 73, 1310– 1321. Ceci, S. J., & Bruck, M. (1993). Suggestibility of the child witness: A historical review and synthesis. Psychological Bulletin, 113, 403–439. Ceci, S. J., & Huffman, M. C. (1997). How suggestible are preschool children? Cognitive and social factors. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 948–958. Chambers, C. T., & Johnston, C. (2002). Developmental differences in children’s use of rating scales. Journal of Pediatric Psychology, 27, 27–36. Cicchetti, D., & Rogosch, F. A. (1996). Equifinality and multifinality in developmental psychopathology. Development and Psychopathology, 8, 597–600. Cicchetti, D., & Rogosch, F. A. (2002). A developmental psychopathology perspective on adolescence. Journal of Consulting and Clinical Psychology, 70, 6–20. Coulton, C. J., Korbin, J. E., & Su, M. (1999). Neighborhoods and child maltreatment: A multi-level study. Child Abuse and Neglect, 23, 1019–1040. Crooks, C. V., Scott, K. L., Francis, K. J., Kelly, T., & Reid, M. (2006). Eliciting change in maltreating fathers: Goals, processes, and desired outcomes. Cognitive and Behavioral Practice, 13, 71–81. Davies, P. T., Harold, G. T., Goeke-Morey, M. C., & Cummings, E. M. (2002). Child emotional security and interparental conflict. Monographs of the Society for Research in Child Development, 67, 3 Serial No. 270. Davies, P. T., & Windle, M. (2001). Interparental discord and adolescent adjustment trajectories: The potentiating and protective role of intrapersonal attributes. Child Development, 72, 1163–1178. De Los Reyes, A., & Kazdin, A. E. (2004). Measuring informant discrepancies in clinical child research. Psychological Bulletin, 16(3), 330–334. De Los Reyes, A., & Kazdin, A. E. (2005). Informant discrepancies in the assessment of childhood psychopathology: A critical review, theoretical framework, and recommendations for further study. Psychological Bulletin, 131, 483–509. Duncan, G. J., & Brooks-Gunn, J. (2000). Family poverty, welfare reform, and child development. Child Development, 71, 188–196. Edelbrock, C., Costello, A. J., Dulcan, M. K., Conover, N. C., & Kala, R. (1986). Parent–child agreement on child psychiatric

Clin Child Fam Psychol Rev (2009) 12:157–173 symptoms assessed via structured interview. Journal of Child Psychology and Psychiatry, 27, 181–190. Elliot, D. S., Huizinga, D., & Morse, B. J. (1986). Self-reported violent offending: A descriptive analysis of juvenile violent offenders and their offending careers. Journal of Interpersonal Violence, 4, 472–514. El-Sheikh, M., & Harger, J. (2001). Appraisals of marital conflict and children’s adjustment, health, and physiological reactivity. Developmental Psychology, 37, 875–885. Ewart, C. K., & Suchday, S. (2002). Discovering how urban poverty and violence affect health: Development and validation of a neighborhood stress index. Health Psychology, 21, 254–262. Faller, K. C. (2003). Research and practice in child interviewing: Implications for children exposed to domestic violence. Journal of Interpersonal Violence, 18, 377–389. Feld, S. L., & Straus, M. A. (1990). Escalation and desistance from wife assault in marriage. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families: Risk factors and adaptation to violence in 8, 145 families (pp. 489–505). New Brunswick, NJ: Transaction Publishers. Flango, C. R. (1988). State courts’ jurisdiction and terminology for child abuse and neglect cases. Williamsburg, VA: National Center for State Courts. Gaylord, N. K., Kitzmann, K. M., & Coleman, J. K. (2003). Parents’ and children’s perceptions of parental behavior: Associations with children’s psychosocial adjustment in the classroom. Parenting: Science and Practice, 3, 23–47. Geddie, L. F., Beer, J., Bartosik, S., & Wuensch, K. L. (2001). The relationship between interview characteristics and accuracy of recall in young children: Do individual differences matter? Child Maltreatment, 6, 59–68. Gottman, J. M., & Katz, L. F. (1989). The effects of marital discord on young children’s peer interaction and health. Developmental Psychology, 25, 373–381. Grant, K. E., Compas, B. E., Stuhlmacher, A. E. T., McMahon, S. D., & Halpert, J. A. (2003). Stressors and child and adolescent psychopathology: Moving from markers to mechanisms of risk. Psychological Bulletin, 129, 447–466. Greenwald, R. L., Bank, L., Reid, J. B., & Knutson, J. F. (1997). A discipline-mediated model of excessively punitive parenting. Aggressive Behavior, 23, 259–280. Grych, J. H., & Fincham, F. D. (1990). Marital conflict and children’s adjustment: A cognitive-contextual framework. Psychological Bulletin, 108, 267–290. Grych, J. H., Fincham, F. D., Jouriles, E. N., & McDonald, R. (2000). Interparental conflict and child adjustment: Testing the mediational role of appraisals in the cognitive-contextual framework. Child Development, 71, 1648–1661. Guterman, N. B., & Lee, Y. (2005). The role of fathers in risk for physical child abuse and neglect: Possible pathways and unanswered questions. Child Maltreatment, 10(2), 136–149. Haskett, M. E., Marziano, B., & Dover, E. R. (1996). Absence of males in maltreatment research: A survey of recent literature. Child Abuse and Neglect, 20, 1175–1182. Henning, K., Leitenberg, H., Coffey, P., Bennett, T., & Jankowski, M. K. (1997). Long-term psychological adjustment to witnessing physical conflict during childhood. Child Abuse and Neglect, 21, 501–515. Holden, G. W., Geffner, R., & Jouriles, E. N. (Eds.). (1998). Children exposed to marital violence: Theory, research, and applied issues. Washington, DC: American Psychological Association. Holliday, R. E. (2003). Reducing misinformation effects in children with cognitive interviews: Dissociating recollection and familiarity. Child Development, 74, 728–751. Jencks, C., & Mayer, S. (1990). The social consequences of growing up in a poor neighborhood. In L. E. Lynn Jr. & G. H. McGeary

171 (Eds.), Inner-city poverty in the United States (pp. 111–186). Washington, DC: National Academy Press. Johnson, M. P., & Ferraro, K. J. (2000). Research on domestic violence in the 1990s: Making distinctions. Journal of Marriage and the Family, 62, 948–963. Jones, D. J., Shaffer, A., Forehand, R., Brody, G., & Armistead, L. P. (2003). Coparent conflict in single mother-headed African American families: Do parenting skills serve as a mediator or moderator of child psychosocial adjustment. Behavior Therapy, 34, 259–272. Jouriles, E. N., McDonald, R., Norwood, W. D., Ware, H. S., Spiller, L. C., & Swank, P. R. (1998). Knives, guns, and interparent violence: Relations with child behavior problems. Journal of Family Psychology, 12, 178–194. Jouriles, E. N., McDonald, R., Spiller, L., Norwood, W. D., Swank, P. R., Stephens, N., et al. (2001a). Reducing conduct problems among children of battered women. Journal of Consulting and Clinical Psychology, 69, 774–785. Jouriles, E. N., Norwood, W. D., McDonald, R., & Peters, B. (2001b). Intimate partner violence and child adjustment. In J. Grych & F. Fincham (Eds.), Child Development and interparental conflict (pp. 315–336). Cambridge: Cambridge University Press. Kim, J. E., Hetherington, E. M., & Reiss, D. (1999). Associations among family relationships, antisocial peers, and adolescents’ externalizing behaviors: Gender and family type differences. Child Development, 70, 1209–1230. Kitzmann, K. M. (2000). Effects of marital conflict on subsequent triadic family interactions and parenting. Developmental Psychology, 36, 3–13. Kitzmann, K. M., Gaylord, N. K., Holt, A. R., & Kenny, E. D. (2003). Child witnesses to intimate partner violence: A meta-analytic review. Journal of Consulting and Clinical Psychology, 71, 339– 352. Klebanov, P. K., Brooks-Gunn, J., McCarton, C., & McCormick, M. C. (1998). The contribution of neighborhood and family income to developmental test scores over the first three years of life. Child Development, 69, 1420–1436. Knutson, J. F., & Bower, M. E. (1994). Physically abusive parenting as an escalated aggressive response. In M. Potegal & J. F. Knutson (Eds.), The escalation of aggression in dyads and groups: Biological and social processes (pp. 195–225). Mahwah, NJ: Lawrence Erlbaum. Knutson, J. F., DeGarmo, D., Koeppl, G., & Reid, J. B. (2005). Care neglect, supervisory neglect and harsh parenting in the development of children’s aggression: A replication and extension. Child Maltreatment, 10, 92–107. Knutson, J. F., DeGarmo, D. S., & Reid, J. B. (2004). Social disadvantage and neglectful parenting as precursors to the development of antisocial and aggressive child behavior: Testing a theoretical model. Aggressive Behavior, 30, 187–205. Knutson, J. F., Fordyce, D. J., & Anderson, D. J. (1980). Escalation of irritable aggression: Control by consequences and antecedents. Aggressive Behavior, 6, 347–359. Knutson, J. F., & Schartz, H. A. (1997). Physical abuse and neglect of children. In T. A. Widiger, A. J. Frances, H. A. Pincus, R. Ross, M. B. First, & W. Davis (Eds.), DSM-IV sourcebook (Vol. 3, pp. 713– 804). Washington, DC: American Psychiatric Association Press. Knutson, J. F., & Viken, R. J. (1984). Animal analogues of human aggression: Studies of social experience and escalation. In D. C. Blanchard, K. J. Flannelly, & R. J. Blanchard (Eds.), Biological perspectives on aggression. (pp. 75–94). New York: Alan R. Liss, Inc. Kohen, D. E., Brooks-Gunn, J., Leventhal, T., & Hertzman, C. (2002). Neighborhood income and physical and social disorder in Canada: Associations with young children’s competencies. Child Development, 73, 1844–1860.

123

172 Kolko, D. J., Kazdin, A. E., & Day, B. T. (1996). Children’s perspectives in the assessment of family violence: Psychometric characteristics and comparisons to parent report. Child Maltreatment, 1, 156–167. Lamb, M. E., Sternberg, K. J., Orbach, Y., Epslin, P. W., Steward, H., & Mitchell, S. (2003). Age differences in young children’s responses to open-ended invitations in the course of forensic interviews. Journal of Consulting and Clinical Psychology, 71, 926–934. Lamb, M. E., Sternberg, K. J., Orbach, Y., Heshkowitz, I., & Esplin, P. W. (1999). Forensic interviews of children. In A. Memon & R. Bull (Eds.), The psychology of interviewing: A handbook (pp. 253–277). New York, NY: Wiley. Langer, A., Lawrence, E., & Barry, R. A. (2008). Using a vulnerability-stress-adaptation framework to predict physical aggression trajectories in newlywed marriage. Journal of Consulting and Clinical Psychology, 76, 756–768. Lawrence, E., & Bradbury, T. N. (2001). Physical aggression as a predictor of marital dysfunction: A longitudinal analysis. Journal of Family Psychology, 15, 135–154. Lawrence, E., & Bradbury, T. N. (2007). Trajectories of change in physical aggression and marital satisfaction. Journal of Family Psychology, 21, 236–247. Lawrence, E., Peek-Asa, C., Langer, A., Brock, R., Yoon, J., Georgia, E., & Siebrecht, J. (2008). Reliability and validity of the Context of Intimate Partner Violence Interview across four samples. Unpublished manuscript. Leonard, K. E., & Roberts, L. J. (1998). Marital aggression, quality, and stability in the first year of marriage: Findings form the Buffalo Newlywed Study. In T. N. Bradbury (Ed.), The developmental course of marital dysfunction (pp. 44–73). New York, NY: Cambridge University Press. Lewis-O’Connor, A., Sharps, P. W., Humphreys, J., Gary, F. A., & Campbell, J. (2006). Children exposed to intimate partner violence. In M. M. Feerick & G. B. Silverman (Eds.), Children exposed to violence. Baltimore, MD: Paul H. Brookes Publishing Company. Linver, M. R., Brooks-Gunn, J., & Kohen, D. E. (2002). Family processes as pathways from income to young children’s development. Developmental Psychology, 38, 719–734. Locke, L. M., & Prinz, R. J. (2003). Measurement of parental discipline and nurturance. Clinical Psychology Review, 22, 895–929. Marche, T. A., & Howe, M. L. (1995). Preschoolers report misinformation despite accurate memory. Developmental Psychology, 31, 554–567. Maughan, A., & Cicchetti, D. (2002). Impact of child maltreatment and interadult violence on children’s emotion regulation abilities and socioemotional adjustment. Child Development, 73, 1525–1542. National Academy of Science. (1993). Understanding child abuse and neglect. Washington, DC: National Academy Press. O’Leary, K. D. (1988). Physical aggression between spouses. In V. B. Van Hasselt, R. L. Morrison, A. S. Bellack, & M. Hersen (Eds.), Handbook of family violence (pp. 31–56). New York: Plenum Press. O’Leary, K. D., Barling, J., Arias, I., Rosenbaum, A., Malone, J., & Tyree, A. (1989). Prevalence and stability of physical aggression between spouses: A longitudinal analysis. Journal of Consulting and Clinical Psychology, 57, 263–268. O’Leary, K. D., Malone, J. E., & Tyree, A. (1994). Physical aggression in early marriage: Prerelationship and relationship effects. Journal of Consulting and Clinical Psychology, 62, 594–602. Offord, D. R., Boyle, M. H., Racine, Y., Szatmari, P., Fleming, J. E., Sanford, M., et al. (1996). Integrating assessment data from multiple informants. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 1078–1085.

123

Clin Child Fam Psychol Rev (2009) 12:157–173 Orbach, Y., Lamb, M. E., Sternberg, K. J., Williams, J. M. G., & Dawud-Noursi, S. (2001). The effect of being a victim or witness of family violence on the retrieval of autobiographical memories. Child Abuse and Neglect, 25, 1427–1437. Pan, H. S., Neidig, P. H., & O’Leary, K. D. (1994). Predicting mild and severe husband-to-wife physical aggression. Journal of Consulting and Clinical Psychology, 62, 975–981. Papp, L. M., Cummings, E. M., & Goeke-Morey, M. C. (2002). Marital conflicts in the home when children are present versus absent. Developmental Psychology, 38, 774–783. Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia, Press. Patterson, G. R., & Reid, J. B. (1970). Reciprocity and coercion: Two facets of social systems. In C. Neuringer & J. L. Michael (Eds.), Behavior modification in clinical psychology (pp. 133–177). New York: Appleton-Century-Crofts. Peterson, C. (1999). Children’s memory for medical emergencies: 2 years later. Developmental Psychology, 35, 1493–1506. Peterson, C., & Rideout, R. (1998). Memory for medical emergencies experienced by 1- and 2- year-olds. Developmental Psychology, 34, 1059–1072. Phares, V., Compas, B. E., & Howell, D. C. (1989). Perspectives on child behavior problems: Comparisons of children’s self-reports with parent and teacher reports. Psychological Assessment, 1, 68–71. Phares, V., Fields, S., Kamboukos, D., & Lopez, E. (2005, October). Still looking for Pappa. American Psychologist, 73 5–736. Pipe, M.-E., Gee, S., Wilson, J. C., & Egerton, J. M. (1999). Children’s recall 1 or 2 years after an event. Developmental Psychology, 35, 781–789. Poole, D. A., & Lamb, M. E. (1998). Investigative interviewing of children. Washington, DC: American Psychological Association. Quas, J. A., Malloy, L. C., Melinder, A., Goodman, G. S., D’Mello, M., & Schaaf, J. (2007). Developmental differences in the effects of repeated interviews and interviewer bias on young children’s event memory & false reports. Developmental Psychology, 43, 823–837. Quigley, B. M., & Leonard, K. E. (1996). Desistance of husband aggression in the early years of marriage. Violence and Victims, 11, 355–370. Renk, K. (2005). Cross informant ratings of the behavior of children and adolescents: The ‘‘Gold Standard’’. Journal of Child and Family Studies, 14, 457–468. Renk, K., & Phares, V. (2004). Cross-informant ratings of social competence in children and adolescents. Clinical Psychology Review, 24, 239–254. Roscoe, B., & Benaske, N. (1985). Courtship violence experienced by abused wives: Similarities in patterns of abuse. Family Relations, 34, 419–424. Runyan, D. K., & English, D. J. (2006). Measuring child abuse and neglect using child protective services records. In M. Feerick, J. F. Knutson, P. Trickett, & S. Flanzer (Eds.), Child abuse and neglect: Definitions, classifications, & a framework for research (pp. 255–292). Baltimore, MD: Brooks Publishing Co. Schwartz-Kenney, B. M., & Goodman, G. S. (1999). Children’s memory of a naturalistic event following misinformation. Applied Developmental Science, 3, 34–46. Schwarz, J. Barton-Henry, M. & Pruzinsky, T. (1985). Assessing child-rearing behaviors: A comparison of ratings made by mother, father, child and sibling on the CRPBI. Child Development, 56(2), 462–479. Selner-O’Hagan, M. B., Kindlon, D. J., Buka, S. L., Raudenbush, S. W., & Earls, F. J. (1998). Assessing exposure to violence in urban youth. Journal of Child Psychology and Psychiatry, 39, 215–224.

Clin Child Fam Psychol Rev (2009) 12:157–173 Sessa, F. M., Avenevoli, S., Steinberg, L., & Morris, A. (2001). Correspondence among informants on parenting: Preschool children, mothers and observers. Journal of Family Psychology, 15, 53–68. Silvern, L., Karyl, J., Waelde, L., Hodges, W. F., Starek, J., Heidt, E., & Min, K. (1995). Retrospective reports of parental partner abuse: Relationships to depression, trauma symptoms and selfesteem among college students. Journal of Family Violence, 10, 177–202. Spaccarelli, S., Sandler, I. N., & Roosa, M. (1994). History of spouse violence against mother: Correlated risks and unique effects in child mental health. Journal of Family Violence, 9, 79–98. Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and dyads. Journal of Marriage and the Family, 38, 15–28. Steinberg, L., & Avenevoli, S. (2000). The role of context in the development of psychopathology: A conceptual framework and some speculative proposals. Child Development, 71, 66–74. Sternberg, K. J., Knutson, J. F., Lamb, M. E., Bradaran, L. P., Nolan, C., & Flanzer, S. (2004). The Child Maltreatment Log: A PCbased program for describing research samples. Child Maltreatment, 9, 30–48. Sternberg, K. J., Lamb, M. E., & Dawud-Noursi, S. (1997). Using multiple informants and cross-cultural research to study the effects of IPV on developmental psychopathology: Illustrations from research in Israel. In S. S. Luthar, J. A. Burack, D. Cicchetti, & J. Weisz (Eds.), Developmental psychopathology:Perspectives on risk and disorder (pp. 121–156). New York and Cambridge: Cambridge University Press. Stith, S. M., Rosen, K. H., Middleton, K. A., Busch, A. L., Lundeberg, K., & Carlton, R. P. (2000). The intergenerational transmission of spouse abuse: A meta-analysis. Journal of Marriage and the Family, 62, 640–654. Straus, M. A. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics (CT) Scales. Journal of Marriage and the Family, 41, 75–88. Straus, M. A., & Gelles, R. J. (1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and the Family, 48, 465–479.

173 Straus, M. A., & Gelles, R. J. (Eds.). (1990). Physical violence in American families: Risk factors and adaptatiosn to violence in 8, 145 families. New Brunswick, NJ: Transaction Publishers. Straus, M. A., Gelles, R. J., & Steinmetz, S. K. (1980). Behind closed doors: Violence in the American family. Garden City, NY: Doubleday/Anchor. Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996). The revised Conflict Tactics Scales (CTS2): Development and preliminary psychometric data. Journal of Family Issues, 17, 283–316. Sullivan, P. M., & Knutson, J. F. (1998). The association between child maltreatment and disabilities in a hospital-based pediatric sample. Child Abuse and Neglect, 22, 271–288. Sullivan, P. M., & Knutson, J. F. (2000a). The prevalence of disabilities and maltreatment among runaway youth. Child Abuse and Neglect, 24, 1275–1288. Sullivan, P. M., & Knutson, J. F. (2000b). Maltreatment and disabilities: A population-based epidemiological study. Child Abuse and Neglect, 24, 1257–1274. Tein, J., Roosa, M. W., & Michaels, M. (1994). Agreement between parent and child reports on parental behaviors. Journal of Marriage and the Family, 56(2), 341–355. Tienda, M. (1991). Poor people, poor places: Deciphering neighborhood effects on poverty outcomes. In J. Humber (Ed.), Macro-micro linkages in sociology (pp. 244–262). Newbury Park, CA: Sage. U.S. Department of Health, Human Services, Administration on Children, Youth, Families. (2006). Child Maltreatment 2004. Washington, DC: U.S. Government Printing Office. Weissman, M., Orvaschel, H., & Padian, N. (1980). Children’s symptom and social functioning self-report scales: Comparison of mothers’ and children’s reports. The Journal of Nervous and Mental Disease, 168, 736–740. Widom, C. S. (1988). Sampling biases and implications for child abuse research. American Journal of Orthopsychiatry, 58, 260–270. Woffordt, S., Mihalic, D. E., & Menard, S. (1994). Continuities in Marital Violence. Journal of Family Violence, 9, 195–225.

123