Risky relationships: Victims' risk of revictimization of intimate partner violence

(1)

Tilburg University

Risky relationships: Victims' risk of revictimization of intimate partner violence

Kuijpers, K.F.

Publication date: 2011

Document Version

Publisher's PDF, also known as Version of record Link to publication in Tilburg University Research Portal

Citation for published version (APA):

Kuijpers, K. F. (2011). Risky relationships: Victims' risk of revictimization of intimate partner violence. [s.n.].

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal Take down policy

(2)

(3)

Cover design Kitty van der Veer, Proefschrift.nu

Layout Renate Siebes, Proefschrift.nu

Printed by Ipskamp Drukkers B.V.

ISBN 978-94-90791-07-0

(4)

Risky Relationships

Victims’ risk of revictimization of

intimate partner violence

Proefschrift

ter verkrijging van de graad van doctor aan Tilburg University, op gezag van de rector magnificus, prof. dr. Ph. Eijlander, in het openbaar te verdedigen ten overstaan van een door

het college voor promoties aangewezen commissie in de aula van de Universiteit op

vrijdag 9 december 2011 om 10.15 uur door

Karlijn Francisca Kuijpers

geboren op 27 april 1984

(5)

Promotor: Prof. mr. dr. F. W. Winkel

Copromotor: Dr. L. M. van der Knaap

Promotiecommissie: Prof. M. A. Dutton, PhD Dr. A. A. P. van Emmerik Prof. dr. F. Lamers-Winkelman Prof. dr. A. J. J. M. Vingerhoets Prof. dr. J. R. L. M. Wolf

(6)

Introduction

(9)

INTRODUCTION

Intimate partner violence (IPV) is one of the most common forms of interpersonal violence across the world (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002). It has been associated with various serious consequences for its victims; not only injuries and physical complaints are frequently reported (for an overview, see Campbell et al., 2002), but also psychological problems such as depression and posttraumatic stress disorder (PTSD) (Campbell, 2002; Dutton et al., 2006; Foa, Cascardi, Zoellner, & Feeny, 2000; Golding, 1999). Although both men and women are victims of IPV, women suffer the most serious forms of abuse by an intimate partner (Archer, 2000; 2002). Estimates of the proportion of women who are physically assaulted by an intimate male partner at some point in their lives range from 10% to 69% (Krug et al., 2002). Moreover, a large proportion of IPV victims is victimized repeatedly. For instance, the British Crime Survey found that two-thirds of female victims of non-sexual domestic violence (threats or force) had experienced more than one incident in the previous year (Walby & Allen, 2004). Therefore, the prevention of repeat IPV is a relevant issue for practitioners who assist IPV victims with their problems.

During the last decades, numerous studies have been conducted into risk factors for (repeat) IPV, with the majority focusing on a variety of perpetrator characteristics that influence risk, such as prior violence committed by the perpetrator, perpetrators’ anger and hostility, substance abuse, borderline traits and trauma symptoms (see, e.g., Dutton, 1995; Hilton et al., 2004; Norlander & Eckhardt, 2005). Yet, research has shown interventions for male batterers are not very effective in the prevention of repeat partner violence (e.g., Babcock, Green, & Robie, 2004; Feder & Wilson, 2005). Moreover, several studies have reported that characteristics of both members of a couple increase risk for repeat IPV (e.g. Capaldi & Kim, 2007; Moffitt, Robins, & Caspi, 2001; Sonis & Langer, 2008). This would suggest that the prevention of repeat partner violence could benefit from a focus on both perpetrator and victim characteristics. Research into victim-related risk factors for IPV (re) victimization however, seems to lag behind. For instance, Bennett Cattaneo and Goodman (2005) systematically reviewed 64 studies across several disciplines and samples for both perpetrator- and victim-related predictors for reabuse. They concluded that victim-related variables were the “significant minority”, and a “major gap in the extant research” into risk factors for repeat IPV (p. 168). Therefore, more research and knowledge on these variables is necessary. Not only to fill the gap in the literature, but moreover because it may assist practitioners in providing appropriate and efficient services to victims of IPV and may help in optimizing these services. Identifying victim characteristics and behaviors that are associated with being revictimized may evoke the suggestion of “blaming the victim” for the partner violence they experience (see Bennett Cattaneo & Goodman, 2005). Yet,

(10)

we would like to stress that we do not mean to imply that victims share any responsibility for their repeated IPV victimization. We do, however, believe that a focus on factors that victims can influence themselves might, when properly embedded, provide a means to victim control and empowerment (Goodman, Dutton, Vankos, & Weinfurt, 2005; Perez & Johnson, 2008).

For a more complete understanding of the relevance of victims’ characteristics, it is important to keep in mind the ongoing controversies in the field of partner violence and gender (see Langhinrichsen-Rohling, 2010). In the literature, two different perspectives on IPV are described: the gender perspective and mutual IPV perspective. Both of them formulated ideas on how victims of IPV may be helped to stop the violence they experience at the hands of their partners. However, the exact ways in which victims may influence their own IPV revictimization risk clearly differ between both IPV perspectives. First, the gender perspective on IPV suggests that IPV is characterized by one-sided violence, initiated by the male partner with the objective to dominate and control his wife (see for example Dobash & Dobash, 1979; Walker, 2009). Within this perspective, Walker (2009) described the so-called cycle of violence. In this cycle, a phase of tension building precedes the acute battering incident which is followed by a period of loving-contrition or absence of tension. A certain proportion of victims experiences this cycle over and over again, and according to Walker (1979), “repeated batterings, like electrical shocks, diminish the woman’s motivation to respond” (p. 50). According to the gender perspective, empowerment of the female victim may help her to be able to stop the cycle of IPV revictimization. This empowerment typically focuses on environmental and societal victim-related factors, such as the promotion of sexual equality, improving opportunities for women’s employment and economic power, and improving levels of female education (see e.g. Jewkes, 2002). On an individual level, victims’ empowerment may include women’s realistic assessment of responsibility for the violence, as well as a realistic evaluation of individual resources available for stopping it (Peled, Eisikovits, Enosh & Winstok, 2000).

The other perspective on IPV, which is labeled the mutual IPV perspective, holds that both men and women perpetrate partner violence approximately to an equal extent (see Archer, 2000) and that IPV can be initiated by men but also by women (see Stets & Straus, 1990). According to this perspective, IPV can mainly be described as an emotional response in reaction to an unpleasant experience and cycles of revictimization are thought to be characterized by mutual emotional aggression (see for example Dutton, 2008; Stets & Straus, 1989). Within this perspective, the role of victims’ own behavior in relation to IPV risk is much more emphasized. While the gender perspective emphasizes the individual strengths available in the victim that may help stop the violence, the mutual IPV perspective also makes explicit victims’ own risky characteristics and behaviors that may contribute

(11)

to IPV revictimization (see e.g. Moffitt et al., 2001). Within the mutual IPV perspective, ideas on how victims may stop cycles of IPV revictimization are focused on the treatment or change of these victim characteristics and behaviors.

While we did not set out to test either one of these perspectives, our study fits best in the mutual IPV perspective as the aim of the current thesis was to prospectively examine victim-related risk factors for revictimization of partner violence. We conducted a systematic review of prior prospective research in this area and set up a comprehensive, prospective study into victim-related risk and protective factors among a sample of 166 help-seeking victims of IPV in the Netherlands. As stated above, we do not mean to raise the suggestion that victims are to blame for what happened to them. We do, however, believe that research on victim-related risk factors for IPV revictimization is necessary in order to increase our knowledge in this relatively scarcely studied field. Moreover, this knowledge may assist practice in developing and optimizing victim interventions to decrease partner violence. In the next section, we will describe how the current thesis contributes to the literature. Furthermore, we will discuss our study design, the recruitment of participants and the procedure of the study. We will end this chapter by providing an outline of the thesis.

Contributions to the literature

The current thesis contributes to the literature in several ways. Firstly, our prospective study among 166 IPV victims extensively and repeatedly measures not only IPV committed by the perpetrator, but also IPV perpetrated by victims themselves in order to see whether and how these forms of IPV contribute to risk for IPV revictimization. In many studies, researchers focus “solely on men’s [i.e., perpetrators’] perpetration of intimate partner violence, while ignoring women’s [i.e., victims’] perpetration of violence within the same context” (see Langhinrichsen-Rohling, 2010, p. 183). This limited approach to the problem of partner violence “seems likely to hamper rather than advance the field” (p. 183), particularly because victims’ own perpetration of partner violence has been reported to be related to (re)victimization of IPV (e.g. Capaldi, Shortt & Crosby, 2003; Stith, Smith, Penn, Ward, & Tritt, 2004). In other words, victims’ use of violence is suggested to play an important role in maintaining cycles of IPV revictimization. In the literature, only a limited number of studies have examined the relation between victim-perpetrated IPV and risk for future IPV. Moreover, this relation was mainly found based on cross-sectional data (Anderson, 2002; Fergusson, Horwood, & Ridder, 2005; Stith et al., 2004). The current study therefore makes an important contribution to the extant knowledge by examining victims’ use of partner violence and risk for IPV revictimization in a prospective design. In addition, the majority of studies into risk assessment of repeat partner violence have focused on official

(12)

IPV recidivism based on police records as an outcome measure (e.g. Hilton et al., 2004; Kropp, Hart, & Belfrage, 2005), whereas the current study focuses on IPV revictimization reported by victims themselves. As not all IPV incidents are reported to the police, using victims’ self-reports might provide a more complete insight in the partner violence that occurs within a couple. Although the number of studies is growing, the body of literature considering victims’ reports of IPV revictimization is still small. Furthermore, a variety of individual victim factors, such as anger, anxiety, attachment, PTSD, borderline traits and quality of life are assessed for their relation to IPV risk simultaneously and prospectively in the current study. The simultaneous assessment of multiple factors offers the possibility to examine which subset of victim-related factors most strongly predicts IPV revictimization (see chapter 5), whether the identified relations between predictors and IPV revictimization may be mediated by some other victim characteristics (see chapter 6), and if certain risk factors show an interaction with each other (see chapter 4). Next, the current thesis contributes to the literature by focusing on dynamic victim-related factors. Prior studies often did not specifically concentrate on dynamic factors and included a wide range of static victim factors, such as victims’ age, education, ethnicity and socio-economic status, as possible predictors for repeat IPV (e.g., Carlson, Harris, & Holden, 1999; Frias & Angel, 2007; Mears, Carlson, Holden, & Harris, 2001). Yet, dynamic victim-related factors are factors that can be changed (in contrast to static factors) and thus are most appropriate for interventions.

Recruitment of participants and study design

Participants of the study were recruited from various victim support services in the Netherlands in four large and four medium-sized cities. The services included a victim support office, a mental health organization, a number of social work organizations, a shelter, victim service points of the Public Prosecution service, and various local domestic violence teams of a health authority. Participants were approached and included in our study 1) if they had been a victim of IPV at least once in the past two years, and 2) if they sufficiently mastered the Dutch language to understand the Dutch questionnaires we used. Participants were considered to be a victim of IPV if they had been abused physically, sexually or psychologically by their current or ex-partner. Victims were recruited through the collaborating victim support organizations by having staff inform eligible clients about this study. Most victims were directly contacted by staff members; others were informed about the study through a letter. Clients who indicated interest in participating received a registration form asking them to provide some personal data (name, address, phone number, and e-mail address) and to return it to the researchers. A researcher then

(13)

telephoned registered participants to discuss any questions about the study they might have. Participants were asked to take part in the study at three different moments in time: after the initial assessment at baseline (Time 1), assessments were repeated two (Time 2) and six months later (Time 3). During each assessment, they had to fill in a questionnaire eliciting information on a wide range of variables, including the partner violence that occurred, participants’ (psychological) complaints, and some demographics of the participant and the perpetrator of the violence. Participants could choose whether they preferred to fill in an online version or a paper version of the questionnaire. It was also possible to plan a personal appointment with a researcher to complete the questionnaire. For any questions during completion of the questionnaire, participants could phone or e-mail the researchers. After completing the questionnaire at all three moments of data collection, they were paid a 100 euro compensation for their time. Data that are reported in this thesis were collected between August 2008 and August 2010.

Figure 1.1 shows the number of victims that registered for participation and the number included in each moment of the study. In total, 181 victims registered for participation, of which 166 victims filled in the questionnaire at Time 1, 162 victims at Time 2 and 159 victims at Time 3. Staff of collaborating victim organizations was instructed

Introduction Chapter 1

Figure 1.1 Flowchart of study participants.

Figure 1. Flowchart of Study Participants

Time 1 166 victims Time 2 162 victims Time 3 159 victims 181 victims registered for

participation 15 victims did not participate:  6 did not meet inclusion criteria  6 were unreachable

 2 indicated they had no time  1 gave no reason

4 victims dropped out:  3 were unreachable  1 indicated participation was

too much of a burden

(14)

to report the total number of victims they invited to participate, including those victims who declined. However, the number that was invited but declined was not reported systematically and hence, we do not know the exact number of victims invited for the study. We only know that eventually 134 victims registered for participation after they were contacted personally about the study by staff members. Yet, for the group of victims that was informed about the study through a letter, we do have exact information. In total, 175 victims received a letter about the study which offered them the choice to participate. As 47 of them registered for participation, the response rate was 26.8%.

Outline of the thesis

Chapter 2 describes a systematic review of prospective studies into victim-related risk and

protective factors for IPV revictimization, against the background of conceptual models of women’s influence on partner violence developed by Foa et al. (2000). It provides the reader with an overview of prior research in this area, before presenting this thesis’ main study on victim-related risk factors for revictimization of partner violence.

In chapter 3, data from the baseline questionnaire (Time 1) of our study are presented. As these data did not yet allow statements concerning (prospective) risk factors for IPV revictimization, we decided to focus on a frequently reported consequence of IPV victimization: victims’ PTSD. Apart from being a consequence of IPV, PTSD can also be a risk factor for future IPV victimizations (see e.g., Perez & Johnson, 2008). In that way, victims of IPV who are at greater risk of developing PTSD symptoms may also be in greater danger for IPV revictimization. In chapter 3 we examine how victims’ borderline traits influence the relation between IPV victimization and development of PTSD symptomatology.

In chapter 4, data from Time 1 and 2 of our study are analyzed. It investigates the victim-related psychological mechanisms that may underlie the link between victimization and revictimization of IPV. Hypotheses regarding possibly relevant mechanisms are derived from two conflicting approaches to IPV: the gender perspective and the mutual IPV perspective. Following these perspectives, the role of victims’ attachment style, victims’ anger and victim-perpetrated IPV in explaining risk for IPV revictimization is examined.

Chapter 5 presents data from all three moments of data collection in our study

(Time 1, 2 and 3). In accordance with the models articulated by Foa et al. (2000), we empirically tested how the three key factors from their models - partner violence, victims’ psychological difficulties and victims’ resilience - related to the occurrence of any future IPV revictimization as well as the severity of future IPV revictimization.

In chapter 6, data from all three moments of data collection are presented; however,

(15)

main analyses are conducted on data from Time 2 and 3. It examines how each of four PTSD symptom clusters (re-experiencing; arousal; avoidance; and numbing) relates to IPV revictimization. Furthermore, we analyze if victims’ own perpetration of partner violence mediates the relation between PTSD and IPV revictimization.

Finally, chapter 7 summarizes the main findings of this thesis (presented in chapter 2 to 6), followed by an overall discussion of these findings. Furthermore, some methodological considerations related to the design of our prospective study are addressed, including possible selection and information biases, and some notes concerning causality. The chapter closes with several recommendations for practice and future research.

REFERENCES

Anderson, K. L. (2002). Perpetrator or victim? Relationships between intimate partner violence and well-being. Journal of Marriage and Family, 64, 851–863.

Archer, J. (2000). Sex differences in aggression between heterosexual partners: A meta-analytic review. Psychological Bulletin, 126, 651–680.

Archer, J. (2002). Sex differences in physically aggressive acts between heterosexual partners: A meta-analytic review. Aggression and Violent Behavior, 7, 313–351.

Babcock, J. C., Green, C. E., & Robie, C. (2004). Does batterers’ treatment work? A meta-analytic review of domestic violence treatment. Clinical Psychology Review, 23, 1023–1053.

Bennett Cattaneo, L., & Goodman, L. A. (2005). Risk factors for reabuse in intimate partner violence: A cross-disciplinary critical review. Trauma, Violence, & Abuse, 6, 141–175.

Campbell, J.C. (2002). Health consequences of intimate partner violence. The Lancet, 359, 1331–1336. Campbell, J. C., Jones, A. S., Dienemann, J., Kub, J., Schollenberger, J., O’Campo, P., Gielen, A. C., &

Wynne, C. (2002). Intimate partner violence and physical health consequences. Archives of Internal Medicine, 162, 1157–1163.

Capaldi, D. M., & Kim, H. K. (2007). Typological approaches to violence in couples: A critique and alternative conceptual approach. Clinical Psychology Review, 27, 253–265.

Capaldi, D. M., Shortt, J. W., & Crosby, L. (2003). Physical and psychological aggression in at-risk young couples: Stability and change in young adulthood. Merrill-Palmer Quarterly, 49, 1–27.

Carlson, M. J., Harris, S. D., & Holden, G. W. (1999). Protective orders and domestic violence: Risk factors for reabuse. Journal of Family Violence, 14, 205–226.

Dobash, R. E., & Dobash, R. P. (1979). Violence against wives: A case against patriarchy. New York: Free Press.

Dutton, D. G. (1995). A scale for measuring propensity for abusiveness. Journal of Family Violence, 10, 203–221.

Dutton, D. G. (2008). My back pages: Reflections on thirty years of domestic violence research. Trauma, Violence, & Abuse, 9, 131–143.

Dutton, M. A., Green, B. L., Kaltman, S. I., Roesch, D. M., Zeffiro, T. A., & Krause, E. D. (2006). Intimate partner violence, PTSD, and adverse health outcomes. Journal of Interpersonal Violence, 21, 955–968. Introduction

(16)

Feder, L., & Wilson, D. B. (2005). A meta-analytic review of court mandated batterer intervention programs: Can courts affect abusers’ behavior? Journal of Experimental Criminology, 1, 239−262. Fergusson, D. M., Horwood, L. J., & Ridder, E. M. (2005). Partner violence and mental health outcomes

in a New Zealand birth cohort. Journal of Marriage and Family, 67, 1103–1119.

Foa, E. B., Cascardi, M., Zoellner, L. A., & Feeny, N. C. (2000). Psychological and environmental factors associated with partner violence. Trauma, Violence, & Abuse, 1, 67–91.

Frias, S. M., & Angel, R. J. (2007). Stability and change in the experience of partner violence among low-income women. Social Science Quarterly, 88, 1281–1306.

Golding, J. M. (1999). Intimate partner violence as a risk factor for mental disorders: A meta analysis. Journal of Family Violence, 14, 99–132.

Goodman, L., Dutton, M. A., Vankos, N., & Weinfurt. K. (2005). Women’s resources and use of strategies as risk and protective factors for reabuse over time. Violence Against Women, 11, 311–336. Hilton, N. Z., Harris, G. T., Rice, M. E., Lang, C., Cormier, C. A., & Lines, K. J. (2004). A brief actuarial

assessment for the prediction of wife assault recidivism: The Ontario Domestic Assault Risk Assessment. Psychological Assessment, 16, 267–275.

Jewkes, R. (2002). Intimate partner violence: Causes and prevention. The Lancet, 359, 1423–1429. Kropp, P. R., Hart, S. D., & Belfrage, H. (2005). Brief Spousal Assault Form for the Evaluation of Risk

(B-SAFER): User manual. Vancouver, British Columbia: ProActive ReSolutions.

Krug, E. G., Dahlberg, L. L., Mercy, J. A., Zwi, A. B., & Lozano, R. (2002). World report on violence and health. Geneva: World Health Organization.

Langhinrichsen-Rohling, J. (2010). Controversies involving gender and intimate partner violence in the United States. Sex Roles, 62, 179–193.

Mears, D. P., Carlson, M. J., Holden, G. W., & Harris, S. D. (2001). Reducing domestic violence revictimization: The effects of individual and contextual factors and type of legal intervention. Journal of Interpersonal Violence, 16, 1260–1283.

Moffitt, T. E., Robins, R. W., & Caspi, A. (2001). A couples analysis of partner abuse with implications for abuse-prevention policy. Criminology and Public Policy, 1, 5–37.

Norlander, B., & Eckhardt, C. (2005). Anger, hostility, and male perpetrators of intimate partner violence: A meta-analytic review. Clinical Psychology Review, 25, 119–152.

Peled, E., Eisikovits, Z., Enosh, G., & Winstok, Z. (2000). Choice and empowerment for battered women who stay: Toward a constructivist model. Social Work, 45, 9–25.

Perez, S., & Johnson, D. M. (2008). PTSD compromises battered women’s future safety. Journal of Interpersonal Violence, 23, 635–651.

Sonis, J., & Langer, M. (2008). Risk and protective factors for recurrent intimate partner violence in a cohort of low-income inner-city women. Journal of Family Violence, 23, 529–538.

Stets, J. E., & Straus, M. A. (1989). The marriage license as a hitting license: A comparison of assaults in dating, cohabiting and married couples. Journal of Family Violence, 4, 161–180.

(17)

Stith, S. M., Smith, D. B., Penn, C. E., Ward, D. B., & Tritt, D. (2004). Intimate partner physical abuse perpetration and victimization risk factors: A meta-analytic review. Aggression and Violent Behavior, 10, 65–98.

Walby, S., & Allen, J. (2004). Domestic violence, sexual assault and stalking: Findings from the British Crime Survey. Home Office Research Study 276. London: Home Office.

Walker, L. E. (1979). The battered woman. New York: Harper & Row.

Walker, L. E. A. (2009). The battered woman syndrome (3rd ed.). New York: Springer. Introduction

(18)

Victims’ influence on

intimate partner violence

revictimization: A systematic

review of prospective evidence

Karlijn F. Kuijpers Leontien M. van der Knaap

Ilse A. J. Lodewijks (2011)

Trauma, Violence, & Abuse, 12, 198-219

(19)

ABSTRACT

Foa, Cascardi, Zoellner and Feeny (2000) developed two models of women’s influence on intimate partner violence (IPV) which integrate victim-related variables associated with the cessation or continuation of partner violence (i.e., repeat IPV). One of the models focuses on psychological factors while the other centers on environmental factors. Central to both models are three key factors: partner violence; psychological difficulties; and resilience. Despite the appeal of these models, empirical, prospective research that specifically tests these models appears to be lacking. This article describes a systematic review of the available literature that examines the prospective link between the three key factors of the models and the risk of IPV revictimization. A synthesis of 15 studies reveals that Foa et al.’s models of revictimization are partly supported by prior prospective research. It is beyond doubt that the key factor partner violence (involving the severity and frequency of prior IPV) is a strong predictor for IPV revictimization; the evidence regarding victims’ psychological difficulties and resilience is more mixed. Findings are discussed in terms of implications for practice and research and might enable practitioners to help victims to take control of their situations and to contribute to their empowerment. The importance of future prospective research into dynamic, victim-related variables is emphasized, in order to further support Foa’s models of victims’ influence on IPV revictimization.

(20)

INTRODUCTION

Intimate partner violence (IPV) is one of the most pervasive social problems all over the world. Although both men and women are victims of IPV, women suffer the most serious forms of abuse by an intimate partner (Archer, 2000; 2002). Estimates of the proportion of women who are physically assaulted by an intimate male partner at some point in their lives range from 10% to 69% (Krug, Dahlberg, Mercy, Zwi, & Lozano, 2002). Furthermore, a large proportion of IPV victims is victimized repeatedly (Walby & Allen, 2004). During the last decades, a large amount of research has focused on a wide array of perpetrator characteristics that influence risk for (repeat) partner violence (see, e.g., Dutton, 1995; Hilton et al., 2004; Norlander & Eckhardt, 2005). Surprisingly, though, research on victim-related risk factors seems to lag behind. For instance, Bennett Cattaneo and Goodman (2005) systematically reviewed 64 studies across several disciplines and samples for both perpetrator- and victim-related predictors for reabuse. They concluded that victim-related variables were the “significant minority”, and a “major gap in the extant research” into risk factors for repeat IPV (p. 168). Knowledge on risk and protective factors that victims themselves can influence is important, however. For one, interventions for male batterers have shown mixed results with regard to their efficacy (e.g., Babcock, Green, & Robie, 2004; Feder & Wilson, 2005). Second, research also shows that characteristics of both members of the couple increase risk for IPV (Moffitt, Robins, & Caspi, 2001) which means that interventions to prevent future violence could be more effective if they also target victims of IPV. Most importantly, though, knowledge on risk and protective factors that are within victims’ sphere of influence can help victims take control and thereby empower them (Goodman, Dutton, Vankos, & Weinfurt, 2005; Perez & Johnson, 2008). Or, as articulated by Foa, Cascardi, Zoellner and Feeny (2000), psychological and environmental interventions for victims “will facilitate women’s agency in reducing partner violence” (p. 69).

In order to stimulate the development of forms of victim support that enable practitioners to help victims take control over their lives, Foa and her colleagues (2000) developed two models of women’s influence on IPV which integrate victim-related variables associated with the cessation or continuation of partner violence (i.e., repeat IPV). One of the models focuses on psychological factors while the other centers on environmental factors. Central to both models are three key factors: partner violence; psychological difficulties; and resilience. Because these factors are multidimensional and complex, they are described along multiple aspects.

First, Foa et al.’s (2000) key factor partner violence involves the severity and frequency of prior partner violence which is hypothesized to relate to repeat IPV. This key factor is operationalized as incorporating physical and emotional abuse, perceived threat, and

(21)

psychological reactions to the abuse. In defining their construct of partner violence, Foa et al. refer to literature showing that physical and emotional abuse are distinct but highly correlated phenomena. According to various studies, emotional abuse is more common than physical abuse (e.g., Dutton & Starzomski, 1993; Straus, Gelles, & Steinmetz, 1980), emotional and physical abuse often occur together (Follingstad, Rutledge, Berg, Hause, & Polek, 1990; Walker, 1984), and emotional abuse predicts future physical abuse (Malamuth, Linz, Heavey, Barnes, & Acker, 1995; O’Leary, Malone, & Tyree, 1994). As a second aspect of partner violence, perceived threat involves the victim’s assessment of the likelihood of future abuse by her partner. This perceived threat not only concerns fear for the victim’s own safety, but also for the safety of her children, family and friends (Campbell, 1995). Foa and her colleagues hypothesized that the victim’s own perception of threat within the relationship might be an important predictor of actual partner violence. Lastly, psychological reactions of the victim to the abuse are assumed to form part of the construct of partner violence as well. They involve the victim’s perceptions of susceptibility to physical and psychological danger, and the loss of power and control within the violent relationship. These psychological reactions are thought to distinguish battered women from non-battered women (see Smith, Tessaro, & Earp, 1995). Although victims’ perceptions of susceptibility to physical and psychological danger seem to be similar to victims’ perceived threat described above, Foa et al. suggest perceived threat to refer to danger from the partner in specific, whereas the other seems to concern a more general susceptibility to physical and psychological danger.

The second key factor of Foa et al.’s (2000) model of women’s influence on IPV relates to psychological difficulties; victims of partner violence suffer from a variety of psychological difficulties (e.g., Follingstad, Hause, Rutledge, & Polek, 1992). During the last decades, research has mainly focused on symptoms of depression, anxiety, posttraumatic stress disorder and substance abuse among IPV victims (Campbell, 2002; Coker et al., 2002; Dutton et al., 2006; Follingstad, Brennan, Hause, Polek, & Rutledge, 1991; Golding, 1999; Kemp, Green, Hovanitz, & Rawlings, 1995; Pico-Alfonso, Garcia-Linares, Celda-Navarro, Herbert, & Martinez, 2004). These adversities are classified as psychological difficulties in Foa et al.’s models, hypothesizing that they serve as precipitating factors by increasing a victim’s risk for future partner violence.

Finally, the third key factor hypothesized to affect women’s influence on IPV is resilience, which involves the ability to successfully cope with, adjust to, or recover from major life stressors, such as partner violence. Resilience can thus be viewed as a protective factor against IPV revictimization. Indeed, battered women employ a wide variety of actions to escape, avoid and protect themselves (Campbell, Rose, Kub, & Nedd, 1998). According to Foa et al. (2000), a central, relatively stable characteristic to resilience is the

(22)

victim’s physical health. Research showed a positive association between health problems and the continuation of partner violence (Campbell & Soeken, 1999). Furthermore, Foa and colleagues suggest optimism, self-esteem, and cognitive flexibility (i.e., the ability to perceive different aspects of events) to be important correlates of physical health and to serve to promote victims’ resilience as well. Therefore, they hypothesize that optimism, self-esteem, flexibility and physical health of the victim combine to represent resilience.

Central to the models of women’s influence on repeat IPV are the mechanisms between the above key factors: partner violence; psychological difficulties; and resilience. Foa et al. (2000) hypothesize that partner violence and psychological difficulties interact in a vicious cycle whereby partner violence causes psychological difficulties that, in turn, put women at greater risk of revictimization by hindering the victim’s ability to curtail future violence. They further argue that victims’ intra-personal resources—resilience—temper the negative psychological impact of partner violence and, thereby, serve to reduce the risk of revictimization.

The relevance of these conceptual models in the field of IPV revictimization is apparent as Foa et al. (2000) are in this journal’s top 10 of most cited articles.1_However,

there appears to be little empirical, longitudinal research that specifically tests these models. Although Foa et al. based their models on a review of the literature, most research was cross-sectional which only allows statements regarding correlates of IPV (re)victimization. In order to describe factors that predict future IPV revictimization, studies with a prospective design are required. We therefore set out to systematically review the existing literature that prospectively links (aspects of) Foa et al.’s three key factors – partner violence; psychological difficulties; and resilience – to the risk for IPV revictimization.

METHOD

Literature search

The search for relevant studies to include in our systematic review was performed using search term combinations including IPV-related terms (domestic*, intimate*, partner*, viol*), revictimization and risk-related terms (victim*, revictim*, risk*, protect*, vulnerab*, trauma*) and terms relating to the type of studies we were looking for (empiric*, prospect*). All aspects of the three key factors in Foa et al.’s (2000) models concern possible risk and

21 Syst ema tic r eview of vic tims ’ influenc e on IPV r evic timiza tion Chapt er 2

(23)

protective factors for revictimization of IPV. We therefore deliberately used these more general search terms instead of terms related to specific key factors to make sure that we would not miss any prospective study in our field of interest. Combinations of above search terms were entered in a variety of databases (Tilburg University Catalogue, JSTOR, Netherlands Central Catalogue, Online Contents book chapters and journal articles Tilburg University and national, PsychArticles, Psychology and Behavioral Sciences Collection, PsychInfo, PubMed, SAGE Journals Online, ScienceDirect, Social Services Abstracts, Sociological Abstracts, SpringerLink, SSRN, Tilburg University Repository, Web of Science, Wiley InterScience) and internet search engines (Google / Google Scholar). Furthermore, we examined the reference sections of the studies we decided to include in our review for other potentially relevant studies. We performed our literature search from September 21, 2009 until November 3, 2009, a period of six weeks.

Selection of literature

Studies were included if they used a prospective design; aspects of the key factors had to be measured at a time point prior to the measurement of revictimization of IPV. Furthermore, they had to include revictimization of IPV as an outcome measure, which was defined in the current review as reoccurrence of any physical, psychological, and / or sexual violence, injuries, and / or threats of violence perpetrated by a current partner or ex-partner. Lastly, studies had to report on at least one aspect of one of the three key factors of Foa et al.’s (2000) models in relation to revictimization of IPV to be included in our review. Our literature search resulted in a total number of 219 studies that seemed relevant for our systematic review on the basis of their title. After reading their abstract, the number of possibly relevant studies was further reduced to 44 studies. Of these studies we obtained and read the full article. Twenty-nine studies were excluded after closer reading. Of these, 13 studies measured victimization of IPV among a sample including both victims and non-victims (e.g., Ehrensaft et al., 2003; Moffitt et al., 2001). However, for these studies no information was given on whether these victimization reports actually concerned a first IPV victimization or an IPV revictimization. Similarly, it was not clear whether the reported factors were risk factors for a first IPV victimization or an IPV revictimization and therefore we decided to exclude this type of studies. Furthermore, 9 studies were excluded because the design was not prospective (i.e., risk factors relevant for our review were not measured at a time point prior to the measurement of revictimization of IPV), and 7 studies were excluded for other reasons (e.g., no aspects of key factors included). This resulted in a total of 15 studies that eventually met the above inclusion criteria (Table 2.1). For the majority of the studies, the outcome variable revictimization of IPV was assessed

(24)

based on victim self-reports, but we also included one study that used police and court records (Mears, Carlson, Holden, & Harris, 2001). We did not formulate any restrictions with regard to publication year, but the vast majority of the included studies has been published in the past 10 years, indicating the relatively recent interest in victim-related risk factors for repeat IPV. Another similarity across the 15 included studies concerns the fact that they were all conducted in the United States of America.

Summarizing and scoring risk and protective factors

The 15 studies that were included in the review were summarized according to a fixed format in which we documented information about the sample of the study, research design, data and analysis plan and the results of the study in terms of identified risk and protective factors for revictimization of IPV. Two studies were summarized by both the first and third author, to reach agreement on what we considered relevant information, and how and at what point in our format this information should be documented. All following studies were summarized by either the first or the third author. After completion, they were read by the other person in order to see if there was any ambiguous information that had to be clarified. In summarizing the results of the included studies we specifically focused on risk and protective factors that relate to the key variables included in Foa et al.’s (2000) predictive models of women’s influence on partner violence. In reporting our findings we decided to present the results from multivariate, rather than bivariate analyses, when available.

Description of the selected studies

The current systematic review is based on 15 studies that prospectively link aspects of the three key factors from Foa et al.’s (2000) models to the risk for IPV revictimization. Although these studies showed similarities on the inclusion criteria, differences can be identified as well, for instance in the nature of the sample or length of follow-up time. Because these differences might in part account for differences in results across studies, we included them in the overview of reviewed studies (Table 2.1) and describe them in more detail below.

Nature of the sample

All 15 studies used a victim sample. For seven of them, the sample consisted of women who had contacts with the police or court, three studies used a sample of women who had been living or still lived in a shelter, another three studies included women seeking health care at medical sites, and two studies recruited their samples from a combination of a shelter,

(25)

Systematic review of victims’ influence on IPV revictimization Chapter 2

Table 2.1

O

ver

view of included studies

Study Na tur e of sample Follo w -up time a Sour ce of follo w -up inf or ma tion Amoun t of revic timiza tion of IPV dur ing f ollo w -up per iod D efinition and measur emen t of revic timiza tion of IPV a t f ollo w -up (dependen t v ar iable) M easur emen t of aspec ts of key fac tors fr om F oa ’s models (independen t var iables) Sta tistical analy ses Repor ted aspec ts of key fac tors fr om

Foa´s models and their r

ela tion with revic timiza tion of IPV (+ = positiv e, - = nega tiv e and 0 = no rela tionship )

Bennett Cattaneo & Goodman (2003)

96 w omen who appear ed a t a c our t in take c en ter in W ash -ingt on (USA ) f ollo wing the ar rest of an abusiv e par tner . O f baseline sample ( N=169) 91.1% w as A frican A mer ican, 1.8% C aucasian, 1.2% La tina, 5.9% missing or other . P rimar ily lo w -inc ome w omen 3 mon ths Vic tim in ter -view 27.7% An y ph ysical abuse , ph ysical thr ea ts , se xual abuse , and/ or destruc tion of the vic tim ’s pr oper ty b y the inde x par tner , and/or un w an ted con tac t bet w een vic

tim and inde

x par tner , measur ed by sev er al it ems (br oad definition) Ph ysical abuse: C TS, psy -cholog ical abuse: P MWI, vic tim ’s assessmen t: one

question about their own assessmen

t of dan -ger ousness , of ten within 24 h af ter off ense Biv ar ia te: opti -mal da ta analy -sis ( OD A) b Ph ysical abuse in y ear pr ior t o baseline (+), dominanc e/isola tion rela ted psy cholog ical abuse in y ear pr ior to baseline (+), vic -tim ’s assessmen t of likelihood of fur ther abuse (+)

Bybee & Sulliv

an (2002) 267 w omen of a do -mestic violenc e shelt er pr og ram in M ichigan (USA ). O f baseline sample ( N=278) 45% w as A frican A mer ican, 42% E ur opean A mer i-can, 7% La tina, 2% Asian A mer ican, 4% Na tiv e A mer ican, A rab Amer ican or mix ed her itage . P rimar ily lo w -inc ome w omen 1 y ear Vic tim in ter -view Not r epor ted Sev er ity of ph ysical abuse per petr at ed by an y par tner (or ig

i-nal and/or new), measur

(26)

Bybee & Sulliv

an (2005) 123 w omen of a do -mestic violenc e shelt er pr og ram in M ichigan (USA ). O f baseline sample ( N=141) 46% w as W hit e, 42% Afr ican A mer ican, 7% La tina, 2% A sian Amer ican, 4% Na tiv e Amer ican, A rab A mer i-can or mix ed her itage . Pr imar ily lo w -inc ome w omen 1 y ear Vic tim in ter -view 19% Sev er ity of ph ysical abuse per petr at ed by an y par tner (or ig

ed b y C TS Ph ysical abuse: C TS, qualit y of lif e: 9-it em scale adapt ed fr om A n-dr ew s & W ithey (1976), social suppor t: 9-it em scale A dult ’s S ocial Sup -por t Q uestionnair e M ultiv ar ia te: hier ar chical log istic r eg res -sion analy sis Ph ysical abuse in 6 mon ths pr ior t o base -line (+), vic tim ’s qual -ity of lif e (-), vic tim ’s social suppor t (-) Cole , L ogan, & Shannon (2008) 632 w omen who rec en

tly had obtained

a PO against a male intima

(27)

Systematic review of victims’ influence on IPV revictimization Chapter 2 Table 2.1 Contin ued Study Na tur e of sample Follo w -up time a Sour ce of follo w -up inf or ma tion Amoun t of revic timiza tion of IPV dur ing f ollo w -up per iod D efinition and measur emen t of revic timiza tion of IPV a t f ollo w -up (dependen t v ar iable) M easur emen t of aspec ts of key fac tors fr om F oa ’s models (independen t var iables) Sta tistical analy ses Repor ted aspec ts of key fac tors fr om

ela tion with revic timiza tion of IPV (+ = positiv e, - = nega tiv e and 0 = no rela tionship ) Cr andall , Na thens , Ker nic , Holt , & R iv ar a (2004) 354 w omen with a polic e- or c our t-r e-por

ted episode of IPV

(W ashingt on, USA ). O f these w omen, 57.1% w as C aucasian, 19.5% Afr ican A mer ican, 7.1% A sian or P acific Islander , 16.4% other .

51.1% full-time and 48.3% under full-time emplo

ymen t. 9 mon ths Vic tim in ter -view 17.2% An y IPV r ela ted injur y inflic ted by inde x par tner , measur ed with one question about an y injur ies e xper ienc ed due t o IPV Ph ysical abuse: C TS, psy cholog ical abuse: measur e not r epor ted , clinical depr ession: CES -D i, alc ohol abuse: NE T measur e j, drug abuse: measur e not r epor ted M ultiv ar ia te: st ep wise log is-tic r eg ression analy sis Ph ysical abuse on the inciden t da te (+), pr ior psy cholog ical abuse (0), vic tim ’s clinical depr ession (0), vic tim ’s alc ohol abuse (0), vic tim ’s drug abuse (0) Fleur y, Sulliv an, & Bybee (2000) 135 w

omen who had

lef t a domestic violenc e shelt er pr og ram in a medium-siz ed industr i-al cit

y the USA and who

w er e still separ at ed from their ba tter ers 10 w eeks af ter shelt er e xit . O f these w omen, 54% w as A frican A mer ican, 36% W hit e, 6% La tina, 1% A sian A mer ican, 3% other . 42% w as emplo yed 2 y ears Vic tim in ter -view 36% An y ph ysical abuse per petr at ed b y inde x par tner , measur ed by C TS Ph ysical violenc e: C TS,

emotional abuse: two questions about frequenc

y of thr ea ts b y per petr at or M ultiv ar ia te: ev en t hist or y analy sis (this

method uses logistic r

(28)

27 Syst ema tic r eview of vic tims ’ influenc e on IPV r evic timiza tion Chapt er 2 G oodman, D utt on, Vankos , & W einfur t (2005) 324 w

omen who had

been vic

tims of IPV in

the past y

ear and who

w er e seek ing ser vic es for violenc e a t 3 sit es

in the USA (shelt

er , pr ot ec tion or der c our t and cr iminal c our t). O f these w omen, 80.5% w as A frican A mer ican, 19.5% other . 60.4% w as emplo yed 1 y ear Vic tim in ter -view 38.3% An y ph ysical and/or se

xual abuse per

-petr at ed b y inde x par tner , measur ed by C TS Social suppor t: I nt er per -sonal Suppor t E valua -tion List k, qualit y of lif e: 9-it em scale adapt ed from A ndr ew s & W ithey (1976) M ultiv ar ia te: log istic r eg res -sion analy sis Social suppor t vic tim (-), vic tim ’s qualit y of life (0) l H irschel & Hut chison (2003) 240 f emale vic tims

who had called t

o the polic e f or an inciden t of IPV ( Char lott e, Nor th C ar olina, USA ). O f baseline sample _(N=419) 70.3% w as Black , 29.0% W hit e, 0.7% other . M ost w omen w er e a t the po ver ty lev el and w or

king class lev

el close t o 6 mon ths m Vic tim in ter -view n Not r epor ted An y ph ysical abuse and/or psy cholog ical abuse per petr at ed by inde x par tner , measur ed b y C TS

and some additional items fr

om Na

tional

Institut

e of Justic

e-funded Spouse Assault R

eplica tion Pr ojec ts c onc er ning e.g . thr ea tening or hitting D esir e f or ar rest: ques

-tion about wha

t vic tim w an ted t o happen when

they called the polic

e,

ph

ysical IPV

: question

about number of times victim w

as hit M ultiv ar ia te: log istic r eg res -sion analy sis Vic tim desir e f or ar rest (+), fr equenc y of ph

ysical IPV bet

w een IPV inciden t f or which the vic

tim called the

(29)

ela tion with revic timiza tion of IPV (+ = positiv e, - = nega tiv e and 0 = no rela tionship ) Kr ause ,

Kaltman, Goodman, & D

utt

on

(2006)

324 w

omen who had

been vic

tims of IPV in

the past y

ear and who

w er e seek ing ser vic es for violenc e a t 3 sit es

in the USA (shelt

er , pr ot ec tion or der c our t and cr iminal c our t). O f baseline sample _(N=405) 81% w as Afr ican A mer ican, 13% W hit e, 1% La tina, 5% other . P rimar ily lo w -inc ome w omen 1 y ear Vic tim in ter -view 36.7% An y ph ysical and/or se

xual abuse per

-petr at ed b y inde x par tner , measur ed by C TS Ph ysical and se xual IPV : CT S, PT SD: PT SD Check -list -C ivilian V ersion o M ultiv ar ia te: log istic r eg res -sion analy sis Sev er ity of ph ysical and se xual IPV in y ear pr ior t o baseline (+), vic tim ’s PT SD numb -ing sympt oms (+) M ears , Car lson,

Holden, & Har

ris (2001) 336 c ouples of which the f emale vic tim rec en

tly had obtained a

PO against the in

tima

te

par

tner and/or of which

the abusiv e par tner w as ar rest ed r ec en tly (T ra vis C oun ty , USA ). O f these w omen, 34% w as Hispanic , 34% W hit e, 32% Black . P rimar ily lo w -inc

ome and medi

-um-inc ome w omen 2 y ears Polic e and cour t r ec or ds 23% An y ph ysical abuse per petr at ed b y male half of the c ouple

which has been r

e-por ted t o the polic e, measur ed b y C TS Ph ysical IPV : inf or ma tion der iv ed fr om polic e and cour t r ec or ds M ultiv ar ia te: Co x r eg ression analy sis Number of pr ior ph

ysical IPV vic

timiza

(30)

29 Syst ema tic r eview of vic tims ’ influenc e on IPV r evic timiza tion Chapt er 2 M iller & K rull (1997) Reanaly sis of thr ee domestic violenc e

studies including cases in which a domestic violenc

(31)

ela tion with revic timiza tion of IPV (+ = positiv e, - = nega tiv e and 0 = no rela tionship ) Per ez & Johnson (2008) 320 w omen seek ing health car e a t medical sit es in inner -cit y Chi -cago (USA ) who had been vic tims of IPV in the past y ear ( CWHRS). O f these w omen, 68% w as A frican A mer ican, 23% H ispanic , 8% Eur opean. P rimar ily lo w -inc ome w omen 10 mon ths (range 3-23 mon ths) Vic tim in ter -view 50% Sev er ity of ph ysi -cal , se xual , and/or psy cholog ical abuse per petr at ed b y an y par tner (or ig inal

and/or new), meas

-ur ed b y the C amp -bell I nciden t S ev er ity Scale Ph ysical IPV : C ampbell Inciden t S ev er ity S cale , PT SD: PSS p, depr essiv e sympt oms: 4 it ems from M edical O ut come Study q, social suppor t: Social Suppor t Net w or k Scale fr om the C WHRS M ultiv ar ia te: hier ar chical linear r eg res -sion analy sis Ph

ysical IPV sev

er ity in y ear pr ior t o base -line (+), vic tim ’s PT SD (+), vic tim ’s depr essiv e sympt oms (0), vic tim ’s social suppor t (-) Sonis (2008) 321 w omen seek ing health car e a t medical sit es in inner -cit y Chi -cago (USA ) who had been vic tims of IPV in the past y ear ( CWHRS). O f these w omen, 68% w as A frican A mer ican, 23% H ispanic , 8% other . P rimar ily lo w -inc ome w omen 3-23 mon ths (85% of sample follo w -up af ter 4-15 mon ths) Vic tim in ter -view 50% An y ph ysical , se xual , and/or psy cholog ical abuse per petr at ed b y an y par tner (or ig inal

and/or new), meas

(32)

Sonis & Langer (2008)

321 w omen seek ing health car e a t medical sit es in inner -cit y Chi -cago (USA ) who had been vic tims of IPV in the past y ear ( CWHRS). O f these w omen, 68% w as A frican A mer ican, 23% H ispanic , 8% other . P rimar ily lo w -inc ome w omen 3-23 mon ths (85% of sample follo w -up af ter 4-15 mon ths) Vic tim in ter -view 50.5% An y ph ysical abuse per petr at ed b y an y par tner (or ig inal

and/or new), meas

-ur ed b y C TS Sev er ity of ph ysical abuse per petr at ed by an y par tner (or ig

ed b y the Campbell I nciden t Sev er ity S cale Ph ysical IPV : C TS, emo -tional abuse: P ow er and Con trol Scale s, PT SD: PSS, depr essiv e sympt oms: four of fiv e it ems fr om M en tal Health I nv en tor y-5 t, social suppor t: f our

item scale dev

eloped f or CWHRS M ultiv ar ia te: log istic r eg res -sion analy sis Fac tors f or an y and sev er ity of IPV r evic -timiza tion: Fr equenc y of ph ysical IPV in y ear pr ior t o baseline in ter view (+ for an y; 0 f or sev er -ity), par tner ’s use of po w er and c on trol tac tics (emotional abuse) (+ f or an y; 0 for sev er ity), vic tim ’s PT SD (0), vic tim ’s depr essiv e sympt oms (0), vic tim ’s social suppor t (0) W eisz,

Tolman, & Saunders (2000)

177 f

emale vic

tims of

IPV whose per

petr at or w as r ec en tly char ged

with IPV (USA

). O f these w omen, 71.6% w as E ur opean A mer i-can, 27.7% A frican Amer ican, 0.6% A sian, 0.6% missing . 58.5% w as full-time em -plo yed , 13.6% w as par t-time emplo yed , 27.7% not emplo yed . 4 mon ths Vic tim in ter -view Not r epor ted An y & sev er ity of sev er e ph ysical

abuse and/or sev

er e psy cholog ical abuse per petr at ed b y inde x par tner , measur ed by C TS Ph ysical abuse: C TS, psy cholog ical abuse: CT S, vic tim ’s assessmen t:

one question about likelihood tha

t par tner will bec ome violen t in ne xt y ear M ultiv ar ia te: log istic and linear r eg res -sion analy sis Fac tors f or both an y and sev er ity of IPV revic timiza tion: Pr ior ph ysical abuse (0), pr ior psy cholog

i-cal abuse (0), vic

(33)

Not

e.

The number in the

“Na tur e of sample ” c olumn r ef ers t o the number of r esponden ts a t f ollo w

-up on which the r

esults r

egar

ding risk and pr

ot ec tiv e fac tors f or r evic timiza tion of IPV w er e based . P er cen tages c onc erning r ac e, inc ome , and emplo ymen t ar e based on the f ollo w -up sample as w ell

, unless specified other

wise

. I

n the c

olumn

“D

efinition and measur

emen t of r evic timiza tion of IPV a t f ollo w -up ” w e sometimes r ef er t o the inde x par tner . T

his is the par

tner

or e

x-par

tner who perpetr

at ed the violenc e r epor ted a t baseline / study enr ollmen t. In the c olumn “Repor ted aspec ts of k ey fac tors fr om F

oa´s models and their r

ela tion with r evic timiza tion of IPV ”, v

ariables with (+) hold a positiv

e, sig nifican t r ela tionship ( p < .05),

variables with (-) a nega

tiv e, sig nifican t r ela tionship ( p < .05), and v

ariables with (0) no sig

nifican t r ela tionship with r evic timiza tion of IPV . CES -D = C en ter f or Epidemiolog ic S tudies-D epr ession CT S = C onflic t T ac tics S cale CWHRS = Chicago W omen ’s Health R isk S tudy IPV = in tima te par tner violenc e PMWI = P sy cholog ical M altr ea tmen t of W omen I nv en tor y PO = pr ot ec tiv e or der PSS = PT SD S ympt om S cale PT SD = posttr auma tic str ess disor der USA = Unit ed S ta tes of America a F ollo w -up time r epr esen ts the amoun t of time bet w een measur emen t of the aspec ts of the thr ee k ey fac tors of F oa

’s models and measur

emen t of r evic timiza tion of IPV b T he OD A method iden

tifies the model tha

t uses the pr

edic

tor sc

or

e in a manner such tha

t it discrimina tes those r evic timiz ed fr om those not r evic timiz ed with optimal ac cur ac y. c T his per cen tage c onc

erns the amoun

t of r

evic

timiza

tion of IPV during f

ollo w -up period b y a new par tner d Robins , Helz er , Cr oughan, & R at cliff , 1981 e Sheehan et al ., 1997 f M cL ellan, L uborsk y, O ’Brien, & W oody , 1980 g Gurley , 1990 h Alc ohol abuse w as sig nifican

tly and positiv

ely r ela ted t o futur e r evic timiza tion a t p < .05. Ho w ev er , in r epor ting their r esults C ole et al . (2008) adjust ed the p-value t o .01. i R adloff , 1977 j B ott oms , M ar tier , & S ok ol , 1989 k C ohen, M ermelst ein, K amar ck , & Hoberman, 1985 l This study t est ed sev erit y of past par tner violenc e as w ell; ho w ev er this w as a br oad c omposit e measur e cr ea ted b

y the authors including ph

ysical , psy cholog ical and se xual abuse ,

injuries and stalk

ing . T his br oad c omposit e measur e of sev erit y of past par tner violenc e sho w ed no sig nifican t main eff ec t on r evic timiza

tion of IPV in multiv

aria te analy sis . m Exac t f ollo w -up time w as not r epor

ted; baseline vic

tim in ter view s t ook plac e shor tly af

ter the IPV inciden

t f

or which vic

tims called the polic

e, the f ollo w -up vic tim in ter view w as

scheduled six mon

ths af

ter the IPV inciden

t tha t r esult ed in a polic e call . n Although polic e r ec or ds w er e used as sour ce of f ollo w -up inf orma tion as w ell , the analy ses which pr oduc

ed the risk fac

tors r

epor

ted in this table only used r

evic timiza tion r epor ts obtained fr om vic tim in ter view s. o W ea thers

, Litz, Herman, Husk

a, & Keane , 1993 p F oa, R iggs , Dancu , & Rothbaum, 1993 q Ha ys , Sherbourne , & M az el , 1995 r Although PT SD sho w ed a moder at e unadjust ed associa tion with r evic timiza tion of an y IPV , af ter c on tr olling f or impor tan t c onf ounding fac

tors (such as sev

erit

y of pr

evious abuse and

(34)

protection order court and criminal court. The majority of the studies we included in this review were conducted among samples consisting of only women (13 studies), one study included couples involved in IPV2_{(Mears et al., 2001) and one study (Miller & Krull, 1997)}

included both male and female victims in their sample.

Follow-up time

The length of follow-up time varies from three months (Bennett Cattaneo & Goodman, 2003) to two years (Fleury, Sullivan, & Bybee, 2000; Mears et al., 2001). Most of the studies used a follow-up period with a maximum of one year (10 of 15 studies).

Source of follow-up information

To measure revictimization, three different sources of follow-up information were used in the studies included in our systematic review. The vast majority of studies retrieved their information directly from the victim through a victim interview (14 studies). One study used police and court records as their source of follow-up information (Mears et al., 2001).

Amount of revictimization of IPV during follow-up

An important difference between studies included in our systematic review is the amount of revictimization during the follow-up period. This amount ranged from 17.2% (Crandall et al., 2004) to 74.1% (Colorado Springs sample of Miller & Krull, 1997). This might have been influenced by various factors such as the source of follow-up data (police and court records versus victim interviews), the length of the follow-up time (ranging from three months to two years) and to what extent researchers were able to retain all the respondents in the study.

Definition of revictimization of IPV

The way in which revictimization of IPV was defined differs strongly across the 15 studies. First, a difference can be identified in the variety of behaviors that have been categorized under IPV. Some studies only include physical partner violence, whereas others use a broader definition also including psychological violence, sexual violence, and / or IPV-related injury (e.g., Cole, Logan, & Shannon, 2008; Crandall, Nathens, Kernic, Holt, &

(35)

Rivara, 2004). A second dimension on which the definition of IPV revictimization varies across studies is whether IPV was conceptualized as a dichotomous (i.e., any IPV) or continuous (i.e., severity of IPV) variable. However, the majority of studies defined their outcome variable as any IPV (eight studies). Thirdly, IPV revictimization is operationalized differently in terms of the perpetrator of the violence. Some studies consider all subsequent IPV victimization by any partner (original and / or new partner, e.g., Bybee and Sullivan, 2002, 2005). Other studies measure IPV revictimization by one specific partner, such as IPV revictimization by the index partner (e.g., Krause, Kaltman, Goodman, & Dutton, 2006). Here, the focus is on the abusive partner described at baseline and whether this partner perpetrated IPV again during follow-up against the same victim. Furthermore, another study measured any physical, sexual or psychological abuse perpetrated by a new partner (Cole et al., 2008).

Measurement of revictimization of IPV

The instruments for measuring IPV as an outcome variable varied across the 15 studies as well. Some studies based their measure of IPV on only one question. For instance, Crandall and colleagues (2004) simply asked whether victims had experienced any injuries due to repeat IPV (Crandall et al., 2004). However, most studies used standardized questionnaires to obtain information about the presence and severity of IPV revictimization. The majority used the Conflict Tactics Scale (CTS; Straus, 1979; Straus & Douglas, 2004; Straus, Hamby, Boney-McCoy, & Sugarman, 1996) or a modified version of it. Other instruments included the Psychological Maltreatment of Women Inventory (PMWI; Tolman, 1989) and the Campbell Incident Severity Scale (Campbell, 1986).

Measurement of aspects of key factors

Similar aspects of Foa et al.’s (2000) key factors were often measured differently, for example in case of depression which was measured in four studies by four different instruments. In contrast, prior abuse (especially prior physical abuse) was quite consistently measured with the CTS. Furthermore, in some studies aspects of key factors were operationalized as binary variables (e.g., depression absent or present), whereas in other studies they were operationalized as continuous variables (e.g., number of depressive symptoms).

Statistical analyses

For all studies in our review, multivariate analyses were conducted except for one (Bennett Cattaneo & Goodman, 2003). Multivariate analyses report on the effect of a particular variable on the outcome while controlling for the effects of other variables in the model,

(36)

whereas bivariate analyses examine the effect of one variable on the outcome without taking into account the effects of other variables. Logistic regression analysis was the multivariate statistical technique that was used the most among the 15 studies, followed by linear regression analysis.

RESULTS

Construct of partner violence

Starting with synthesizing the evidence on the prospective relation between prior partner violence and revictimization, the construct of partner violence consists of three aspects: 1) physical and emotional abuse; 2) perceived threat; and 3) psychological reactions to the abuse. Table 2.2 summarizes the results from this review of the evidence on the link between partner violence and revictimization.

Physical and emotional abuse

A fair number of studies included prior physical abuse as predictor and the majority of these studies show a significant and positive relation between prior physical abuse and revictimization of IPV (see Table 2.2). Several studies used different measures of prior abuse, including a count of the number of prior victimizations, frequency of prior abuse, and measures of severity of prior abuse. This relationship was furthermore reported in different samples of victims, ranging from victims with a police- or court-reported case (Bennett Cattaneo & Goodman, 2003; Crandall et al., 2004; Hirschel & Hutchison, 2003; Miller & Krull, 1997), to victims seeking help in shelters (Bybee & Sullivan, 2005), mixed samples (Krause et al., 2006) and victims selected in a population sample of hospital patients (Perez & Johnson, 2008). Although Bennett Cattaneo and Goodman (2003) only used bivariate analysis to show a significant relationship between physical abuse and IPV revictimization, there were other studies that showed this relationship in multivariate analyses as well. The majority of the studies used the Conflict Tactics Scale (CTS; Straus, 1979; Straus & Douglas, 2004; Straus et al., 1996) to assess prior abuse, but other measures such as the Campbell Incident Severity Scale (Perez & Johnson, 2008) and items from National Institute of Justice-funded Spouse Assault Replication Projects (Hirschel & Hutchison, 2003) were also used. These differences in definitions of prior abuse, sample selection, and measures to assess prior abuse result in a convincing substantiation of the predictive effect of prior physical abuse. Nevertheless, there is also a number of studies that report no significant effect of prior physical abuse (Fleury et al., 2000; Mears et al., 2001;

(37)

Weisz, Tolman, & Saunders, 2000) or significance for specific aspects of prior abuse but not for others (Sonis & Langer, 2008). However, Fleury et al. (2000) restricted their analyses to women who left their abusive partners and only included revictimization by that same man. Furthermore, whereas all studies that reported a significant relation between prior

Table 2.2 Evidence for the relation between Foa et al.’s construct of partner violence and revictimization of IPV

Note. IPV = intimate partner violence.

a_{The article of Miller and Krull (1997) reported results from three domestic violence studies, based on three distinct samples}

from Milwaukee, Colorado Springs and Omaha. We report aspects of Foa et al.’s key factors for each of these samples.

b_{Bennett Cattaneo and Goodman (2003) concluded that psychological dominance-isolation, and not emotional-verbal}

abuse, was a significant predictor for revictimization of IPV.

Construct Study

Positive(+), negative (-) or no (0) relationship to revictimization of IPV

Partner violence

Physical and emotional abuse

Physical abuse Bennett Cattaneo & Goodman

(2003)

+

Bybee & Sullivan (2005) +

Crandall et al. (2004) +

Fleury et al. (2000) 0 (frequency of prior violence)

Hirschel & Hutchison (2003) + (frequency of prior violence)

Krause et al. (2006) + (severity of prior violence)

Mears et al. (2001) 0

Miller & Krull (1997)a

(Colorado Springs sample)

+ (IPV related injury)

Perez & Johnson (2008) + (severity of prior violence)

Sonis & Langer (2008) 0 (severity of prior violence)

+ (frequency of prior violence)

Weisz et al. (2000) 0

Emotional abuse Bennett Cattaneo & Goodman

(2003)

+ (psychological

dominance-isolationb₎

Crandall et al. (2004) 0

Fleury et al. (2000) + (threats)

Sonis & Langer (2008) + (power & control tactics)

Weisz et al. (2000) 0

Perceived threat Bennett Cattaneo & Goodman

(2003)

+

Hirschel & Hutchison (2003) + (desire to have perpetrator arrested)

Weisz et al. (2000) +

Psychological reactions to the abuse