Violence Against Women

3 downloads 0 Views 317KB Size Report
Mar 12, 2012 - ... violence advisor services (Study commissioned by the Hestia Fund and funded by the Henry Smith Charity and the Sigrid Rausing Trust).
Violence Against Women http://vaw.sagepub.com/

Women's Perceptions of Safety and Risk Following Police Intervention for Intimate Partner Violence Melissa E. Dichter and Richard J. Gelles Violence Against Women 2012 18: 44 DOI: 10.1177/1077801212437016 The online version of this article can be found at: http://vaw.sagepub.com/content/18/1/44

Published by: http://www.sagepublications.com

Additional services and information for Violence Against Women can be found at: Email Alerts: http://vaw.sagepub.com/cgi/alerts Subscriptions: http://vaw.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Mar 12, 2012 What is This?

Downloaded from vaw.sagepub.com by guest on October 12, 2013

437016 6Dichter and GellesViolence Against Women 2012© The Author(s) 2011

VAWXXX10.1177/107780121243701

Reprints and permission: http://www. sagepub.com/journalsPermissions.nav

Article

Women’s Perceptions of Safety and Risk Following Police Intervention for Intimate Partner Violence

Violence Against Women 18(1) 44­–63 © The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/1077801212437016 http://vaw.sagepub.com

Melissa E. Dichter1 and Richard J. Gelles2

Abstract Police intervention is a primary response to intimate partner violence (IPV) but does not guarantee a victim’s future safety. This study sought to identify factors associated with IPV survivors’ perceptions of safety and risk of revictimization following police intervention. One hundred sixty-four women completed a questionnaire, and 11 of those women also took part in qualitative interviews. The findings revealed that feeling unsafe and perceiving oneself to be at risk of future violence is associated with experiencing particular forms of IPV, including battering, lethality threats, and sexual violence. Having support from others and distance from the partner helps women feel safe. Keywords battering, intimate partner violence, revictimization

Background and Significance As with other forms of violence, intimate partner violence (IPV) is a criminal act in the United States, and the primary societal response is arrest and prosecution. Criminal system intervention, however, does not guarantee a victim’s future safety. Unlike other violent crimes, IPV victimization is typically not a one-time event and, in many cases, a victim experiences multiple incidents of victimization from the same “offender” over time, sometimes even after an offender has been arrested (Cattaneo & Goodman, 2005; Miller, 2003; Sonis & Langer, 2008; Tjaden & Thoennes, 2000; Ventura & Davis, 2005). Identifying

1

Philadelphia VA Medical Center, Philadelphia, PA, USA University of Pennsylvania, Philadelphia, PA, USA

2

Corresponding Author: Melissa E. Dichter, Center for Health Equity Research & Promotion, Philadelphia VA Medical Center, 3900 Woodland Avenue, Philadelphia, PA 19104, USA Email: [email protected]

Dichter and Gelles

45

individuals who are at risk of revictimization and understanding victims’ perceptions of risk of future violence may be important for informing interventions to protect future safety. Numerous studies (e.g., Cattaneo, Bell, Goodman, & Dutton, 2007; Cattaneo & Goodman, 2005; Hilton et al., 2004; Weisz, Tolman, & Saunders, 2000) have found that female IPV victims’ assessment of risk is associated with actual subsequent victimization. Individual risk perception is important not only because it may be an accurate predictor of future victimization but also because perception of risk is associated with taking action toward protective or risk-avoidant behaviors (Brewer et al., 2007). Although individuals are certainly not responsible for their own victimization, those who perceive themselves to be in danger may be able to take steps to decrease their risk of violence victimization if they are afforded adequate resources to do so. Prior literature has identified some risk factors for revictimization. However, we know little about what factors are associated with women’s perceptions of risk, nor do we have much data about the differences between women who feel at risk and those who do not. The purpose of this study is to identify the factors associated with IPV survivors’ perceptions of safety and risk of future violence following police intervention for IPV.

Revictimization Risk Factors In previous studies, women who reported that their partners controlled their freedom and independence, such as monitoring and restricting their activities, and who were jealous or suspicious about the women’s relationships, were more likely to experience ongoing repeat victimization than women who reported that their partners did not engage in such behaviors (Cattaneo & Goodman, 2005; Robinson, 2007; Sonis & Langer, 2008). Robinson (2007) also found an association between women feeling “very frightened” of their partners’ violence and experience of repeat victimization. Having access to resources that facilitate independence and freedom, such as money, employment, and social support, however, is associated with less repeated violence (Bybee & Sullivan, 2005; Goodman, Dutton, Vankos, & Weinfurt, 2005; Perez & Johnson, 2008). Findings related to separation or relationship termination are mixed. For example, Sonis and Langer (2008) found that women who had left an abusive partner were at decreased risk of future violence, whereas Robinson (2007) found that relationship separation was associated with an increase in repeat violence. Leaving the relationship does not guarantee safety and, in fact, may increase risk of further violence. For some women, the point of highest risk of violence is when they make known to their partners the decision to leave the relationship (Fleury, Sullivan, & Bybee, 2000) and assaults may increase in severity after leaving (Anderson & Saunders, 2003). Women who are victims of intimate partner homicide are more likely to have left the relationship and to be separated from the partner than to be still in the relationship at the time of their death (Anderson & Saunders, 2003; Dobash, Dobash, Cavanagh, & Medina-Ariza, 2007; Wilson & Daly, 1993). Leaving threatens the partner’s dominance, which may then lead to violence in an attempt to regain power, among those perpetrators for whom violence is related to an attempt to dominate and control the partner (Campbell, 1992; Mahoney, 1991, 1994).

46

Violence Against Women 18(1)

Taken together, prior studies on risks for revictimization suggest that revictimization is associated with a particular form of IPV, referred to as battering. Battering is a pattern of violence used to gain coercive control over the victim; through violence and threats of violence, the batterer establishes dominance in the relationship and decreases the victim’s independence. Not all violence used against partners is battering. Violence used in anger, frustration, retaliation, or self-defense; that does not have coercive, fear-inducing, and threatening power over the victim, is nonbattering IPV (Dutton & Goodman, 2005; Smith, Thornton, DeVellis, Earp, & Coker, 2002; Stark, 2007). Battering is measured not by individual assault incidents, but by the dynamic of the violence in the relationship—by measuring behaviors reflective of control and dominance, a perpetrator’s motive to control the victim, and/or the victim’s experience of being dominated and controlled through threats of violence.

Reporting to Police Approximately one in four cases of physical IPV against women, and one in four cases of intimate partner rape, are reported to the police, either by the female victims themselves or by others (e.g., bystanders, children, doctors; Tjaden & Thoennes, 2000). The cases that reach police tend to be more severe than those that do not (Akers & Kaukinen, 2009; Bonomi, Holt, Martin, & Thompson, 2006) and, for many of these cases, involving the police has a deterrent effect on future violence (Felson, Ackerman, & Gallagher, 2005). Police intervention may be necessary, but not sufficient, for arrest. When police intervene, they may or may not make an arrest; if they do make an arrest, they may arrest either partner, or both partners, depending on the evidence suggestive of criminal activity. Recent studies have found little to no relationship between arrest of the male partner and reassault (e.g., Felson et al., 2005; Hilton, Harris, & Rice, 2007). This study focused on women who had experienced police intervention because they are likely to have endured severe and/or ongoing violence and because the police contact presents an opportunity to connect victims to a range of health and social services.

Method This study used a nonexperimental, cross-sectional design, employing both quantitative and qualitative data collection and analysis methods. A self-report questionnaire was the primary method of data collection; a subset of women who completed the questionnaire also took part in an in-depth qualitative interview.

Setting and Sample Data collection for the study took place in Philadelphia, PA, a large city on the east coast of the United States, between September 2006 and June 2008. Participants included English-speaking adult (18 to 64) women who had experienced police response to an incident of violence between themselves and a current or former male intimate partner.

Dichter and Gelles

47

Recruitment and Data Collection Participants were recruited from a hospital emergency department (ED) as well as four organizations in Philadelphia that provide counseling and support to women who have experienced domestic violence. In the hospital, research assistants screened female adult patients for IPV and eligibility for this study (women having had the police called to respond to an incident of fighting or violence between themselves and a male partner). Patients who revealed a police call were then invited to participate in the study by completing the questionnaire during their ED visit. Participants who were interested in participating in an in-depth interview were referred to the principal investigator (first author). At the organizations, flyers and counselors informed clients about the study. Potential participants then arranged individual meetings with the study investigator. Data collection took place in a private space at the hospital or organization from which the participant was recruited. Participants were provided with the option of reading and completing the questionnaire themselves or having the questions read to them. The principal investigator conducted all of the in-depth interviews, using a semistructured interview guide that asked the participant to elaborate on themes raised in the questionnaire through open-ended questions (e.g., “Can you tell me more about that?”). All participants provided oral informed consent before beginning the questionnaire or interview. To protect participant safety and confidentiality, we did not solicit or record any individual identifying information. Interviews were audio-recorded and transcribed verbatim.

Key Variables and Measures Relationship Factors. Participants were asked to provide responses about a relationship with a current or former male intimate partner in which the police were called to respond to an incident of IPV. Relationship variables included status of the relationship (current/ former), cohabitation, and whether the couple had children in common. Participants were also asked who in the couple, if anyone, was arrested as a result of police intervention.

Violence Victimization and Perpetration. Violence in the relationship was measured in multiple ways. Physical and sexual assaults by either partner against the other were measured with questions from the short form of the Revised Conflict Tactics scales (CTS2S; Straus & Douglas, 2004) that ask about specific behaviors (e.g., pushing, shoving, slapping, punching, insisting on or forcing sex). Three variables that have been identified as highly associated with lethality risk— threats to kill, attempts to choke, and threats with a lethal weapon—were measured using the corresponding items from Campbell’s (1995) Danger Assessment (DA) scale. Participants were asked if their partners had ever used any of these behaviors against them to yield a dichotomous measure (yes/no). The respondent’s subjective experience of battering—entrapment and disempowerment resulting from her partner’s behavior—was measured with the Women’s Experience with Battering Scale (WEB; Smith, Earp, & DeVellis, 1995). The WEB has 10 statements

48

Violence Against Women 18(1)

reflective of the experience women have when battered (e.g., “I feel owned and controlled by him”; “I feel like he keeps me prisoner”; “He can scare me without laying a hand on me”). Based on recommendations from participants in a pilot test, we used a dichotomous scoring system (agree/disagree) in which agreement with two or more of the ten items indicates battering. The dichotomous scoring system corresponds well to the original version and has strong sensitivity (79.8%) and specificity (99.4%; Coker et al. 2007).

Resources and Social Support. Employment was used as a proxy of resource availability: access to income. Women were asked whether they were currently (at the time of data collection) employed in full- or part-time work for which they earned an income. Social support is a personal characteristic based on interpersonal relationships. Women’s social support was measured with two questions: Do you have someone you can . . . (a) stay with in an emergency? (b) borrow money from if needed? These questions were taken from the Social Support Network Scale developed as part of the Chicago Women’s Health Risk Study (Block, 2000) and scored dichotomously (yes/no).

Perceptions of Safety and Risk. Safety and risk are measured based on the women’s own perceptions with two close-ended questions: (a) How safe do you feel in terms of this relationship? That is, how safe do you feel around your partner/ex-partner currently? and (b) How likely do you think it is that your partner/ex-partner will use violence against you in the future? Participants responded on a 4-point Likert-type scale (very safe to very unsafe or very unlikely to very likely) and responses were collapsed to a dichotomous measure (safe/unlikely vs. unsafe/likely) to facilitate analysis. The questions were asked in reference to the partner or ex-partner with whom the women had the relationship in which the police were called. The two concepts—safety and risk—are related to, but distinct from, one another. Sense of safety is an emotion, possibly based on past trauma even in the absence of future risk. Risk perception, however, is a cognitive assessment that may or may not be related to feelings of safety.

Data Analysis Quantitative Data. SPSS 16.0 was used to conduct statistical analyses of quantitative data. Crosstabs, along with the Pearson chi-square test of statistical significance, were used to measure association between each of the key independent variables and each of the outcome variables (feeling unsafe and feeling at risk). Binary logistic regression models were used to identify odds of each of the outcomes (feeling unsafe and feeling at risk), controlling for all other key variables in the model. To optimize model efficiency, only key variables that produced chi-square p values of