A Meta-Analysis of Risk Markers for Intimate Partner Violence in Same-Sex Relationships

Abstract

Research on intimate partner violence (IPV) has largely focused on heterosexual relationships, but, in recent years, researchers have expanded their focus to include same-sex relationships. Using meta-analytic techniques, this study was conducted to examine the relative strength of various risk markers for men and women being perpetrators and victims of physical IPV in same-sex relationships. Articles were identified through research search engines and screened to identify articles fitting the inclusion criteria, a process that resulted in 24 studies and 114 effect sizes for the meta-analysis. The strongest risk marker among those with at least two effect sizes for both male and female perpetration was psychological abuse perpetration. The strongest risk marker among those with at least two effect sizes for IPV victimization was also perpetration of psychological abuse for males and psychological abuse victimization for females. Among same-sex-specific risk markers, internalized homophobia and fusion were the strongest predictors for being perpetrators of IPV for men and women, respectively. HIV status and internalized homophobia were the strongest risk markers for IPV victimization for men and women, respectively. Of 10 comparisons between men and women in risk markers for IPV perpetration and victimization, only 1 significant difference was found. The results suggest that although same-sex and heterosexual relationships may share a number of risk markers for IPV, there are risk markers for physical IPV unique to same-sex relationships. Further research and increased specificity in measurement are needed to better study and understand the influence of same-sex-specific risk markers for IPV.

Keywords

intimate partner violence same-sex relationships risk markers meta-analysis

Efforts to address the issue of intimate partner violence (IPV) have primarily targeted heterosexual relationships and largely neglected same-sex relationships (Baker, Buick, Kim, Moniz, & Nava, 2013). As Carvalho, Lewis, Derlega, Winstead, and Viggiano (2011) noted, “domestic violence programs and shelters are often unprepared to deal with victims of same-sex IPV” (p. 502). A growing number of researchers have responded to the imperative to conduct research on IPV in same-sex relationships, and a meta-analysis of research on risk markers for IPV in same-sex relationships is necessary to inform subsequent research. Because IPV is very complex with many types of violence (i.e., physical, psychological, sexual) that have differing prevalence rates (Black et al., 2011) and consequences (Dutton, Goodman, & Bennett, 2001; Pico-Alfonso, 2005), we chose to focus solely on physical IPV in order to gain a clearer understanding of one type of IPV in same-sex relationships. Furthermore, in this article, we use the term, IPV as synonymous with physical IPV, even though there are different types of IPV. Ultimately, a meta-analysis of the research on risk markers for physical IPV perpetration and victimization in same-sex relationships may facilitate the development and implementation of prevention and treatment programs for physical IPV in same-sex relationships.

Despite a recent influx of research focused on examining risk markers for physical IPV in same-sex relationships, it has not yet been unified through a meta-analytic review, making it difficult to assess the progress, or lack thereof, researchers have made in terms of this collaborative enterprise. In this study, we sought to use a meta-analysis to draw actionable conclusions from the incipient area of research on physical IPV for both males and females in same-sex relationships and to identify deficiencies in the research that can be targeted in future studies.

Existing Meta-Analytic Research on Risk Markers for IPV

Most previous meta-analytic work has addressed risk markers for IPV in the general population and has not specifically addressed IPV in same-sex relationships. For example, Stith, Smith, Penn, Ward, and Tritt (2004) included evidence from 85 studies to identify risk markers most strongly related to IPV perpetration and victimization. They were able to calculate effect sizes for 16 perpetration risk markers and 9 victimization risk markers. The strongest effect sizes were found between perpetration and the following five risk markers: psychological abuse, forced sex, illicit drug use, attitudes condoning marital violence, and marital satisfaction. More recently, Stith et al. (2014) were able to identify 80 risk markers for IPV in the general population. The strongest risk markers for perpetration were prior victimization, stalking, prior psychological abuse, prior perpetration, borderline personality disorder, anger, and approval of violence. The strongest risk markers for IPV victimization were prior perpetration, prior victimization, psychological abuse, and perpetrators’ use of power and control. Although we are beginning to learn more about risk markers for IPV, in general, research on risk markers for IPV in same-sex relationships is much more limited.

Existing Research on Risk Markers for IPV in Same-Sex Relationships

In a review of the literature of the risk markers for all types of IPV in same-sex relationships, Edwards, Sylaska, and Neal (2015) concluded that increased risk for IPV victimization was associated with a number of variables, including witnessing IPV as a child, victimization in peer networks, and physical and psychological health problems. Although these findings contribute to the literature related to same-sex IPV, it is important to note that the researchers did not make a distinction between risk markers for the different kinds of IPV; rather, studies were incorporated into the review that examined the relationship between various potential risk markers and physical, psychological, or sexual IPV. Furthermore, the investigation did not clearly distinguish between risk markers for males and females, so it is uncertain whether risk markers differ between male and female same-sex relationships. In the current meta-analysis, therefore, we sought to expand upon Edwards and colleagues’ (2015) work by exploring whether or not the risk markers are moderated by gender.

In a recent meta-analytic study, Buller, Devries, Howard, and Bacchus (2014) explored risk markers for IPV in men who have sex with men. Nineteen studies met the researchers’ inclusion criteria. They found that being a victim of IPV in same-sex relationships was associated with substance use, being HIV positive, depressive symptoms, and engagement in unprotected anal sex. Perpetration of IPV, on the other hand, was only associated with substance use. However, like the investigation from Edwards and colleagues (2015), the researchers included studies that measured at least one form of IPV (i.e., physical, sexual, or psychological). Consequently, it is uncertain whether the risk markers would remain statistically significant if only studies that assessed physical IPV were included. Another limitation of this study is that articles examining IPV in female same-sex relationships were not included, so the degree to which the risk markers for IPV differ between male and female same-sex relationships could not be examined. Although there have been meta-analyses in which researchers examine the prevalence of IPV in female same-sex relationships (e.g., Badenes-Ribera, Frias-Navarro, Bonilla-Campos, Pons-Salvador, & Monterde-i-Bort, 2015), we were not able to find a meta-analysis in which risk markers for IPV in female same-sex relationships were explored.

Another limitation of previous literature reviews and meta-analyses has been the operationalization of same-sex relationships. Previous reviews include articles based on lesbian, gay, and bisexual (LGB) identity (e.g., Edwards, Sylaska, & Neal, 2015; Mason et al., 2014) or sexual behavior (e.g., Buller, Devries, Howard, & Bacchus, 2014). This can be problematic because participants in the included studies may not have been in a same-sex relationship—that is, they may have had sex with someone of the same-sex or identify as LGB but not have been in a same-sex relationship. This critique of studies on IPV in same-sex relationships has been detailed by previous researchers (see Baker et al., 2013). Thus, our meta-analysis adds to understanding risk markers for physical IPV in same-sex relationships by including studies that explicitly indicate that participants were reporting on IPV in a current same-sex relationship or in a specific previous same-sex relationship rather than by sexual behavior or sexual identity.

IPV in same-sex relationships is a relatively new research area, but finding threads that link existing studies using meta-analysis may help to draw general conclusions that can inform program development and implementation. In the present study, we use a meta-analysis in an attempt to synthesize findings from previous researchers. Although researchers have conducted meta-analyses in exploring the prevalence of same-sex IPV and risk markers for physical, sexual, or psychological IPV for men in same-sex relationships, this is, to our knowledge, the first meta-analysis of risk markers for physical IPV for both males and females in same-sex couples.

Sexual Minority Stressors and Same-Sex IPV

Same-sex and heterosexual couples may share a number of risk markers, but there may also be differences between these populations in terms of risk markers for IPV, which may be explained, in part, by the differences in the nature of gender socialization experienced by men and women (McClennen, 2005). That is, beyond the ostensible similarities between risk markers for IPV among same-sex couples and heterosexual couples, it is also important to consider that individuals in same-sex relationships may contend with additional risk markers. For example, “fusion” has been studied as a risk factor for IPV in female same-sex relationships. Fusion is a term that refers to the high level of closeness or lack of boundaries between partners that may lead to anxiety, tension, or a loss of individuality in a relationship. Although lack of differentiation has been studied as a risk marker for IPV in heterosexual relationships, fusion, which can result from lack of differentiation, has been frequently examined with regard to lesbian relationships (e.g., Causby, Lockhart, White, & Green, 1995; Milletich, Gummienny, Kelley, & D’Lima, 2014). Providing a conceptual framework for risk markers in same-sex relationships, Meyer (1995, 2013) posited that the higher prevalence of mental health disorders in sexual minorities relative to heterosexuals could be explained by various forms of minority stress, which include forms of stress due to stigma, prejudice, and discrimination within one’s social milieu.

Several sexual minority stressors have been identified within previous research. In this study, we examine a variety of types of minority stressors. “Internalized homophobia,” for example, is a minority stressor that refers to the internalization of negative societal views about the minority population of which one is a member (Newcomb & Mustanski, 2011). “Stigma consciousness” is a minority stressor used to describe the extent to which stigmatized groups expect to be stereotyped and experience discrimination (Carvalho, Lewis, Derlega, Winstead, & Viggiano, 2011). Another minority stressor is called “outness,” which refers to the degree to which one is open about their sexual identity with family, friends, and colleagues (Steele, Tinmouth, & Lu, 2006). We also examined the same-sex-specific risk marker called “homophobic controlling behaviors,” which is a term used to describe acts perpetrated by partners that internalize negative attitudes about homosexuality or threatening to reveal the partner’s sexual orientation to friends, family, and other individuals who are important in his or her life (Badenes-Ribera et al., 2015).

For individuals in same-sex relationships, minority stressors may increase the risk of perpetrating physical IPV, being a victim of physical IPV, or both. However, despite the recent profusion of research that has provided empirical support linking minority stressors to a number psychological outcomes including depressive symptoms (Hatzenbuehler, Nolen-Hoeksema, & Erickson, 2008), anxiety (Lehavot & Simoni, 2011), suicide ideation (Meyer, 1995), and loneliness (Szymanski, Chung, & Balsam, 2001), the association between minority stressors and IPV has received little attention from researchers. To integrate Meyer’s theory into prevention and intervention programs, more research is needed to establish an empirical relationship between minority stressors and IPV in same-sex relationships. We address this need by including same-sex-specific risk markers in our analysis to examine the strength of these risk markers.

Present Study

In this study, we endeavored to make progress toward three primary goals, including (1) examine the relative strength of various risk markers for IPV in same-sex relationships, (2) compare risk markers for IPV perpetration versus victimization for men and women in same-sex relationships, and (3) compare the strength of risk markers between men and women in same-sex relationships to explore possible gender differences. Results from this research can aid in the development of efforts for education, advocacy, and policy related to IPV in same-sex relationships. In this meta-analysis, we elected to only include studies which assessed the relationships between risk markers and physical IPV. Therefore, if the study did not distinguish between various types of abuse and the identified risk marker, it was not included in this article.

Method

We searched published and unpublished studies from 1980 to 2016 which were found in search engines (ProQuest Research Library, ProQuest Theses and Dissertations Global, PsycINFO, Social Services Abstracts, Sociological Abstracts) by using key terms related to couple (i.e., couple, intimate partner, romantic partner, dating partner, marital, spouse, husband, wife), sexual orientation (i.e., lesbian, same-sex, gay, men who have sex with men, women who have sex with women), and partner aggression (i.e., abuse, aggression, domestic violence, batter, maltreatment, violence, victim). We also screened the reference lists from four meta-analysis/reviews reporting about IPV in same-sex relationships (i.e., Badenes-Ribera et al., 2015; Buller et al., 2014; Burke & Follingstad, 1999; Edwards et al., 2015).

Studies were included in the present study if (a) the outcome measured perpetration or victimization of physical IPV in same-sex relationships, (b) the study included statistical information sufficient for calculating one (or more) bivariate effect size, (c) the study was published in English, and (d) the study had an adult sample which identified as being in current or previous same-sex adult romantic relationships. Studies were excluded if (a) they did not differentiate between different forms of IPV (e.g., physical, psychological, or sexual), (b) they did not differentiate between lesbian, gay, bisexual, transgender, and questioning (LGBTQ) samples, or (c) coders could not determine whether the risk markers were measured about a same-sex relationship—this usually manifested in studies that recruited participants only by their sexual identity or only as having sex with the partners of the same sex but not that the participants had actually been in a same-sex relationship where physical IPV was present. Furthermore, psychological abuse was only examined as a risk marker when the abuse occurred in the same relationship as the IPV.

A total of 690 studies (553 from search engines and 142 from reference lists) were screened, and 24 studies (114 effect sizes) met our criteria for inclusion (see Figure 1). Of these 114 effect sizes, 43 were for men in same-sex relationship, while 71 effect sizes were for women in same-sex relationships. We followed recommended meta-analysis coding procedures (Card, 2012; Hunter & Schmidt, 2004). A graduate research coding team used a 37-item code sheet to capture pertinent information from each included study. Each study was cross-coded, and if cross-coders could not resolve coding discrepancies, they met with a project leader to collectively arrive at a deeper understanding of the data (Hawkins, Blanchard, Baldwin, & Fawcett, 2008). Overall coding agreement was 99.2%. We used Comprehensive Meta-Analysis software version 2.0 (Borenstein, Hedges, Higgins, & Rothstein, 2005) to enter, calculate, and analyze effect sizes for physical IPV perpetration and victimization. The overall effect size for all same-sex-specific risk markers with at least two reported were effect sizes included. However, other risk markers were not included if they did not have enough effect sizes to run at least one comparative analysis, either within men and women groups or between men and women groups. In order to compare the strength of two groups, each group must have at least two effect sizes.

Figure 1.
Flow process in selecting primary studies. ^aPercentages add up to more than 100% because studies had multiple reasons for exclusion.

We theorized there would be real population differences between studies, so we used a random-effects model to aggregate mean effect sizes for each risk marker. The random-effects approach accounts for both between-study variance and within-study variance (fixed-effect only accounts for within-study variance), which allows for generalized inferences that can extend beyond the studies in this meta-analysis (Card, 2012). A random-effects model was also used to compare males and females as well as perpetration and victimization within men and women. Every meta-analysis suffers from the “file drawer problem” (Hunter & Schmidt, 2004). One way we addressed this was by including unpublished articles (i.e., theses and dissertations) which may have nonsignificant findings. We also made attempts to contact authors (n = 8) who reported usable data but did not separate types of IPV, sexual orientation, or separate the sample by relationship status with a 25% success rate (n = 2). Finally, we conducted three tests to evaluate the possible effects of publication bias: the trim-and-fill test (Duval & Tweedie, 2000), the fail-safe N (Rosenthal, 1979), and Orwin’s fail-safe N (Orwin, 1983). Finally, we interpreted the magnitude of mean effect sizes using Cohen’s (1992) suggested criteria of trivial (r < .01), small (r = .10), medium (r = .30), and large (r = .50).

Results

Same-Sex-Specific Risk Markers

The online supplementary material includes references for all studies in the final analysis, along with the characteristics of all of the effect sizes in the final analysis organized by study, gender, and risk marker. Consistent with previous reviews, we did not find many studies that explored same-sex-specific risk markers related to physical IPV perpetration or victimization. The same-sex-specific risk markers with at least two effect sizes were examined for men and women (see Table 1). For men, internalized homophobia (r = .23, p < .001) was found to be a significant same-sex-specific risk marker of perpetration of physical IPV, but this was not the case for outness and stigma consciousness. We also did not find evidence that HIV status or stigma consciousness was a significant risk marker for physical IPV victimization for men.

Table 1.
Same-Sex-Specific Risk Markers for Men’s Perpetration and Victimization of Physical IPV.

Risk Marker k Mean ES LL UL

Men

Physical IPV Perpetration

Internalized homophobia 2 .23* .11 .33

Outness 2 −.10 −.44 .26

Stigma consciousness 2 .02 −.07 .12

Physical IPV victimization

HIV status 2 .20 −.06 .44

Stigma consciousness 2 .03 −.07 .13

Women

Physical IPV perpetration

Fusion 3 .14 .06 .22

Internalized homophobia 3 .09* .00 .18

Physical IPV victimization

Internalized homophobia 4 .11 −.02 .24

Outness 2 .09 −.33 .48

Fusion 5 .01 −.25 .28

Note. Random effects models calculate an effect size using both sample error within and between studies. These are simple reports of the overall effect size for a risk marker in one gender and in one direction, so only within study sample error is used in the calculation; therefore, effects sizes from Table 1 may differ from those in Tables 2 –5. k = number of studies; mean ES = mean effect size; mean SE = mean standard error; LL = lower limit of confidence interval; UL = upper limit of confidence interval; IPV = intimate partner violence.

p < .05. p < .01. p < .001 (two-tailed).

For women in same-sex relationships, one of the two significant same-sex-specific risk markers for physical IPV perpetration was fusion (r = .14, p < .01). Internalized homophobia (r = .09, p < .05) was the other significant same-sex-specific risk marker for physical IPV perpetration. The three risk markers for physical IPV victimization, which included internalized homophobia, outness, and fusion, did not reach statistical significance.

Comparing Strengths of Risk Markers in Men’s Same-Sex Relationships

Tables 2 and 3 present results of moderator analyses with direction of violence (perpetration or victimization) with those variables (same-sex specific and nonspecific) when two or more effect sizes were available. There were no significant differences in strength of risk markers for male physical IPV perpetration and victimization in same-sex relationships (see Table 2). The strongest risk markers for IPV perpetration were “perpetrating psychological abuse” (r = .55, p < .001), “victim of psychological abuse” (r = .48, p < .001), alcohol abuse (r = .27, p < .001), experiencing child abuse in family of origin (FOO; r = .15, p < .001), and witnessing parental IPV (r = .09, p < .05). Stigma consciousness—the extent to which stigmatized groups expect to be stereotyped and experience discrimination—and education level were not significantly related to IPV perpetration. Perpetration of psychological abuse and being a victim of psychological abuse were large effect sizes, and alcohol abuse, experiencing child abuse in FOO, and witnessing parental IPV were small effect sizes for IPV perpetration among men in same-sex relationships.

Table 2.
Risk Markers for Men in Same-Sex Relationships.

Risk Marker k Mean ES LL UL Mean SE Direction k Mean ES LL UL Q^b

Alcohol abuse 5 .31* .23 .39 .01 Perpetration 3 .27* .14 .40 0.68

Victimization 2 .34* .23 .45

Child abuse in family of origin 8 .16* .08 .24 .01 Perpetration 4 .15* .03 .27 0.04

Victimization 4 .17* .06 .27

Education 4 .004 −.22 .23 .004 Perpetration 2 .04 −0.28 .36 0.10

Victimization 2 −.03 −.32 .27

Perpetrating psych abuse 5 .53* .47 .59 .01 Perpetration 2 .55* .41 .66 0.07

Victimization 3 .53* .46 .59

Stigma consciousness 4 .02 −.04 .09 .00 Perpetration 2 .02 −.07 .12 0.01

Victimization 2 .03 −.06 .12

Witnessing IPV 8 .13* .04 .22 .00 Perpetration 5 .09* .00 .17 2.10

Victimization 3 .18*** .09 .27

Note. k = number of studies; mean ES = mean effect size; mean SE = mean standard error; LL = lower limit of confidence interval; UL = upper limit of confidence interval; Q^b = Cochran’s Q test of heterogeneity between studies; IPV = intimate partner violence.

p < .05. p < .01. p < .001 (two-tailed).

Table 3.
Risk Markers for Women in Same-Sex Relationships.

Risk Marker k Mean ES LL UL Mean SE Direction k Mean ES LL UL Q^b

Alcohol abuse 14 .16* .11 .002 .01 Perpetration 8 .19* .13 .24 1.78

Victimization 6 .13* .08 .19

Anger 4 .23* .19 .28 .002 Perpetration 2 .26* .20 .31 1.10

Victimization 2 .21* .16 .27

Child abuse in family of origin 4 .15* .06 .25 .01 Perpetration 2 .16* .04 .26 0.01

Victimization 2 .15 −.03 .32

Fusion 8 .05 −.12 .22 .04 Perpetration 3 .11 −.17 .36 0.28

Victimization 5 .01 −.20 .23

Internalized homophobia 7 .11* .02 .19 .01 Perpetration 3 .11 −.02 .22 0.02

Victimization 4 .11 −.003 .22

Poor mental health^a 4 .13 −.19 .43 .07 Perpetration 2 −.05 −.39 .31 1.88

Victimization 2 .28 −.03 .54

Perpetrating psych abuse 6 .54* .46 .61 .01 Perpetration 3 .57* .49 .65 1.54

Victimization 3 .50* .40 .58

Self-esteem 7 −.04 −.21 .14 .03 Perpetration 4 .04 −.19 .27 1.03

Victimization 3 −.14 −.40 .13

Victim of psych abuse 10 .50* .44 .55 .003 Perpetration 4 .47* .39 .53 1.75

Victimization 6 .52* .47 .57

Note. k = number of studies; mean ES = mean effect size; mean SE = mean standard error; LL = lower limit of confidence interval; UL = upper limit of confidence interval; Q^b = Cochran’s Q test of heterogeneity between studies.

^aStudies for mental health were reporting on psychological problems and lower levels of mental health.

p < .05. p < .01. p < .001 (two-tailed).

The strongest risk marker for IPV victimization for men in same-sex relationships was perpetrating psychological abuse (r = .53, p < .001), followed by being a victim of psychological abuse (r = .49, p < .001), alcohol abuse (r = .34, p < .001), witnessing parental IPV (r = .18, p < .001), and experiencing child abuse in FOO (r = .17, p < .001). Stigma consciousness was not a significant risk marker for victimization. The effect sizes for being a victim of physical IPV were large for perpetrating and being a victim of psychological abuse, medium for alcohol abuse, and small for witnessing parental IPV and experiencing child abuse in FOO.

Comparing Strengths of Risk Markers in Women’s Same-Sex Relationships

There were no significant differences in risk markers for female IPV perpetration or victimization (see Table 3). The strongest risk marker for female perpetration in same-sex relationships was perpetrating psychological abuse (r = .57, p < .001), followed by being a victim of psychological abuse (r = .47, p < .001), anger (r = .26, p < .001), alcohol abuse (r = .19, p < .001), and child abuse in FOO (r = .16, p < .05). Overall, fusion, internalized homophobia, self-esteem, and poor mental health were not significant risk markers for female IPV perpetration. Perpetrating and being a victim of psychological violence were large effect sizes, and anger, alcohol abuse, and child abuse in FOO were small effect sizes for female IPV perpetration.

The strongest risk markers for physical IPV victimization for women in same-sex relationships was being a victim of psychological abuse (r = .52, p < .001), perpetrating psychological abuse (r = .50, p < .001), anger (r = .21, p < .001), and alcohol abuse (r = .13, p < .05). Child abuse in FOO, fusion, internalized homophobia, self-esteem, and poor mental health were not significant risk markers for IPV victimization for women in same-sex relationships. Perpetrating and being a victim of psychological abuse were large effect sizes and anger and alcohol abuse were small effect sizes.

Gender Differences

There were no significant differences in strength of risk markers for male versus female perpetration of physical IPV in same-sex relationships (see Table 4). There was a significant difference, Q (2) = 22.56, p < .001, for alcohol abuse as a risk marker for physical IPV victimization—alcohol abuse was a stronger risk marker for IPV victimization for men (r = .36, p < .001) than women (r = .14, p < .001; see Table 5).

Table 4.
Men Versus Women Perpetration in Same-Sex Relationships.

Risk Markers k Mean ES LL UL Mean SE Gender k Mean ES LL UL Q^b

Alcohol abuse 11 .21* .13 .28 .003 Women 8 .19* .06 .33 1.43

Men 3 .27* .02 .45

Child abuse in family of origin 6 .15* .08 .22 .01 Women 2 .16* .04 .26 0.01

Men 4 .15** .06 .24

Education 5 .15 −.06 .34 .04 Women 3 .24* .01 .44 1.40

Men 2 .02 −.25 .29

Internalized homophobia 5 .15* .02 .28 .01 Women 3 .09* .01 .18 2.88

Men 2 .23 .11 .36

Perpetrating psych abuse 5 .48* .43 .69 .01 Women 3 .57* .41 .56 0.11

Men 2 .55* .49 .65

Witnessing parental IPV 7 .09* .02 .17 .01 Women 2 .11 −.01 .24 0.11

Men 5 .09 −.0002 .17

Victim of psychological abuse 7 .47* .39 .54 .01 Women 4 .46* .36 .55 0.11

Men 3 .49*** .37 .59

Note. It is not unusual to see slight shifts in ES when looking at different moderators, but these shift are negligible when there are a large number of ES—this shift may more noticeable here due to the small number of ES. Thus, risk markers ES and significance levels from Tables 2 and 3 for perpetration may differ slightly in Table 4. k = number of studies; mean ES = mean effect size; mean SE = mean standard error; LL = lower limit of confidence interval; UL = upper limit of confidence interval; Q^b = Cochran’s Q test of heterogeneity between studies; IPV = intimate partner violence.

p < .05. p < .01. p < .001 (two-tailed).

Table 5.
Men Versus Women Victimization in Same-Sex Relationships.

Risk Marker k Mean ES LL UL Mean SE Gender k Mean ES LL UL Q^b

Alcohol abuse 8 .24* .03 .44 .01 Women 6 .14* .09 .18 22.56*

Men 2 .36* .28 .43

Child abuse in family of origin 6 .16* .05 .27 .01 Women 2 .14 −.09 .35 0.06

Men 4 .17* .04 .29

Perpetrating psych abuse 6 .51* .43 .58 .01 Women 3 .50* .42 .57 0.29

Men 3 .52*** .45 .59

Note. It is not unusual to see slight shifts in ES when looking at different moderators, but these shift are negligible when there are a large number of ES—this shift may more noticeable here due to the small number of ES. Thus, risk markers ES and significance levels from Tables 2 and 3 for victimization may differ slightly in Table 5. k = number of studies; mean ES = mean effect size; mean SE = mean standard error; LL = lower limit of confidence interval; UL = upper limit of confidence interval; Q^b = Cochran’s Q test of heterogeneity between studies.

p < .05. p < .01. p < .001 (two-tailed)

Publication Bias

Due to the low number of empirical same-sex IPV research reports available at the time of this meta-analysis, we were unable to test for the potential impact of publication bias on several risk markers (see Table 6). This lack of available effect sizes challenges the assumption behind publication bias tests regarding the presence of an ample pool of research, which may not reflect this particular field. For the risk markers that did have enough effect sizes to test for publication bias, we first used Duval and Tweedie’s (2000) trim-and-fill test, which uses a funnel plot to analyze (using random effects) asymmetrical distributions of included effect sizes to assess the publication bias of potential missing studies on our aggregated effect size. The trim-and-fill results suggested the impact of publication bias was trivial. Next, fail-safe Ns were estimated for each risk factor to evaluate the number of potential missing null studies needed to push the mean effect size above the p < .05 threshold (Rosenthal, 1979).

Table 6.
Duval and Tweedie’s Trim and Fill (Random Effects), Classic Fail-Safe N, and Orwin’s Fail-Safe N Tests for Risk Markers Associated With Perpetration and Victimization of IPV.

Risk Factor k Trim and Fill Classic Fail-Safe N Orwin’s Fail-Safe N

Imputed Studies r to .10

Female IPV perpetration

Alcohol abuse 8 0 194 7

Fusion 3 2 4 2

Internalized homophobia 3 0 6 —

Perpetrating psych abuse 3 0 283 17

Self-esteem 4 1 0 —

Victim of psychological abuse 4 0 287 18

Female IPV victimization

Alcohol abuse 6 1 81 2

Fusion 5 0 — —

Internalized homophobia 4 0 14 1

Perpetrating psych abuse 3 1 199 14

Self-esteem 3 0 — —

Victim of psychological abuse 6 0 1,163 30

Male IPV perpetration

Alcohol abuse 3 2 16 6

Child abuse in FOO 4 1 7 2

Witnessing parental IPV 5 0 — —

Victim of psychological abuse 3 2 129 14

Male IPV victimization

Child abuse in FOO 4 0 14 3

Perpetrating psych abuse 3 0 265 21

Witnessing parental IPV 3 0 8 3

Note. k = number of studies; IPV = intimate partner violence; FOO = family of origin.

Rosenthal (1979) recommends a threshold of 5k + 10 to guard against biased results, but it has been noted that this threshold has not been empirically tested since it was developed (Card, 2012). It has also been suggested that the number of studies expected to be unpublished in an area will likely depend on the field (Rosenthal, 1979). Thus, given the small number studies on risk markers for physical IPV in same-sex relationships, it is not surprising that there were a number of risk markers that did not meet this threshold (see Table 6). For female perpetrators, fusion, self-esteem, and internalized homophobia did not meet this threshold. For female victims of physical IPV, internalized homophobia did not meet this threshold. For male perpetrators of physical IPV, alcohol abuse and child abuse in FOO did not meet this criterion. Lastly, for male victims of physical IPV, witnessing parental IPV and child abuse in the FOO did not meet this threshold. With the exception of self-esteem, all of the risk markers mentioned above would require multiple studies to render the significance null. Clearly, more studies are needed to examine the significance of relationships between these risk markers and physical IPV in same-sex relationships and results should be interpreted with this in mind.

Finally, Orwin’s fail-safe Ns were estimated by testing how many potential missing studies with an effect size magnitude of r = .00 would be needed to reduce our mean effect size down to r = .10 (the lower limit of small effect sizes; Cohen 1992; Orwin, 1983). Although results indicated alcohol abuse and female victimization needed only one missing effect size of r = .00, the other results suggest our risk markers were robust against publication bias.

Discussion

Table 7 summarizes the key findings from the meta-analysis. Overall, our finding that there were no significant differences in the strength of risk markers in terms of their associations with perpetration versus victimization is not surprising given the high rate of bidirectional IPV found in previous research with heterosexual couples. For example, Whitaker, Haileyesus, Swahn, and Saltzman (2007) found that a quarter of all couples will experience violence at some point in their relationship, and over half of this violence will be mutual. This high rate of bidirectional violence is likely to apply to individuals in same-sex relationships also. Similarly, the lack of differences in the prediction of IPV risk factors for men versus women in same-sex relationships is also supported by the research in heterosexual relationships (Stith et al., 2004, 2014, 2016).

Table 7.
Critical Findings.

Significant risk markers for both IPV perpetration and victimization for men in same-sex relationships include perpetration of psychological abuse, being a victim of psychological abuse, alcohol abuse, and witnessing parental IPV

For men in same-sex relationships, child abuse in family of origin was a significant risk marker for IPV perpetration but not for IPV victimization

For women in same-sex relationships, the significant risk markers of both IPV perpetration and victimization were alcohol abuse, anger, psychological abuse perpetration, and psychological abuse victimization

Fusion was a significant risk marker for IPV perpetration in female, but not male, same-sex relationships

Research regarding same-sex-specific risk markers is extremely limited

Note. IPV = intimate partner violence.

For men in same-sex relationships, significant risk markers for both IPV perpetration and victimization included perpetration of psychological abuse, being a victim of psychological abuse, alcohol abuse, and witnessing parental IPV. We did not find evidence that these risk markers differed in strength as risk markers for IPV perpetration versus victimization. Although experiencing child abuse in FOO was a significant risk marker of IPV perpetration for men in same-sex relationships, it was not a significant risk marker for IPV victimization.

For women in same-sex relationships, several risk markers were associated with IPV perpetration and victimization, including alcohol abuse, anger, psychological abuse perpetration, and psychological abuse victimization. It is important to note that there was not statistical evidence that the strength of the risk markers differed between those who perpetrated IPV and those who were victims of IPV. The finding that psychological abuse was a strong risk marker for IPV perpetration and victimization was similar to findings in general meta-analyses (Stith et al., 2004, 2014). When comparing risk markers for IPV between men and women, we found that alcohol abuse was a significantly stronger risk marker for male IPV victimization than for female IPV victimization.

A primary finding in our review of same-sex-specific risk markers was that the research on them was extremely limited. In fact, several same-sex-specific risk markers for physical IPV victimization in men had one effect size and thus could not be examined in the meta-analysis, including being a victim of homophobic controlling behaviors, having a history of sex work, risky sexual behaviors, and outness. For physical IPV victimization in women, only one effect size was found for being a victim of homophobic controlling behaviors. This highlights the need for researchers of IPV in same-sex relationships to include same-sex-specific risk markers in subsequent studies.

Although there is a scarcity of research exploring same-sex-specific risk markers of physical IPV, we were able to demonstrate that internalized homophobia was a significant risk marker for IPV perpetration for both men and women. These findings are in line with previous findings that internalized homophobia is related to physical IPV perpetration. On the other hand, we did not find a gender difference between men and women in same-sex relationships. This suggests that gender does not moderate the association between internalized homophobia and perpetration of physical IPV. Because few previous studies have compared men and women for physical IPV in same-sex relationships, further research is needed to elucidate this finding. The way in which internalized homophobia relates to IPV has not been firmly established, but there has been speculation that the perpetrator may sometimes intentionally exacerbate feelings of homophobia in the victim. For example, some researchers have argued that IPV perpetrators may leverage the victim’s awareness of hatred of sexual minorities in mainstream society to reinforce the victim’s sense of responsibility for the abuse (Balsam & Szymanski, 2005) or to convince the victim that seeking help would be futile or even dangerous (Gehring & Vaske, 2015). When examining IPV perpetration, it may be that the perpetrator is projecting the negative feelings they have about themselves onto their partner, leading to the use of violence (Carvalho et al., 2011). Previous research has demonstrated associations between internalized homophobia and a number of previously identified risk markers for IPV, such as psychological maladjustment (Pepper & Sand, 2015), low self-esteem (Szymanski et al., 2001), and alcohol abuse (Lehavot & Simoni, 2011).

Despite not being included in the meta-analysis, we found one effect size for men and one effect size for women indicating that being victim of homophobic controlling behaviors was a risk marker for IPV victimization. It was the strongest risk marker that we found for both men and women’s IPV victimization. It has been speculated that individuals who experience homophobia may be more socially isolated than those who do not experience homophobia (Stephenson, de Voux, & Sullivan, 2011). The social isolation that victims of IPV can experience may be intensified for sexual minorities who are experiencing homophobic controlling behaviors (Allen & Leventhal, 1999). Because there was only one effect size for both men and women, it is difficult to draw conclusions, but this may be a specific type of psychological abuse that can occur in same-sex relationships. Although being a victim of homophobic controlling behaviors was a strong risk marker for being a victim of IPV for both men and women in same-sex relationships, there were not enough effect sizes to include in our meta-analysis; this suggests that homophobic controlling behaviors should be assessed in future research regarding same-sex IPV (see Balsam & Szymanski, 2005, for an example of how homophobic control behaviors can be measured).

For men in same-sex relationships, risky sexual behaviors and a history of sex work were risk markers for IPV victimization in one study. It is important to note that risky sexual behaviors could be a result of coercion and that a history of sex work could be involuntary. IPV victimization could impact an individual’s ability to advocate for their own sexual safety and ability to successfully negotiate with one’s partner to practice safe sexual behaviors, such as condom use (Relf, Huang, Campbell, & Catania, 2004).

A significant risk marker for IPV in female, but not male, same-sex relationships was fusion, or enmeshment with one’s partner. Fusion in lesbian relationships can lead to conflict when there is resentment surrounding isolation from friends and family members (Hill, 1999). It is possible that fusion contributes to resentment toward one’s partner, which may exacerbate the risk of IPV perpetration in female same-sex relationships. Our finding that only one risk marker that significantly differed between men and women mirrors research on risk markers for IPV perpetration and victimization in heterosexual relationships that have found few differences between risk markers for men and women (Spencer, Cafferky, & Stith, 2016). Although we did find that alcohol abuse was a stronger risk marker for same-sex IPV victimization in men compared to women, we are unaware of any theoretical orientation that could be used to adequately explain this finding. Subsequent studies are needed, therefore, to buttress the empirical support for this finding and to elucidate the factors that underlie it.

Limitations and Future Research

Overall, the lack of studies examining IPV perpetration and victimization in same-sex relationships made it difficult to thoroughly examine a wide variety of risk markers in the meta-analysis. Future research examining a multitude of potential risk markers for IPV perpetration and victimization in same-sex relationships is needed. More specifically, due to the lack of studies in which minority stressors are assessed and examined in relation to IPV, it is imperative for researchers to make a concerted effort to include measures of minority stressors in subsequent studies. This is one limitation of this meta-analysis, as many of the same-sex-specific risk markers in this study only had two effect sizes, it makes it far more difficult to draw definitive conclusions from these results. The inclusion of measures of minority stressors in studies of same-sex IPV will be useful in garnering more insight into the way in which minority stressors play a role in IPV in various populations of sexual minorities and, thus, inform the work of clinicians who assess and intervene with IPV perpetrators and victims who are in same-sex relationships.

Another consideration is that our conclusions about risk factors for men in same-sex relationships may not be as reliable as those about risk factors for women in same-sex relationships because we have 71 effect sizes for women in same-sex relationships and only 43 for men. More research on IPV in male relationships and research exploring differences between men and women in terms of risk markers for same-sex IPV is needed to facilitate the development of prevention and intervention programs that are aimed to address same-sex IPV. Studies are also needed to better understand the factors that contribute to such differences. For example, subsequent research may be useful in elucidating the factors that underlie the asymmetry in the strength of the association between alcohol abuse in men and women that was found in the present study.

Conclusion

Findings from this study seem to parallel findings in our ongoing meta-analysis of risk markers for IPV in heterosexual relationships (Stith et al., 2014, 2016). That is, relationship factors, including IPV victimization, seem to be the strongest risk markers for IPV perpetration. Furthermore, we found that alcohol abuse was a stronger risk marker for same-sex IPV in men compared to women, which demonstrates that the strength of the association between risk markers and IPV may differ among sexual minority populations. This study highlights the need for an increase in empirical research on risk markers for IPV in same-sex relationships and we hope our fellow researchers will join us as we take on this important endeavor.

Implications for Practice, Policy, and Research

In order to develop successful prevention and intervention programs that address intimate partner violence (IPV) perpetration and victimization in same-sex couples, research is needed that explores risk markers for IPV in same-sex relationships

Measures of same-sex-specific risk markers should be included in subsequent studies of IPV in same-sex relationships or in studies of risk markers of IPV that include individuals who are sexual minorities

Research is needed to explore whether risk markers of physical IPV perpetration and victimization differ between men and women in same-sex relationships

It is important for clinicians who assess and intervene with IPV perpetrators and victims who are in same-sex relationships to consider the role of sexual minority stressors in the cycle of abuse

Although there may be many similar risk markers of physical IPV in same-sex and heterosexual relationships, existing prevention and intervention programs designed primarily for those in heterosexual relationships may not adequately serve those in same-sex relationships, as these programs do not address sexual minority stressors

Supplemental Material

Supplementary_Material - A Meta-Analysis of Risk Markers for Intimate Partner Violence in Same-Sex Relationships

Supplementary_Material for A Meta-Analysis of Risk Markers for Intimate Partner Violence in Same-Sex Relationships by Jonathan G. Kimmes, Allen B. Mallory, Chelsea Spencer, Austin R. Beck, Bryan Cafferky and Sandra M. Stith in Trauma, Violence, & Abuse

Risk Marker	k	Mean ES	LL	UL
Men
Physical IPV Perpetration
Internalized homophobia	2	.23***	.11	.33
Outness	2	−.10	−.44	.26
Stigma consciousness	2	.02	−.07	.12
Physical IPV victimization
HIV status	2	.20	−.06	.44
Stigma consciousness	2	.03	−.07	.13
Women
Physical IPV perpetration
Fusion	3	.14**	.06	.22
Internalized homophobia	3	.09*	.00	.18
Physical IPV victimization
Internalized homophobia	4	.11	−.02	.24
Outness	2	.09	−.33	.48
Fusion	5	.01	−.25	.28

Risk Marker	k	Mean ES	LL	UL	Mean SE	Direction	k	Mean ES	LL	UL	Q^b
Alcohol abuse	5	.31***	.23	.39	.01	Perpetration	3	.27***	.14	.40	0.68
Victimization	2	.34***	.23	.45
Child abuse in family of origin	8	.16***	.08	.24	.01	Perpetration	4	.15*	.03	.27	0.04
Victimization	4	.17***	.06	.27
Education	4	.004	−.22	.23	.004	Perpetration	2	.04	−0.28	.36	0.10
Victimization	2	−.03	−.32	.27
Perpetrating psych abuse	5	.53***	.47	.59	.01	Perpetration	2	.55***	.41	.66	0.07
Victimization	3	.53***	.46	.59
Stigma consciousness	4	.02	−.04	.09	.00	Perpetration	2	.02	−.07	.12	0.01
Victimization	2	.03	−.06	.12
Witnessing IPV	8	.13*	.04	.22	.00	Perpetration	5	.09*	.00	.17	2.10
Victimization	3	.18***	.09	.27

Risk Marker	k	Mean ES	LL	UL	Mean SE	Direction	k	Mean ES	LL	UL	Q^b
Alcohol abuse	14	.16***	.11	.002	.01	Perpetration	8	.19***	.13	.24	1.78
Victimization	6	.13***	.08	.19
Anger	4	.23***	.19	.28	.002	Perpetration	2	.26***	.20	.31	1.10
Victimization	2	.21***	.16	.27
Child abuse in family of origin	4	.15*	.06	.25	.01	Perpetration	2	.16*	.04	.26	0.01
Victimization	2	.15	−.03	.32
Fusion	8	.05	−.12	.22	.04	Perpetration	3	.11	−.17	.36	0.28
Victimization	5	.01	−.20	.23
Internalized homophobia	7	.11*	.02	.19	.01	Perpetration	3	.11	−.02	.22	0.02
Victimization	4	.11	−.003	.22
Poor mental health^a	4	.13	−.19	.43	.07	Perpetration	2	−.05	−.39	.31	1.88
Victimization	2	.28	−.03	.54
Perpetrating psych abuse	6	.54***	.46	.61	.01	Perpetration	3	.57***	.49	.65	1.54
Victimization	3	.50***	.40	.58
Self-esteem	7	−.04	−.21	.14	.03	Perpetration	4	.04	−.19	.27	1.03
Victimization	3	−.14	−.40	.13
Victim of psych abuse	10	.50***	.44	.55	.003	Perpetration	4	.47***	.39	.53	1.75
Victimization	6	.52***	.47	.57

Risk Markers	k	Mean ES	LL	UL	Mean SE	Gender	k	Mean ES	LL	UL	Q^b
Alcohol abuse	11	.21***	.13	.28	.003	Women	8	.19***	.06	.33	1.43
Men	3	.27***	.02	.45
Child abuse in family of origin	6	.15***	.08	.22	.01	Women	2	.16*	.04	.26	0.01
Men	4	.15**	.06	.24
Education	5	.15	−.06	.34	.04	Women	3	.24*	.01	.44	1.40
Men	2	.02	−.25	.29
Internalized homophobia	5	.15*	.02	.28	.01	Women	3	.09*	.01	.18	2.88
Men	2	.23**	.11	.36
Perpetrating psych abuse	5	.48***	.43	.69	.01	Women	3	.57***	.41	.56	0.11
Men	2	.55***	.49	.65
Witnessing parental IPV	7	.09*	.02	.17	.01	Women	2	.11	−.01	.24	0.11
Men	5	.09	−.0002	.17
Victim of psychological abuse	7	.47***	.39	.54	.01	Women	4	.46***	.36	.55	0.11
Men	3	.49***	.37	.59

Risk Marker	k	Mean ES	LL	UL	Mean SE	Gender	k	Mean ES	LL	UL	Q^b
Alcohol abuse	8	.24*	.03	.44	.01	Women	6	.14***	.09	.18	22.56***
Men	2	.36***	.28	.43
Child abuse in family of origin	6	.16***	.05	.27	.01	Women	2	.14	−.09	.35	0.06
Men	4	.17*	.04	.29
Perpetrating psych abuse	6	.51***	.43	.58	.01	Women	3	.50***	.42	.57	0.29
Men	3	.52***	.45	.59

Risk Factor	k	Trim and Fill	Classic Fail-Safe N	Orwin’s Fail-Safe N
Female IPV perpetration
Alcohol abuse	8	0	194	7
Fusion	3	2	4	2
Internalized homophobia	3	0	6	—
Perpetrating psych abuse	3	0	283	17
Self-esteem	4	1	0	—
Victim of psychological abuse	4	0	287	18
Female IPV victimization
Alcohol abuse	6	1	81	2
Fusion	5	0	—	—
Internalized homophobia	4	0	14	1
Perpetrating psych abuse	3	1	199	14
Self-esteem	3	0	—	—
Victim of psychological abuse	6	0	1,163	30
Male IPV perpetration
Alcohol abuse	3	2	16	6
Child abuse in FOO	4	1	7	2
Witnessing parental IPV	5	0	—	—
Victim of psychological abuse	3	2	129	14
Male IPV victimization
Child abuse in FOO	4	0	14	3
Perpetrating psych abuse	3	0	265	21
Witnessing parental IPV	3	0	8	3

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

The online data supplements are available at .

References

Allen

Leventhal

(1999). History, culture, and identity: What makes GLBT battering different. In Leventhal

Lundy

S. E.

(Eds.), Same-sex domestic violence: Strategies for change (pp. 73–81). Thousand Oaks, CA: Sage.

Badenes-Ribera

Frias-Navarro

Bonilla-Campos

Pons-Salvador

Monterde-i-Bort

(2015). Intimate partner violence in self-identified lesbians: A meta-analysis of its prevalence. Sexuality Research and Social Policy, 12, 47–59.

Baker

N. L.

Buick

J. D.

Kim

S. R.

Moniz

Nava

K. L.

(2013). Lessons from examining same-sex intimate partner violence. Sex Roles, 69, 182–192.

Balsam

K. F.

Szymanski

(2005). Relationship quality and domestic violence in women’s same-sex relationships: The role of minority stress. Psychology of Women Quarterly, 29, 258–269.

Black

M. C.

Basile

K. C.

Breiding

M. J.

Smith

S. G.

Walters

M. L.

Merrick

M. T.

… Stevens

M. R.

(2011). The national intimate partner and sexual violence survey (NISVS): 2010 summary report. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

Borenstein

Hedges

Higgins

Rothsteine

(2005). Comprehensive meta-analysis (Version 2) [Computer software]. Englewood, NJ: Biostat.

Buller

A. M.

Devries

K. M.

Howard

L. M.

Bacchus

L. J.

(2014). Associations between intimate partner violence and health among men who have sex with men: A systematic review and meta-analysis. PLoS Medicine, 11, e1001609.

Burke

L. K.

Follingstad

D. R.

(1999). Violence in lesbian and gay relationships: Theory, prevalence, and correlational factors. Clinical Psychology Review, 19, 487–512.

Cohen

(1992). A power primer. Psychological Bulletin, 112(1), 155–159.

10.

Card

N. A.

(2012). Applied meta-analysis for social science research. New York, NY: Guilford Press.

11.

Carvalho

A. F.

Lewis

R. J.

Derlega

V. J.

Winstead

B. A.

Viggiano

(2011). Internalized sexual minority stressors and same-sex intimate partner violence. Journal of Family Violence, 26, 501–509.

12.

Dutton

M. A.

Goodman

L. A.

Bennett

(2001). Court involved battered women’s responses to violence: The role of psychological, physical, and sexual abuse. In O’Leary

K. D.

Maiuro

(Eds.), Psychological abuse in violent domestic relationships (pp. 177–195). New York, NY: Springer.

13.

Duval

Tweedie

(2000). Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics, 56, 455–463.

14.

Edwards

K. M.

Sylaska

K. M.

Neal

A. M.

(2015). Intimate partner violence among sexual minority populations: A critical review of the literature and agenda for future research. Psychology of Violence, 5, 112–121.

15.

Gehring

K. S.

Vaske

J. C.

(2015). Out in the open: The consequences of intimate partner violence for victims in same-sex and opposite-sex relationships. Journal of Interpersonal Violence, 32, 3669–3692. doi:10.1177/0886260515600877.

16.

Hatzenbuehler

M. L.

Nolen-Hoeksema

Erickson

S. J.

(2008). Minority stress predictors of HIV risk behavior, substance use, and depressive symptoms: Results from a prospective study of bereaved gay men. Health Psychology, 27, 455–462.

17.

Hawkins

Blanchard

Baldwin

Fawcett

(2008). Does marriage and relationship education work? A meta-analytic study. Journal of Consulting and Clinical Psychology, 76, 723–734.

18.

Hill

(1999). Fusion and conflict in lesbian relationships? Feminism & Psychology, 9, 179–185.

19.

Hunter

J. E.

Schmidt

F. L.

(Eds.). (2004). Methods of meta-analysis: Correcting error and bias in research findings. London, England: Sage.

20.

Lehavot

Simoni

J. M.

(2011). The impact of minority stress on mental health and substance use among sexual minority women. Journal of Consulting and Clinical Psychology, 79, 159–170.

21.

McClennen

J. C.

(2005). Domestic violence between same-gender partners recent findings and future research. Journal of Interpersonal Violence, 20, 149–154.

22.

Meyer

I. H.

(1995). Minority stress and mental health in gay men. Journal of Health and Social Behavior, 36, 38–56.

23.

Meyer

I. H.

(2013). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychology of Sexual Orientation and Gender Diversity, 1, 3–26.

24.

Newcomb

M. E.

Mustanski

(2011). Moderators of the relationship between internalized homophobia and risky sexual behavior in men who have sex with men: A meta-analysis. Archives of Sexual Behavior, 40, 189–199.

25.

Orwin

R. G.

(1983). A fail-safe N for effect size in meta-analysis. Journal of Educational Statistics, 8(2), 157–159.

26.

Pico-Alfonso

M. A.

(2005). Psychological intimate partner violence: The major predictor of posttraumatic stress disorder in abused women. Neuroscience and Biobehavioral Reviews, 29, 181–193.

27.

Relf

M. V.

Huang

Campbell

Catania

(2004). Gay identity, interpersonal violence, and HIV risk behaviors: An empirical test of theoretical relationships among a probability-based sample of urban men who have sex with men. Journal of the Association of Nurses in AIDS Care, 15, 14–26.

28.

Rosenthal

(1979). The file drawer problem and tolerance for null results. Psychological Bulletin, 86, 638–641.

29.

Spencer

Cafferky

Stith

S. M.

(2016). Gender differences in risk markers for perpetration of physical partner violence: results from a meta-analytic review. Journal of Family Violence, 31(8), 981–984.

30.

Steele

L. S.

Tinmouth

J. M.

(2006). Regular health care use by lesbians: A path analysis of predictive factors. Family Practice, 23, 631–636.

31.

Stephenson

de Voux

Sullivan

P. S.

(2011). Intimate partner violence and sexual risk-taking among men who have sex with men in South Africa. Western Journal of Emergency Medicine, 12, 343–347.

32.

Stith

S. M.

Cafferky

Bird

Lawson

Smith

Luu

Mallory

Barros Gomez

(2014). Overview of a meta-analytic review of risk markers for IPV. International Family Violence and Child Victimization Research Conference, Portsmouth, NH.

33.

Stith

S. M.

Smith

D. B.

Penn

C. E.

Ward

D. B.

Tritt

(2004). Intimate partner physical abuse perpetration and victimization risk factors: A meta-analytic review. Aggression and Violent Behavior, 10(1), 65–98.

34.

Szymanski

D. M.

Chung

Y. B.

Balsam

K. F.

(2001). Psychosocial correlates of internalized homophobia in lesbians. Measurement and Evaluation in Counseling and Development, 34, 27–41.

35.

Whitaker

D. J.

Haileyesus

Swahn

Saltzman

L. S.

(2007). Differences in frequency of violence and reported injury between relationships with reciprocal and nonreciprocal intimate partner violence. American Journal of Public Health, 97(5), 941–947.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.14 MB

Risk Factor	k	Trim and Fill	Classic Fail-Safe N	Orwin’s Fail-Safe N
Risk Factor	k	Imputed Studies	Classic Fail-Safe N	r to .10
Female IPV perpetration
Alcohol abuse	8	0	194	7
Fusion	3	2	4	2
Internalized homophobia	3	0	6	—
Perpetrating psych abuse	3	0	283	17
Self-esteem	4	1	0	—
Victim of psychological abuse	4	0	287	18
Female IPV victimization
Alcohol abuse	6	1	81	2
Fusion	5	0	—	—
Internalized homophobia	4	0	14	1
Perpetrating psych abuse	3	1	199	14
Self-esteem	3	0	—	—
Victim of psychological abuse	6	0	1,163	30
Male IPV perpetration
Alcohol abuse	3	2	16	6
Child abuse in FOO	4	1	7	2
Witnessing parental IPV	5	0	—	—
Victim of psychological abuse	3	2	129	14
Male IPV victimization
Child abuse in FOO	4	0	14	3
Perpetrating psych abuse	3	0	265	21
Witnessing parental IPV	3	0	8	3