Profiles of Childhood Victimization as Predictors of Sexual Aggression and Intimate Partner Violence Perpetration

Abstract

The study utilized a latent profile analysis to evaluate distinct combinations of men's childhood victimization experiences and their communal predictive ability for sexual aggression (SA) and intimate partner violence (IPV) perpetration. Men living in the United States (N = 399) completed assessments of emotional, physical, and sexual childhood victimization, as well as SA and IPV toward women. The results indicated that members of profiles characterized by moderate to high frequencies of childhood victimization—especially sexual victimization—were at increased risk of SA and IPV perpetration. Contrary to expectations, adult- versus peer-perpetrated victimization did not appear to substantially alter risk of perpetration.

Keywords

childhood abuse sexual assault sexual violence relationship violence domestic violence

Introduction

Sexual aggression is a form of violence that can consist of physical, verbal, or psychological coercion in an attempt to engage in sexual activity when consent is not received or freely given, whereas intimate partner violence includes physical, psychological, and sexual violence directed toward a romantic partner. Despite considerable attention in the literature, male-perpetrated sexual aggression and intimate partner violence toward women remain persistent issues in the United States. Recent estimates from nationwide samples of United States residents suggest that approximately one in three men have perpetrated sexual aggression (Abbey et al., 2021; Casey et al., 2017). Similar estimates are found for intimate partner violence, for which approximately one in five men reported prior intimate partner violence perpetration (Desmarais et al., 2012; Singh et al., 2014).

Due to the fairly large proportions of men with histories of violent behavior toward women, researchers continue to seek empirical knowledge about the nuanced effects of risk factors to inform evidence-based prevention efforts. Childhood victimization (i.e., emotional, physical, or sexual victimization occurring before the age of 18) is a common risk factor that has received ample evidence as a predictor of both types of perpetration (Casey et al., 2017; Godbout et al., 2019). Although much research in the extant literature utilized composite measures of childhood victimization, researchers recently highlighted the differential risk that each type of victimization (e.g., emotional, physical, sexual) may hold in the development of aggressive behavior (Casey et al., 2017; Richards et al., 2017). Unfortunately, research assessing the risk associated with each type of childhood victimization has primarily used variable-centered approaches, which can discount their combined experience and compounded effects (Davis et al., 2018). In light of this limitation, the current study utilized a latent profile analysis (a person-centered approach) to explore distinct combinations of men's childhood victimization experiences (i.e., sexual, physical, and emotional) and their communal predictive ability for sexual aggression and intimate partner violence perpetration. The study also extended prior research by including indicators of peer-perpetrated victimization, alongside indicators of victimization from adults, to examine their combined impact on risk of future perpetration.

Childhood Victimization and Sexual Aggression

Childhood victimization has long been recognized as a risk factor for sexual aggression. Theoretical models (e.g., the confluence model of sexual aggression; Malamuth et al., 1991), for instance, consider childhood victimization among men to be a distal risk factor in the etiology of sexual aggression (Abbey et al., 2006, 2011; Davis et al., 2012; Malamuth & Hald, 2016). However, the level of risk appears somewhat dependent on the type of childhood victimization that was experienced. Childhood sexual and physical victimization perpetrated by adults are both associated with sexual aggression (Casey et al., 2009, 2017; White & Smith, 2004), but there is some evidence to suggest that sexual victimization might be more strongly linked to future perpetration (Casey et al., 2009, 2017).

Casey et al. (2009, 2017) demonstrated that only sexual victimization in childhood is directly related to sexual aggression when included in a statistical model alongside physical victimization from adults and witnessed parental violence. In addition to these results, Casey et al. (2017) conducted a follow-up analysis to investigate the additive effects of poly-victimization (i.e., experiencing multiple forms of victimization) on subsequent risk of sexual aggression. Men who experienced multiple forms of childhood victimization (i.e., sexual, physical from adults, witnessed partner violence) were more likely to perpetrate rape compared to men who reported only sexual victimization or physical victimization from adults, or had no history of victimization. These results provide some evidence for the additive effects that multiple types of victimization can have on risk of sexual aggression, but because victimization experiences were coded as binary (i.e., participants were categorized as “experienced victimization” or “did not experience victimization” for each type of victimization), the researchers were limited in their ability to examine the influence of varying frequencies of victimization on risk of future perpetration. More research is needed to better understand the additive effects that distinct types of childhood victimization can have on future perpetration, particularly work that considers the frequency at which victimization experiences occurred.

Childhood Victimization and Intimate Partner Violence Perpetration

In addition to being a risk factor for sexual aggression, childhood victimization also is a risk factor for intimate partner violence perpetration. Empirical evidence supports the associations between intimate partner violence perpetration and several types of childhood victimization, including emotional, physical, and sexual victimization (Brassard et al., 2014; Brennan et al., 2020; Daigneault et al., 2009; Fang & Corso, 2007; Richards et al., 2017; Smith-Marek et al., 2015; Stith et al., 2000). Further evidence is provided through meta-analytic results, which indicated that all three types of childhood victimization are predictive of men's violent behavior toward intimate partners (Godbout et al., 2019).

There is no clear indication of which type of childhood victimization is most detrimental to future behavioral outcomes (if any), but the existing literature could offer hypotheses. For example, some results suggest that sexual victimization is the strongest individual predictor of future intimate partner violence perpetration (Richards et al., 2016, 2017). However, emotional childhood victimization from adults is exceptionally prevalent (Finkelhor et al., 2015), which has led some to argue that it may have wider impact than other forms of victimization (Richards et al., 2017). Unfortunately, emotional victimization is greatly understudied in the literature, leaving its impact relatively unclear. Additionally, little research has examined the compounded effects of various forms of childhood victimization on intimate partner violence perpetration. The current research attempts to address these latter deficiencies.

Adult- and Peer-Perpetrated Victimization

The literature examining sexual aggression and intimate partner violence has almost exclusively focused on children's victimization experiences from adults, which ignores the potential impact that peer-perpetrated victimization may have. An example of its impact was demonstrated in a meta-analysis conducted by Reijntjes et al. (2010), which found that peer-perpetrated childhood victimization was associated with subsequent internalizing problems (i.e., mental health issues; somatic symptoms). More recent research indicated that bullying victimization in childhood was associated with violence-related cognitions (e.g., gun-supportive attitudes) and behaviors (e.g., gun carrying; physical assault) in adulthood (Hoffman et al., 2017; Ray et al., 2021; Turner et al., 2016). These results suggest that peer-perpetrated aggression has the potential to substantially affect its victims, which may compound the risk of adverse behavioral outcomes produced by adult-perpetrated aggression (Khan et al., 2015). Put differently, both types of victimization may uniquely contribute to risk of future perpetration due to their differences in power dynamics and dissimilar locations (e.g., children's homes versus school). Thus, the combination of peer- and adult-perpetrated victimization may enhance risk of future perpetration, compared to victimization from just peers or adults, suggesting both types should be considered. However, further research is needed to better understand the effects of peer-perpetrated victimization on men's violence perpetration toward women, especially its compounded effects with adult-perpetrated victimization.

The Use of Latent Profile Analysis in Childhood Victimization Research

Researchers have highlighted the need to conduct empirical investigations in ways that are conducive to better understanding distinct types of childhood victimization and how their combined experience may pose risk for maladaptive outcomes (Casey et al., 2017; Davis et al., 2018; Turner et al., 2017). The combined experiences of childhood victimization are best captured through person-centered approaches. As opposed to variable-centered approaches (which examine linear relationships between variables), person-centered approaches offer the flexibility to detect common patterns of variables among individuals and determine how such patterns vary in regards to outcome risk. Within the childhood victimization literature, most research utilized a variable-centered approach to observe the outcomes associated with composite scores of victimization (i.e., all experiences are accounted for within one variable) or the independent relationships between outcomes and predictor variables representing a single victimization type (e.g., sexual victimization). If utilizing a latent profile analysis, a person-centered approach, common patterns of experiences should emerge (i.e., profiles), thus allowing for the inspection of childhood victimization experiences’ interactive impact. In this way, latent profile analyses can simplify the interpretation of complex interactions relative to variable-centered approaches (e.g., moderated multiple regression). Such analyses could provide further conceptual clarity about men's childhood victimization experiences and how common patterns of experiences relate to future perpetration.

Davis et al. (2018) conducted a recent profile analysis examining men's childhood victimization experiences and their associations with sexual aggression and intimate partner violence perpetration. Five types of childhood victimization were examined: emotional neglect, emotional abuse, physical neglect, physical abuse, and sexual abuse. From their analysis, four profiles emerged: ‘Low Maltreatment’ (i.e., low frequencies of all types of victimization), ‘Emotional and Physical Maltreatment’ (i.e., high frequencies of all types of victimization except sexual abuse), ‘Emotional and Sexual Maltreatment’ (i.e., high frequencies of emotional neglect, emotional abuse, and sexual abuse), and ‘Poly-Victimized’ (i.e., high frequencies of all types of victimization). Their results indicated that the ‘Emotional and Physical Maltreatment’ profile had higher rates of intimate partner violence perpetration when compared to the ‘Low Maltreatment’ profile, but there were no significant pairwise differences between any of the remaining profiles. Additionally, there were no significant differences between profiles in regards to sexual aggression. Although these results are quite informative, the authors note limitations that warrant further research. Specifically, there were small numbers of participants in the ‘Emotional and Sexual Maltreatment’ and ‘Poly-Victimized’ profiles, which the authors indicated might have altered their results in ways that are inconsistent with previous research (e.g., no significant group differences for sexual aggression). Thus, additional research utilizing a latent profile analysis may be beneficial.

There are several gaps in the literature for which the current research aims to address. Besides the study conducted by Davis et al. (2018), no research has examined profiles of men's childhood victimization experiences predicting sexual aggression and intimate partner violence perpetration. More research using person-centered approaches is needed to better understand the nuance of perpetration risk factors, such as childhood victimization, especially research that aims to conceptually replicate or extend initial findings. As previously mentioned, there has been little attention devoted to the effects of victimization experiences based on the male child's relationship to the perpetrator. Thus, it remains unclear how peer-perpetrated victimization might influence risk of sexual aggression and intimate partner violence, including its combined effects with adult-perpetrated victimization. A profile analysis that incorporates indicators of peer- and adult-perpetrated childhood victimization can address this research gap.

Current Study

The current study utilized a latent profile analysis to examine discrete patterns of childhood victimization experiences (i.e., emotional, physical, sexual) and their association with sexual aggression and intimate partner violence perpetration. The study extends prior research (e.g., Davis et al., 2018) by examining emotional and physical childhood victimization perpetrated by peers, in addition to adults. Given the low prevalence of sexual victimization in childhood (Finkelhor et al., 2015), we sought to avoid similar issues (i.e., low power within profiles; few observed instances of sexual victimization) to those of Davis et al. (2018) by examining sexual victimization in a general sense (as opposed to separating sexual victimization by peer and adult perpetrators). Therefore, a total of five victimization experiences were examined in the current study (i.e., peer-perpetrated emotional, peer-perpetrated physical, sexual, adult-perpetrated emotional, adult-perpetrated physical).

Although latent profile analysis is a relatively exploratory approach, the current study had several hypotheses. Consistent with the majority of prior research, it was hypothesized that profiles with generally high frequencies of childhood victimization experiences would be associated with greater likelihoods of sexual aggression and intimate partner violence perpetration when compared to profiles with low frequencies (Hypothesis 1). It also was expected that profiles characterized by high frequencies of sexual victimization would be at particularly increased risk of perpetration (i.e., sexual aggression and intimate partner violence) when compared to profiles with low frequencies of sexual victimization (Hypothesis 2). Finally, it was hypothesized that profiles characterized by high frequencies of adult perpetrators (i.e., emotional and physical victimization) would be more predictive of perpetration than profiles characterized by peer perpetrators (Hypothesis 3).

Methodology

Participants

Participants were 399 men living in the United States. The average age was 30.70 (SD = 8.59) and ranged from 18 to 66 years old. The majority of participants were Caucasian/White (n = 284; 71.2%), while the remaining participants reported that their race/ethnicity was African-American/Black (n = 36; 9.0%), Hispanic/Latino (n = 28; 7.0%), Asian-American/Asian (n = 25; 6.3%), American Indian/Alaskan Native (n = 11; 2.8%), Multiracial (n = 11; 2.8%), or not listed (n = 4; 1.0%). Nearly all participants reported at least some college education (n = 361; 90.5%), with just over half holding a Bachelor's degree or higher (n = 216; 54.1%). The region of residence was diverse with participants living in the Southeast (n = 121; 30.3%), Midwest (n = 97; 24.3%), Northeast (n = 79; 19.8%), West (n = 70; 17.5%), and Southwest (n = 32; 8.0%).

Procedure

Participants were recruited from Amazon Mechanical Turk (Mturk) to complete an online questionnaire. Mturk is an online crowdsourcing marketplace that has been shown to produce valid data (Casler et al., 2013) that is significantly more diverse than college samples and slightly more diverse than standard internet samples (Buhrmester et al., 2011). A screening procedure with ambiguous inclusion criteria (i.e., criteria that was not evident to participants) was used to obtain participants who met eligibility criteria. The procedure was consistent with suggestions for “best research practices” offered by past researchers utilizing Mturk samples. Specifically, methodological research suggests ambiguous screening procedures can defend against participants who lie on screener items (when inclusion criteria is explicit) to gain access to questionnaires and subsequent compensation (Chandler & Paolacci, 2017).

Interested participants accessed the screener questionnaire through a web-link posted on the Mturk website. The screener questionnaire consisted of various demographic questions. In order to qualify for the study, participants must have reported they were (1) male, (2) 18 years of age or older, and (3) live in the United States. Respondents who did not meet these criteria were informed of their ineligibility to participate. Mturk participants who attempted to take the screener questionnaire more than once were automatically redirected to an ineligibly page. A total of 1,284 individuals completed the screener questionnaire; 557 (43.4%) did not meet inclusion criteria. Of these individuals, 545 (97.8%) were excluded because they were female, one (0.2%) was under the age of 18, five (0.9%) lived outside of the United States, and five (0.9%) failed to meet multiple inclusion criteria. Those who met inclusion criteria were given access to the study's questionnaire. “Attention check” items were interspersed throughout the questionnaire to reduce careless responding and ensure the quality of the data. Participants who failed an attention check (n = 328) were removed from the study, resulting in a total of 399 men. Participants who completed the study questionnaire were paid $3.00 through their Mturk account. All study procedures were approved by the affiliated Institutional Review Board.

Measures

Childhood victimization

Childhood victimization was measured using the Juvenile Victimization Questionnaire—Short Form (JVQ; Finkelhor et al., 2005; Hamby et al., 2011). The JVQ is a 21-item self-report measure assessing the frequencies of various forms of childhood victimization on a scale from 0 (0 times) to 5 (5 or more times). Five items were used for the purposes of the current study. The items assessed emotional victimization from peers (i.e., “Did you get scared or feel really bad because kids were calling you names, saying mean things to you, or saying they didn't want you around?”), physical victimization from peers (i.e., “Did any other kid ever hit you on purpose?”), sexual victimization (i.e., “Did someone make you do sexual things when you didn't want to?”), emotional victimization from adults (i.e., “Did you get scared or feel really bad because grown-ups called you names, said mean things to you, or said they didn't want you?”), and physical victimization from adults (i.e., “Not including spanking on your bottom, did a grown-up in your life hit you?”).

Sexual aggression

Sexual aggression was measured using the Sexual Experiences Survey (SES; Abbey et al., 2005; Koss et al., 1987). The SES is comprised of 24 self-report items assessing the frequency of sexually aggressive behaviors. Participants were asked to indicate how many times they have engaged in behaviors such as “I put my penis or I put my fingers or objects into someone's vagina without their consent by threatening to physically harm them or someone close to them” on a scale from 0 (0 times) to 3 (3 or more times). Overall scores (α = .98) were recoded into a dichotomous (i.e., yes/no) variable, wherein participants indicating that they had perpetrated one or more sexually aggressive behavior were coded as 1 (yes) and participants reporting that they have not perpetrated any behaviors were coded as 0 (no).

Intimate partner violence

Intimate partner violence was measured using the Relationship Violence Scale—Revised for Perpetration (RVS; Hamby & Turner, 2013). The RVS is a 5-item self-report measure assessing violent behavior toward an intimate partner. Participants were asked to indicate how often they have engaged in behavior such as “Not including horseplay or joking around, I have hit my partner” and “Not including horseplay or joking around, I have beat up my partner” on a scale from 0 (Never) to 4 (Often). Overall scores (α = .96) were recoded into a dichotomous (i.e., yes/no) variable, wherein participants indicating that they had perpetrated one or more behaviors were coded as 1 (yes) and participants reporting that they have not perpetrated any behaviors were coded as 0 (no).

Analytic Approach

A Latent Profile Analysis was conducted using Mplus 7 (Muthén & Muthén, 2012). Prior to exploring profile solutions, patterns of missing data were examined. In the current data, there were nine patterns of missing data all of which had a frequency of five or less—indicating that data were missing at random. A full information maximum likelihood estimator was used to adjust for incidents of missing data. Profile solutions were compared by, first, specifying a one-profile model and incrementally introducing profiles in subsequent models. Multiple fit indices were examined to determine the best profile solution, such as likelihood ratio tests, criteria classification quality (entropy), and information criterion. Information criterion consisted of Akaike (AIC), constant Akaike (CAIC), Bayesian (BIC), and sample-size adjusted Bayesian (aBIC) information criteria values. Relative increases in fit between profile solutions were determined by substantial decreases in information criteria and significant p-values (p < .05) for Vuong–Lo–Mendell–Rubin (VLMR) likelihood ratio and Lo–Mendell–Rubin adjusted LRT (LMR) tests. Theoretical relevance also was considered in concluding the optimal number of profiles within the model.

Once the optimal profile solution was determined, profile memberships were specified as predictors of the study's outcome variables (i.e., sexual aggression and intimate partner violence perpetration). The outcome variables were entered into the model using an auxiliary command, which allows the structure of the profile solution to be unaffected by the inclusion of additional variables (McLarnon & O’Neill, 2018). The outcome variables were coded as binary and, thus, the DCAT specification was used as part of the auxiliary command. The DCAT specification provides odds ratios indicating the likelihood that membership in a particular profile will predict classification in the outcome variable compared to a reference profile. For the current study, odds ratios can be interpreted as “risk” of perpetration (i.e., sexual aggression; intimate partner violence) for members of one profile compared to members of another.

Results

Descriptive Statistics and Bivariate Correlations

Most participants (n = 359; 90.0%) experienced at least one form of childhood victimization, 294 participants (73.7%) experienced multiple forms, and 58 participants (14.5%) experienced all forms. Nearly half of the participants (n = 179; 44.9%) reported a history of either sexual aggression or intimate partner violence, and 70 participants (17.5%) perpetrated both forms of aggression. Additional descriptive statistics are provided in Table 1, which includes the frequency and percentage of the sample that reported each variable at least once, as well as the means, standard deviations, and ranges for all variables. Table 2 provides bivariate correlations between the five types of childhood victimization and the two types of perpetration. All variables were significantly and positively correlated, with the exception of sexual aggression with emotional and physical victimization from peers, which were non-significant. Relative to the other significant associations, there were particularly strong correlations between emotional victimization from peers and physical victimization from peers (r = .54, p < .001) as well as emotional victimization from adults and physical victimization from adults (r = .53, p < .001). Additionally, sexual victimization was most strongly associated with sexual aggression (r = .27, p < .001) and intimate partner violence perpetration (r = .32, p < .001), relative to the other types of victimization.

Table 1.

Descriptives of Study Variables for the Full Sample (N = 399).

	n	%	M	SD	Range
Victimization Type
Peer, Emotional	275	68.9	2.26	1.97	0.00–5.00
Peer, Physical	306	76.7	2.19	1.75	0.00–5.00
Sexual	86	21.6	0.61	1.34	0.00–5.00
Adult, Emotional	163	40.9	1.35	1.88	0.00–5.00
Adult, Physical	198	49.6	1.48	1.89	0.00–5.00
Perpetration Type
Sexual Aggression	121	30.3	4.90	12.54	0.00–64.00
Intimate Partner Violence	128	32.1	1.88	4.26	0.00–18.00

Note. n = number of participants reporting the respective experience at least once; % = proportion of whole sample that reported the respective experience at least once. The means, standard deviations, and ranges in the ‘Sexual Aggression’ and ‘Intimate Partner Violence’ rows are representative of their respective summed score prior to dichotomizing the variables.

Table 2.

Bivariate Correlations Between Types of Childhood Victimization and Perpetration.

	1	2	3	4	5	6	7
1. Peer, Emotional	—
2. Peer, Physical	0.54***	—
3. Sexual	0.22***	0.26***	—
4. Adult, Emotional	0.33***	0.43***	0.37***	—
5. Adult, Physical	0.31***	0.41***	0.40***	0.53***	—
6. Sexual Aggression	0.08	0.10	0.27***	0.19***	0.14**	—
7. Intimate Partner Violence	0.14**	0.14**	0.32***	0.20***	0.20***	0.44***	—

Note. The associations between childhood victimization variables are represented by Pearson correlation coefficients. Point-Biserial correlation coefficients are provided for all associations with a perpetration variable.

**p < .01.

***p < .001.

Latent Profile Analysis

Fit indices for each of the six profile solutions can be seen in Table 3. The five-profile model was determined to be the optimal solution for a number of reasons. First, information criteria continued to decrease through the six-profile solution, but the improvement from the five-profile model to the six-profile model was trivial, which was supported by the non-significant VLMR and LMR tests. Additionally, the six-profile solution had a non-replicated loglikelihood value, which suggested that the results of the model were unreliable. The five-profile model was concluded as superior to the four-profile model because the significant VLMR and LMR tests indicated substantial decrease in criterion values. The five-profile model also added a profile (i.e., ‘High Peer Victimization,’ which is described below) that was theoretically relevant and allowed for essential empirical comparisons.

Table 3.

Latent Profile Analysis Model Fit Indices.

# Profiles	#fp	LL	LLc	AIC	CAIC	BIC	aBIC	Entropy	VLMR	LMR
1	10	−3895.52	0.99	7811.04	7851.93	7850.93	7819.20	—	—	—
2	16	−3594.75	1.27	7221.50	7286.32	7285.32	7234.55	.94	<.001	<.001
3	22	−3400.85	1.45	6845.69	6934.45	6933.45	6863.64	.96	<.001	<.001
4	28	−3247.18	1.35	6550.35	6663.04	6662.05	6573.20	.97	<.001	<.001
5	34	−3177.91	1.38	6423.82	6560.44	6559.44	6451.56	.93	<.001	<.001
6	40	−3124.13	1.46	6328.26	6488.81	6487.82	6360.89	.94	.303	.314

Note. #fp = number of free parameters estimated in each model; LL = model loglikelihood; LLc = scaling correction factor for loglikelihood; AIC = Akaike information criterion; CAIC = consistent AIC; BIC = Bayesian information criterion; aBIC = sample-size adjusted BIC; Entropy = index of classification quality; VLMR = Vuong–Lo–Mendell–Rubin likelihood ratio test p-value; LMR = Lo–Mendell–Rubin adjusted LRT test p-value.

The bolded values highlight the retained profile solution.

Profile Interpretation

The five-profile solution revealed distinct profiles of childhood victimization experiences, which can be seen in Figure 1. Profile means and standard errors are reported in Table 4. The ‘Low Victimization’ profile (n = 195) consisted of participants who reported low frequencies of occurrences for all childhood victimization types. Explained similarly, the ‘Moderate Victimization’ (n = 36) profile contained participants who reported moderate frequencies for all victimization types, while the ‘High Victimization’ (n = 30) profile encompassed participants who reported high frequencies of all childhood victimization types. The fourth profile, ‘High Victimization, Non-Sexual’ (n = 61), consisted of participants who reported high frequencies of emotional and physical victimization from both peers and adults, but a low frequency of childhood sexual victimization. The final profile, ‘High Peer Victimization’ (n = 77), contained participants who reported high frequencies of emotional and physical victimization, but only from peers.

Figure 1.

Means of Identifiers for Each Profile.

Table 4.

Means and Standard Errors for Victimization Type Across the Five-Profile Solution.

	M (SE)
	Peer, Emotional	Peer, Physical	Sexual	Adult, Emotional	Adult, Physical
1. Low Victimization	0.78 (0.09)	1.16 (0.10)	0.05 (0.02)	0.28 (0.06)	0.59 (0.09)
2. Moderate Victimization	2.73 (0.29)	2.92 (0.29)	2.38 (0.08)	2.76 (0.24)	2.51 (0.27)
3. High Victimization	3.53 (0.25)	3.50 (0.25)	4.53 (0.09)	3.13 (0.33)	3.67 (0.31)
4. High Victimization, Non-Sexual	3.45 (0.28)	3.36 (0.23)	0.11 (0.04)	4.55 (0.13)	2.96 (0.33)
5. High Peer Victimization	4.40 (0.11)	3.06 (0.21)	0.08 (0.03)	0.31 (0.11)	1.27 (0.22)

Outcomes

There were several significant associations between profiles and the aggression outcomes. Specifically, the ‘Moderate Victimization’ profile had increased risk of sexual aggression, relative to the ‘Low Victimization’ (OR: 3.72, p = .003) and ‘High Peer Victimization’ (OR: 3.76, p = .005) profiles, whereas the ‘High Victimization’ profile had increased risk of sexual aggression, relative to the ‘Low Victimization’ (OR: 8.21, p < .001) ‘High Victimization, Non-Sexual’ (OR: 4.45, p = .002), and ‘High Peer Victimization’ (OR: 8.30, p < .001) profiles. A similar pattern of results was observed for intimate partner violence, wherein the ‘Moderate Victimization’ and ‘High Victimization’ profiles had increased risk of intimate partner violence perpetration, relative to the ‘Low Victimization’ (OR: 6.69, p < .001; OR: 11.57, p < .001), ‘High Victimization, Non-Sexual’ (OR: 3.73, p = .003; OR: 6.46, p < .001), and ‘High Peer Victimization’ (OR: 3.07, p = .008; OR: 5.31, p < .001) profiles, respectively. It is worth noting that members of the ‘High Victimization’ profile had slightly increased risk of sexual aggression and intimate partner violence perpetration compared to members of the ‘Moderate Victimization’ profile, but this increased risk was non-significant. The risk of perpetration was relatively equal for members of the ‘Low Victimization,’ ‘High Victimization, Non-Sexual,’ and ‘High Peer Victimization’ profiles. A complete list of odds ratios, standard errors, and p-values for sexual aggression are presented in Table 5, while those for intimate partner violence perpetration are presented in Table 6.

Table 5.

Odds Ratios of Sexual Aggression by Profile Membership Relative to a Reference Profile.

	Odds Ratio (Standard Error)
	1	2	3	4	5
1. Low Victimization	—	0.27 (0.12)	0.12 (0.06)	0.54 (0.18)	1.01 (0.36)
2. Moderate Victimization	3.72 (1.60)**	—	0.45 (0.28)	2.02 (0.99)	3.76 (1.83)**
3. High Victimization	8.21 (4.13)***	2.21 (1.35)	—	4.45 (2.47)**	8.30 (4.62)***
4. High Victimization, Non-Sexual	1.85 (0.62)	0.50 (0.24)	0.23 (0.13)	—	1.87 (0.78)
5. High Peer Victimization	0.99 (0.35)	0.27 (0.13)	0.12 (0.07)	0.54 (0.22)	—

Note. Reference profiles are represented by the numbers listed horizontally at the top of the table (1 = ‘Low Victimization,’ 2 = ‘Moderate victimization,’ 3 = ‘High Victimization,’4 = ‘High Victimization, Non-Sexual,’ 5 = ‘High Peer Victimization’), hence the missing values on the diagonal (e.g., the odds of perpetration for members of the ‘High Victimization’ profile, relative to members the ‘Low Victimization’ profile, is listed in row 3, column 1); **p < .01, ***p < .001; p-values below .05 represent an increased probability of reporting prior sexual aggression relative to the reference profile.

Table 6.

Odds Ratios of Intimate Partner Violence Perpetration by Profile Membership Relative to a Reference Profile.

	Odds Ratio (Standard Error)
	1	2	3	4	5
1. Low Victimization	—	0.15 (0.06)	0.09 (0.04)	0.56 (0.21)	0.46 (0.15)
2. Moderate Victimization	6.69 (2.77)***	—	0.58 (0.33)	3.73 (1.76)**	3.07 (1.37)**
3. High Victimization	11.57 (5.61)***	1.73 (0.99)	—	6.46 (3.48)***	5.31 (2.74)***
4. High Victimization, Non-Sexual	1.79 (0.66)	0.27 (0.13)	0.16 (0.08)	—	0.82 (0.38)
5. High Peer Victimization	2.18 (0.73)	0.37 (0.15)	0.19 (0.10)	1.22 (0.50)	—

Note. Reference profiles are represented by the numbers listed horizontally at the top of the table (1 = ‘Low Victimization,’ 2 = ‘Moderate victimization,’ 3 = ‘High Victimization,’4 = ‘High Victimization, Non-Sexual,’ 5 = ‘High Peer Victimization’), hence missing values on the diagonal (e.g., the odds of perpetration for members of the ‘Moderate Victimization’ profile, relative to members the ‘High Victimization, Non-Sexual’ profile, is listed in row 2, column 4); **p < .01, ***p < .001; p- values below .05 represent an increased probability of reporting prior intimate partner violence perpetration relative to the reference profile.

Discussion

The current study examined profiles of childhood victimization experiences and their associations with sexual aggression and intimate partner violence perpetration in a sample of United States men. Results contribute to the extant literature by revealing common patterns of men's childhood victimization experiences, offering information about the potential impact of peer-perpetrated childhood victimization, and providing evidence for the compounded risk of childhood victimization experiences on perpetration. The results provided general support for poly-victimization—particularly when encompassing sexual victimization—as a severe risk factor for maladaptive interpersonal behaviors later in life. It is worth noting that the patterns of results are fairly distinct from those of Davis et al. (2018). Davis et al. (2018) did not find particularly strong effects for profiles encompassing sexual victimization; in fact, members of the ‘Emotional and Sexual Maltreatment’ and ‘Poly-Victimized’ profiles were not at increased risk of sexual aggression and intimate partner violence perpetration, relative to members of the ‘Low Maltreatment’ profile, which contrasted with the observed patterns in the current study. Although there were discrepancies between our results and those of Davis et al. (2018), perhaps due the studies’ inclusion of different profile indicators (e.g., peer-perpetrated victimization in the current study) and outcome variables (e.g., separation of physical and psychological intimate partner violence in Davis et al., 2018), the results of the current study were generally consistent with prior research. The results supported most hypotheses and further highlighted the utility of assessing childhood victimization to predict aggressive behavior.

It was expected that profiles characterized by high frequencies of childhood victimization experiences would be associated with increased odds of sexual aggression and intimate partner violence perpetration when compared to profiles characterized by low frequencies. This hypothesis was supported and provides further evidence for the notion that victimization experiences can have additive effects (e.g., Casey et al., 2017; Turner et al., 2017; Wolfe, 2018); thus, poly-victimization may be detrimental to future outcomes, especially when victimization experiences occur at high rates (Wolfe, 2018). In particular, the results indicated that members of the ‘High Victimization’ profile were more likely to engage in sexually aggressive behavior, and to perpetrate intimate partner violence, compared to members of the ‘Low Victimization’ profile. Additionally, members of the ‘Moderate Victimization’ profile—compared to members of the ‘Low Victimization’ profile—were more likely to perpetrate sexual aggression and intimate partner violence, respectively. Although non-significant, there also was some evidence to suggest that members of the ‘High Victimization’ profile were at slightly increased risk of perpetration relative to members of the ‘Moderate Victimization’ profile. Consistent with the implications of prior work (Wolfe, 2018), these results suggest that the frequency at which multiple forms of childhood victimization occur is an important consideration in determining risk of future perpetration, not merely exposure.

We also expected that members of profiles with high frequencies of sexual victimization would be at particularly increased risk of perpetration compared to members of profiles with low frequencies of sexual victimization. The results supported this hypothesis; evidence was provided through the comparison between members of the ‘High Victimization’ and ‘High Victimization, Non-Sexual’ profiles. With the exception of sexual victimization, these two profiles had a roughly similar collection of experiences, yet members of the ‘High Victimization’ profile were more likely to sexually aggress and perpetrate intimate partner violence when compared to members of the ‘High Victimization, Non-Sexual’ profile. Additionally, members of the ‘Moderate Victimization’ profile had slightly increased risk of sexual aggression—and significantly increased risk of intimate partner violence perpetration—relative to members of the ‘High Victimization, Non-Sexual’ profile, despite having lower frequencies of emotional and physical victimization experiences. As was found by Casey et al. (2017), these results suggest that sexual victimization may be particularly impactful as both an individual (see Table 2) and compounded risk factor for future perpetration.

Lastly, we hypothesized that profiles characterized by high frequencies of adult-perpetrated victimization (i.e., emotional and physical) would be more predictive of sexual aggression and intimate partner violence perpetration than profiles characterized by peer-perpetrated victimization. The results, generally, did not support this hypothesis. The ‘High Victimization, Non-Sexual’ profile and the ‘High Peer Victimization’ profile—which are distinguished by adult-perpetrated emotional and physical victimization—did not have differential risk of sexual aggression or intimate partner violence perpetration. This suggests that emotional and physical victimization from adults did not substantially augment the effects of peer-perpetrated victimization on future perpetration. Although this result alone could be interpreted as support for the detrimental impact of peer-perpetrated victimization, such a claim would not be supported by the current study's overall results. Members of the ‘High Peer Victimization’ exhibited only a slightly increased risk of intimate partner violence perpetration—which was non-significant—and no increased risk of sexual aggression, when compared to members of the ‘Low Victimization’ profile. This suggests that peer- and adult-perpetrated verbal and physical victimization, alike, may be limited in their effects on future perpetration. Rather, the results of the current study suggest that sexual victimization, particularly at high frequencies and in combination with emotional or physical victimization, may be most likely to produce adverse behavioral outcomes in men.

Implications

The results of the current study further highlight the importance of considering several forms of childhood victimization experiences, as they may produce additive risk for subsequent outcomes. For researchers, specifically, the results suggest that sexual victimization often co-occurs with other forms of victimization, yet this comorbidity may not represent all victimization types. Therefore, although our patterns of results indicated that sexual victimization might be particularly detrimental, it is difficult to decipher its individual impact. Researchers should accordingly practice caution when measuring sexual victimization as a sole representative of childhood victimization, because its effects may be conflated with the effects of other forms of childhood victimization.

These results also have implications for interventionists and clinicians whose aim is to reduce the prevalence of sexual aggression and intimate partner violence. For instance, the results call for the creation of screening procedures (such as those that could be implemented in schools, hospitals, or similarly frequented public facilities) that capture profiles of childhood victimization experiences, particularly those characterized by high frequencies of poly-victimization accompanied by sexual victimization. This could help pinpoint children who are at increased risk of future perpetration and improve the efficacy of targeted interventions with such children, thus mitigating the occurrence of future violence. Additionally, clinicians working with children who have experienced poly-victimization, especially if it encompassed sexual victimization, should consider addressing aggressive tendencies that may have been fostered through their lived experiences.

Limitations

The sample and measurements utilized in the current study presented some limitations that should be considered. First, participants were primarily Caucasian/White and fairly well educated. This could limit the ability to generalize results to diverse racial, ethnic, cultural, and socioeconomic groups, especially those outside of the United States. Additionally, the research included some participants who had never been in a romantic relationship. These participants were included because sexual aggression is not confined to intimate relationships, yet the inclusion of these individuals may have resulted in conservative estimates for intimate partner violence, given that some at-risk participants did not have the contextual experiences necessary to perpetrate such violence.

Regarding measurement, retrospective self-report measures are common in violence and aggression research but are inherent limitations due to the uncertainty surrounding participants’ recall accuracy and honesty of responses. An additional limitation is the measurement of adult-perpetrated victimization. These items refer to “an adult in your life” making it difficult to discern if participants are reporting within-home victimizations (e.g., by a parent or caregiver) or elsewhere. Such contexts could produce reasonably different results, but we are unable to determine how these contexts diverge to affect risk of future perpetration. Additionally, sexual victimization was not separated by peer-perpetrated and adult-perpetrated sexual victimization. This similarly limits our ability to understand situational contexts and how they may impact risk of future perpetration. Further, perceived severity of childhood victimization experiences was not measured, which may be another important consideration in determining risk of future perpetration. With regard to intimate partner violence, the utilized measure did not assess psychological abuse and thus may vary in its relationship to the observed profiles relative to other measurement tools. The final limitation is that the outcomes were coded as dichotomous variables. This was primarily done to aid the interpretation of results by providing risk comparisons between profiles, but the dichotomization also limits our ability to differentiate men who perpetrated multiple acts of violence from those who perpetrated one.

Future Directions

Although the current study provides clarifying information about childhood victimization experiences’ compounded risk on sexual aggression and intimate partner violence perpetration, future research could benefit by addressing the limitations of the current study. For example, a stratified sampling technique could allow for a diverse sample that is more representative of the population. Similarly, the inclusion of women may be useful to determine if the association between childhood victimization profiles and future perpetration is stronger for men than women, or if there is an interdependent relationship within couples as it relates to childhood victimization and intimate partner violence perpetration. Further insight could be gained by examining contextual effects of childhood victimization on future perpetration, including the perceived severity of victimization, differential effects of caregiver- versus other adult-perpetrated victimization, and how the effects of sexual victimization might differ when perpetrated by adults versus peers. It also would be informative to determine if the results of the current study are consistent when utilizing other measurement tools, including those that assess intimate partner psychological abuse and differentiate between serial and non-serial perpetration.

In addition to these adjustments, future research also could benefit from understanding how men's childhood victimization experiences may lead to later perpetration. More specifically, research examining potential mediators and moderators in the relationship between profile membership and aggressive outcomes would be informative. Integrating a profile analysis into well-established aggression models (e.g., the confluence model of sexual aggression; Malamuth et al., 1991) would serve this purpose. Additionally, it may be beneficial to observe therapy exposure or emotion regulation skills, as these could be effective protective factors. For example, Parkhill and Pickett (2016) found that sexual victimization in childhood was related to sexual aggression through impulse control difficulties; a person-centered approach assessing various forms of childhood victimization could extend these results. Future research, such as this, is crucial to further understand the etiology of sexual aggression and intimate partner violence perpetration, and to mitigate their occurrence.

Footnotes

Acknowledgments

The authors sincerely thank Dr. Matthew McLarnon for his statistical guidance, review, and comments on an early draft of this manuscript. We also thank Dr. Rollin Cook for his work on a larger project from which these data were derived. Finally, we thank the anonymous reviewers for their invaluable feedback that helped to enhance the manuscript.

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

ORCID iD

Travis N. Ray

Author Biographies

Travis N. Ray is a PhD candidate concentrating in social and health psychology within Oakland University's Department of Psychology. His research focuses on men's conceptualizations of masculinity and how it relates to violence against women, aggression toward LGBTQ + individuals, and gun violence/ownership. He also is interested in emotional components of prejudice and aggression, such as disgust and emotion regulation.

Michele R. Parkhill, PhD, is an associate professor of psychology and director of integrative studies at Oakland University. Her research is interdisciplinary in nature and scope and focuses on predictors of various types of violence including sexual violence; violence against women; violence against lesbian, gay, bisexual, transgender, and queer/questioning individuals; gun violence; and police violence against minority populations.

References

Abbey

Helmers

B. R.

Jilani

Mcdaniel

M. C.

Benbouriche

(2021). Assessment of men’s sexual aggression against women: An experimental comparison of three versions of the sexual experiences survey. Psychology of Violence, 11(3), 253–263. https://doi.org/10.1037/vio0000378

Abbey

Jacques-Tiura

A. J.

LeBreton

J. M.

(2011). Risk factors for sexual aggression in young men: An expansion of the confluence model. Aggressive Behavior, 37(5), 450–464. https://doi.org/10.1002/ab.20399

Abbey

Parkhill

M. R.

BeShears

Clinton-Sherrod

A. M.

Zawacki

(2006). Cross-sectional predictors of sexual assault perpetration in a community sample of single African American and Caucasian men. Aggressive Behavior, 32(1), 54–67. https://doi.org/10.1002/ab.20107

Abbey

Parkhill

M. R.

Koss

M. P.

(2005). The effects of frame of reference on responses to questions about sexual assault victimization and perpetration. Psychology of Women Quarterly, 29(4), 364–373. https://doi.org/10.1111/j.1471-6402.2005.00236.x

Brassard

Darveau

Péloquin

Lussier

Shaver

P. R.

(2014). Childhood sexual abuse and intimate partner violence in a clinical sample of men: The mediating roles of adult attachment and anger management. Journal of Aggression, Maltreatment & Trauma, 23(7), 683–704. https://doi.org/10/1080/10926771.2014.933464

Brennan

C. L.

Borgman

R. A.

Watts

S. S.

Wilson

R. A.

Swartout

K. M.

(2020). Childhood neglect history, depressive symptoms, and intimate partner violence perpetration by college students. Journal of Interpersonal Violence, 36(23-24): NP12576–NP12599. https://doi.org/10.1177/0886260519900307

Buhrmester

Kwang

Gosling

S. D.

(2011). Amazon’s mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6(1), 3–5. https://doi.org/10.1037/14805-009

Casey

E. A.

Beadnell

Lindhorst

T. P.

(2009). Predictors of sexually coercive behavior in a nationally representative sample of adolescent males. Journal of Interpersonal Violence, 24(7), 1129–1147. https://doi.org/10.1177/0886260508322198

Casey

E. A.

Masters

N. T.

Beadnell

Hoppe

M. J.

Morrison

D. M.

Wells

E. A.

(2017). Predicting sexual assault perpetration among heterosexually active young men. Violence Against Women, 23(1), 3–27. https://doi.org/10.1177/1077801216634467

10.

Casler

Bickel

Hackett

(2013). Separate but equal? A comparison of participants and data gathered via Amazon’s MTurk, social media, and face-to-face behavioral testing. Computers in Human Behavior, 29(6), 2156–2160. https://doi.org/10.1016/j.chb.2013.05.009

11.

Chandler

J. J.

Paolacci

(2017). Lie for a dime: When most prescreening responses are honest but most study participants are impostors. Social Psychological and Personality Science, 8(5), 500–508. https://doi.org/10.1177/1948550617698203

12.

Daigneault

Hébert

McDuff

(2009). Men’s and women’s childhood sexual abuse and victimization in adult partner relationships: A study of risk factors. Child Abuse & Neglect, 33(9), 638–647. https://doi.org/10.1016/j.chiabu.2009.04.003

13.

Davis

K. C.

Masters

N. T.

Casey

Kajumulo

K. F.

Norris

George

W. H.

(2018). How childhood maltreatment profiles of male victims predict adult perpetration and psychosocial functioning. Journal of Interpersonal Violence, 33(6), 915–937. https://doi.org/10.1177/0886260515613345

14.

Davis

K. C.

Schraufnagel

T. J.

Jacques-Tiura

A. J.

Norris

George

W. H.

Kiekel

P. A.

(2012). Childhood sexual abuse and acute alcohol effects on men’s sexual aggression intentions. Psychology of Violence, 2(2), 179–193. https://doi.org/10.1037/a0027185

15.

Desmarais

S. L.

Reeves

K. A.

Nicholls

T. L.

Telford

R. P.

Fiebert

M. S.

(2012). Prevalence of physical violence in intimate relationships, part 2: Rates of male and female perpetration. Partner Abuse, 3(2), 170–198. https://doi.org/10.1891/1946-6560.3.2.170

16.

Fang

Corso

P. S.

(2007). Child maltreatment, youth violence, and intimate partner violence: Developmental relationships. American Journal of Preventive Medicine, 33(4), 281–290. https://doi.org/10.1016/j.amepre.2007.06.003

17.

Finkelhor

Hamby

S. L.

Ormrod

Turner

(2005). The juvenile victimization questionnaire: Reliability, validity, and national norms. Child Abuse & Neglect, 29(4), 383–412. https://doi.org/10.1016/j.chiabu.2004.11.001

18.

Finkelhor

Turner

H. A.

Shattuck

Hamby

S. L.

(2015). Prevalence of childhood exposure to violence, crime, and abuse: Results from the national survey of children’s exposure to violence. JAMA Pediatrics, 169(8), 746–754. https://doi.org/10.1001/jamapediatrics.2015.0676

19.

Godbout

Vaillancourt-Morel

M. P.

Bigras

Briere

Hébert

Runtz

Sabourin

(2019). Intimate partner violence in male survivors of child maltreatment: A meta-analysis. Trauma, Violence, & Abuse, 20(1), 99–113. https://doi.org/10.1177/1524838017692382

20.

Hamby

Finkelhor

Turner

Kracke

(2011). The Juvenile Victimization Questionnaire toolkit. http://www.unh.edu/ccrc/jvq/index_new.html.

21.

Hamby

Turner

(2013). Measuring teen dating violence in males and females: Insights from the national survey of children’s exposure to violence. Psychology of Violence, 3(4), 323–339. https://doi.org/10.1037/a0029706

22.

Hoffman

C. Y.

Phillips

M. D.

Daigle

L. E.

Turner

M. G.

(2017). Adult consequences of bully victimization: Are children or adolescents more vulnerable to the victimization experience? Youth Violence and Juvenile Justice, 15(4), 441–464. https://doi.org/10.1177/1541204016650004

23.

Khan

McCormack

H. C.

Bolger

E. A.

McGreenery

C. E.

Vitaliano

Polcari

Teicher

M. H.

(2015). Childhood maltreatment, depression, and suicidal ideation: Critical importance of parental and peer emotional abuse during developmental sensitive periods in males and females. Frontiers in Psychiatry, 6(42), 42–72. https://doi.org/10.3389/fpsyt.2015.00042

24.

Koss

M. P.

Gidycz

C. A.

Wisniewski

(1987). The scope of rape: Incidence and prevalence of sexual aggression and victimization in a national sample of higher education students. Journal of Consulting and Clinical Psychology, 55(2), 162–170. https://doi.org/10.1037/0022-006X.55.2.162

25.

Malamuth

N. M.

Hald

G. M.

(2016). The confluence mediational model of sexual aggression. In Boer

D. P.

(Ed.), The Wiley handbook on the theories, assessment and treatment of sexual offending (pp. 53–71). John Wiley & Sons.

26.

Malamuth

N. M.

Sockloskie

R. J.

Koss

M. P.

Tanaka

J. S.

(1991). Characteristics of aggressors against women: Testing a model using a national sample of college students. Journal of Consulting and Clinical Psychology, 59(5), 670–681. https://doi.org/10/1037/0022-006X.59.5.670

27.

McLarnon

M. J.

O’Neill

T. A.

(2018). Extensions of auxiliary variable approaches for the investigation of mediation, moderation, and conditional effects in mixture models. Organizational Research Methods, 21(4), 955–982. https://doi.org/10.1177/1094428118770731

28.

Muthén

L. K.

Muthén

B. O.

(2012). MPlus: Statistical analysis with latent variables—User’s guide.

29.

Parkhill

M. R.

Pickett

S. M.

(2016). Difficulties in emotion regulation as a mediator of the relationship between child sexual abuse victimization and sexual aggression perpetration in male college students. Journal of Child Sexual Abuse, 25(6), 674–685. https://doi.org/10.1080/10538712.2016.1205161

30.

Ray

T. N.

Parkhill

M. R.

Cook

R. D.

(2021). Bullying, masculinity, and gun-supportive attitudes among men: A path analysis testing the structural relationships between variables. Psychology of Violence, 11(4), 395–404. https://doi.org/10.1037/vio0000370

31.

Reijntjes

Kamphuis

J. H.

Prinzie

Telch

M. J.

(2010). Peer victimization and internalizing problems in children: A meta-analysis of longitudinal studies. Child Abuse & Neglect, 34(4), 244–252. https://doi.org/10.1016/j.chiabu.2009.07.009

32.

Richards

T. N.

Tillyer

M. S.

Wright

E. M.

(2017). Intimate partner violence and the overlap of perpetration and victimization: Considering the influence of physical, sexual, and emotional abuse in childhood. Child Abuse & Neglect, 67, 240–248. https://doi.org/10.1016/j.chiabu.2017.02.037

33.

Richards

T. N.

Tomsich

Gover

A. R.

Jennings

W. J.

(2016). The cycle of violence revisited: Distinguishing intimate partner violence offenders only, victims only, and victim-offenders. Violence and Victims, 31(4), 573–590. https://doi.org/10/1891/0886-6708.VV-D-15-00056

34.

Singh

Tolman

Walton

Chermack

Cunningham

(2014). Characteristics of men who perpetrate intimate partner violence. The Journal of the American Board of Family Medicine, 27(5), 661–668. https://doi.org/10.3122/jabfm.2014.05.130247

35.

Smith-Marek

E. N.

Cafferky

Dharnidharka

Mallory

A. B.

Dominguez

High

Mendez

(2015). Effects of childhood experiences of family violence on adult partner violence: A meta-analytic review. Journal of Family Theory & Review, 7(4), 498–519. https://doi.org/10.1111/jftr.12113

36.

Stith

S. M.

Rosen

K. H.

Middleton

K. A.

Busch

A. L.

Lundeberg

Carlton

R. P.

(2000). The intergenerational transmission of spouse abuse: A meta-analysis. Journal of Marriage and Family, 62(3), 640–654. https://doi.org/10.1111/j.1741-3737.2000.00640.x

37.

Turner

H. A.

Shattuck

Finkelhor

Hamby

(2017). Effects of poly-victimization on adolescent social support, self-concept, and psychological distress. Journal of Interpersonal Violence, 32(5), 755–780. https://doi.org/10/1177/0886260515586376

38.

Turner

M. G.

Phillips

M. D.

Tigri

H. B.

Williams

M. A.

Hartman

J. L.

(2016). On the association between repeat bully victimizations and carrying a firearm: Evidence in a national sample. International Journal of Offender Therapy and Comparative Criminology, 60(8), 871–896. https://doi.org/10.1177/0306624X15573547

39.

White

J. W.

Smith

P. H.

(2004). Sexual assault perpetration and reperpetration: From adolescence to young adulthood. Criminal Justice and Behavior, 31(2), 182–202. https://doi.org/10.1177/0093854803261342

40.

Wolfe

D. A.

(2018). Why polyvictimization matters. Journal of Interpersonal Violence, 33(5), 832–837. https://doi.org/10.117/0886260517752