Violence, Victimization, Criminal Justice Involvement, and Substance Use Among Drug-Involved Men

This research identified three subgroups of drug-involved men based on severity of self-reported violence perpetration against intimate or nonintimate partners among a sample of 148 men selected from a subsample of participants in the Kentucky National Institute on Drug Abuse (NIDA) AIDS Cooperative Agreement. Men in the No Violence group accounted for approximately 19% of the total respondents, men in the Moderate Violence Severity and Extreme Violence Severity groups comprises 56% and 25% of the sample, respectively. Men in the Extreme Violence Severity group experienced significantly more psychological victimization as children and more frequent physical childhood abuse than did their peers. Men in the Extreme Violence Severity group reported having earlier involvement in the criminal justice system and lawbreaking behavior; they also reported higher frequency of marijuana and crack use. Implications for treatment and future research are presented.

Keywords: violence; men; substance use; victimization; criminal justice involvement

Considerable attention has been given to the role of substance use in violence against women and it has been identified as a significant factor in men's perpetration of vio- lence against intimate female partners (El-Bassel, Gilbert, Wu, Chang, & Fontdevila, 2007). Estimates indicate that up to 70% of male substance users have perpetrated some form of violence against their intimate partners (Bennett & Williams, 2003; El-Bassel et al., 2007). These data underscore the importance of targeting drug-involved men in prevention and intervention efforts aimed at reducing and eliminating violence against women.

Although research examining men's violence has typically focused on either violence toward intimate partner violence (IPV) or violence toward nonintimates, there exists evidence to suggest that there is an association between the use of violence within intimate relationships and outside, with nonintimates (Fagan & Brown, 1994; Holtzworth-Munroe, 2000; Holtzworth-Munroe & Meehan, 2004; Holtzworth-Munroe & Stuart, 1994; Lawson et al., 2003). The identification of patterns of violence perpetration across any type of victim is critical in providing a greater understanding of this behavior and aiding in the development of targeted prevention and intervention strategies for violent men.

For drug-involved men in particular, violence perpetration (whether it is against an inti- mate or a nonintimate) may be part of a larger pattern of criminal behavior. There appears to be a reciprocal relationship among substance use, particularly illicit substances, and crime such that greater involvement in one maybe associated with greater involvement in the other (Inciardi & Pottieger, 1994; White & Gorman, 2000). As such, men who use illicit drugs often have a history of lawbreaking and involvement in the criminal justice (CJ) sys- tem and, correspondingly, criminal offenders frequently have histories of illicit drug use. Not surprisingly, it is estimated that approximately 80% of male jail and prison inmates have at least one substance use problem (Feucht & Gfroerer, 2011; Sabol & Couture, 2008).

Notwithstanding these findings, the relationship between violence and substance use is "complex"; there appears to be various individual, environmental, and contextual factors that influence the potential for violence when drugs are being used (for a comprehensive review of these issues, see Boles & Miotto, 2003). In particular, certain substances (e.g., alcohol and crack cocaine) have been found to be more closely associated with violent behavior than others because of their psychopharmacological effects on users, the social process of ingestion and/or distribution, and/or the means necessary to obtain the substances (either the substance itself or money to purchase the substance; Boles & Miotto, 2003; Goldstein, 1985; Inciardi & Pottieger, 1994).

Related to this, there is ample evidence that childhood victimization is a significant risk factor for adult involvement in violence and crime (Maxfield & Widom, 1996; Smith & Thornberry, 1995; Stouthamer-Loeber, Loeber, Homish, & Wei, 2001; Widom, Marmorstein, & White, 2006; Zingraff, Leiter, Myers, & Johnson, 1993). In fact, Widom et al. (2006) found a direct relationship between childhood victimization and arrest for violent crimes in adult- hood. Childhood abuse and neglect increases the likelihood of arrest as a juvenile by 59%, as an adult by 28%, and for a violent crime by 30%, regardless of gender (Widom & Ames, 1994). Not surprisingly, a history of childhood victimization and neglect is common among criminal offenders (Weeks & Widom, 1998). The existing data show that 4.1%-14.3% of men in the CJ system have experienced some form of childhood sexual violence and as many as 64.5% report childhood physical abuse (Harlow, 1999; McClellan, Farabee, & Crouch, 1997; Weeks & Widom, 1998; Wolff & Shi, 2012). In comparison, among a random sample of U.S. men, 1.9% report experiencing childhood sexual victimization whereas 53.8% report the experience of child physical victimization (Tjaden & Thoennes, 2000). Interestingly, for men, unlike their female counterparts, childhood victimization does not appear to be a risk factor for subsequent adult substance use (illicit drug and/or alcohol).1 However, childhood victimization does appear to increase their risk for experiencing problems (e.g., CJ involve- ment) as a result of any substance use (Widom, Ireland, & Glynn, 1995).

In comparison to childhood victimization, much less is known about men's own experiences of victimization in adulthood (IPV and/or nonintimate partner victimization) and the research that does exist is primarily focused on IPV. These relatively limited findings indicate that childhood victimization, specifically childhood physical abuse, is the strongest risk factor for subsequent IPV victimization among men (Coker et al., 2002). Similarly, the experience of IPV victimization significantly increases the odds of men experiencing a substance use disorder (Afifi et al., 2009; Nowinski & Bowen, 2012).

Taken together, this research underscores the intersection of violence, victimization, lawbreaking, and substance use among drug-involved men. Although prior research has examined subsets of these issues, there is a dearth of research examining them in unison; a more comprehensive approach is necessary to better inform practitioners and policy makers working with this population. Similarly, prior research has established the value of examining different patterns or subgroups of individuals based on the level of violence perpetration (e.g., Logan & Leukefeld, 2000; Logan, Walker, & Leukefeld, 2001). Therefore, to address this gap and advance our understanding of the intersection of violence, victimization, lawbreaking, and substance use among drug-involved men, this study sought to meet the following research aims: (a) identify subgroups of drug-involved men based on the severity of violence perpetration, and (b) examine between group differences across the domains of violence perpetration (IPV; nonintimate victims), victimization (childhood; IPV; nonintimate), CJ involvement and lawbreaking, and substance use. The identifica- tion of subgroups provides a unique opportunity to examine heterogeneity of violence perpetration among drug-involved men. Similarly, examining between group differences will illuminate how distinctions in levels of violence perpetration may influence and be influenced by behavior and functioning in other critical domains. This strategy will provide a greater understanding of the diversity that exists among drug-involved men and will thus enhance the ability to develop more tailored intervention strategies for this population.

METHOD

Participants and Procedures

Respondents were selected from a target sample of men (N 5 875) who participated in the Kentucky NIDA AIDS Cooperative Agreement study. The NIDA Cooperative Agreement study focused on out-of-treatment current drug abusers recruited to participate in a community-based HIV prevention intervention study. The sample used for the research presented here was composed of men (N 5 148) who met the following criteria: (a) self- identified as heterosexual or bisexual, or if they identified as gay men, they reported having sex with partners of the opposite sex in the 30 days before entry into the NIDA AIDS Cooperative Agreement study2; and (b) reported crack cocaine use in the NIDA AIDS Cooperative Agreement study. Data for this study were collected by trained interviewers between February 1999 and August 2002, approximately five years after participation in the NIDA AIDS Cooperative Agreement study. Interviews lasted approximately 2 hr; men were paid $75 for participating. Of the original study's participants who met eligibility criteria for this study, no significant differences in select demographic (e.g., age, educa- tion level, marital status), drug use (e.g., number of days used alcohol, marijuana, crack in the past 30 days), or sexual behavior (e.g., number of sex partners and days had sex in the past 30 days, number of days had sex exchange encounters) variables were found between individuals who participated and those who did not participate in this study.

Measures

Measures were used to (a) identify subgroups of men based on the self-reported severity of violence perpetrated against any type of victim and (b) compare the identified subgroups across indicators of sociodemographic characteristics, cumulative violence and victimiza- tion, CJ involvement and lawbreaking, and substance use.

Violence Perpetration Severity Subgroups. Following procedures established by Logan and Leukefeld (2000), subgroups were developed by calculating total self-reported violence severity perpetrated against any type of victim (intimate partner or nonintimate) by the respondents in their lifetime. First, total violence severity was estimated by calculating the weighted sum of 16 questions assessing various types of violence perpetration.3 Questions were adapted from the Conflict Tactics Scale, the National Violence Against Women Survey (NVAWS; Straus, Hamby, & Warren, 2003; Tjaden & Thoennes, 2000) as well as research by Logan and colleagues (Logan & Leukefeld, 2000; Logan et al., 2001; Table 1 provides a list of all the items). Guidelines for weighting the items were taken from Straus et al., 2003 (mean for total violence severity 5 7.689; SD 5 7.827; observed range 5 0-35).4 The mean score for total violence perpetration severity for the sample as a whole was 7.689 (SD 5 7.827; observed range 5 0-35). Based on the established protocol for identifying subgroups, three groups were identified. Participants who indicated no lifetime violence were assigned to the No Violence group (approximately 19% of the total respondents); men at the 75th percentile or higher were assigned to the Extreme Violence Severity group (25% of the total respondents); and those that reported violence, but who fell lower than the 75th percentile, were classified into the Moderate Violence Severity group (56% of the total respondents). Table 1 displays the specific violent acts reported by each violence perpetration severity group. Larger proportions of men in the Extreme Violence Severity group self-reported perpetrating more of the specific violent acts than men assigned to the Moderate Violence Severity group. The mean score for total violence perpetration severity for the Extreme Violence Severity group was 5.925 (SD 5 .974) and 3.229 (SD 5 .354) for the Moderate Violence Severity group.

Demographic Characteristics. Five demographic variables (age, race/ethnicity, i ntimate partner status, education level, and current employment) were examined. Age was measured in years. Three categories were used to describe the race/ethnicity of the participants: Black non-Hispanic, White non-Hispanic, and American Indian. Dichotomous variables (yes 5 1; no 5 0) were used to measure the next two variables. Intimate partner status assessed whether a respondent reported being married or cohabitating with a sexual partner at the time of the interview; current employment indicated whether a respondent reported being employed part-time or more at the time of the interview. Finally, education level specified whether the respondent had less than a high school diploma/general equivalency diploma (GED), had earned a high school diploma or GED, or more education than a high school diploma/GED.

Physical Violence. Four variables were computed to examine the perpetration of physical violence against intimate and nonintimate partners, respectively. Two dichoto- mous variables (yes 5 1; no 5 0) were created assessing whether a respondent had ever perpetrated any of five acts of physical violence (Items 9-13 in Table 1) against an intimate or nonintimate partner, respectively. Two additional variables assessed the frequency of perpetrating these same acts of physical violence against the corresponding victim type (observed range, IPV 5 0-138; observed range, nonintimate 5 0-1,040).

Victimization

Childhood Victimization. Eight variables were used to measure childhood victimization. Psychological, physical, and sexual victimization, in addition to witnessing the physical and/ or sexual abuse of one's mother, were assessed. Two variables, one dichotomous and the other reflecting the number of different abusive acts experienced, respectively, were created for each type of victimization (and witnessing). Eight questions assessed potentially psychologically abusive childhood experiences (e.g., "insulted, shamed, or humiliated you in front of oth- ers"). Physical abuse was assessed by four questions similar to Items 10 through 13 in Table 1 (the stem was changed to reflect the victimization of the respondent as a child by his parents and/or other caretakers). Sexual victimization was measured by items similar to 14 through 16 in Table 1 (stems changed accordingly); witnessing was measured by two variables reflect- ing whether a respondent reported witnessing the physical and/or sexual abuse of his mother. An affirmative response to any specific question was treated as an indication of that type of victimization (yes 5 1; no 5 0). Affirmative responses to questions were summed to arrive at the number of different abusive acts experienced (possible ranges: psychological abuse, 0-8; physical abuse, 0-4; sexual abuse, 0-3; witnessing. 0-2).

Adult Victimization. As with childhood victimization, eight total variables were created to measure physical victimization of the respondent by intimate and nonintimate partners, respectively (behaviors are the same as those measured by items 10-13 in Table 1; the stem was changed to ask whether the respondent had been victimized in this way by an inti- mate or nonintimate partner). Dichotomous variables (yes 5 1; no 5 0) measured whether the man had experienced any type of physical victimization by either an intimate and/or nonintimate partner, respectively, and frequency variables reflected the number of times the man reported experiencing the particular type of violence.

Criminal Justice Involvement and Lawbreaking. CJ involvement was operationalized by three variables. These included whether or not a participant reported having been in juvenile detention (yes 5 1; no 5 0), their age at first adult arrest (provided in years), the number of lifetime arrests,5 and an indicator reflecting the frequency of lawbreaking behav- ior. Engagement in lawbreaking was examined by an additional set of variables. Frequency of lawbreaking behavior was calculated by summing affirmative responses to 14 questions (yes 5 1; no 5 0) that assessed engagement in different types of criminal behavior. In addi- tion, the 14 individual items were examined separately to provide a greater depth of infor- mation regarding the types of lawbreaking behaviors in which individuals were engaged.

Substance Use. Alcohol, marijuana, and crack were selected as measures of substance use because these were the most frequently used substances within this sample. Each type of substance was measured by a single variable that assessed a respondent's pattern of use over the past 24 months. Patterns of alcohol, marijuana, and crack use were calculated based on data obtained through the life history calendar method of collecting event histo- ries. Respondents were asked to first select the response option that best characterized their frequency of use (1-2 times per month; 1-2 times per week; 3-5 times per week; almost every day). Following this, they were then asked to characterize how intoxicated they usu- ally became during use (never to extreme intoxication; sometimes to extreme intoxication; usually to extreme intoxication). Responses for pattern of substance use were then coded 1 (1-2 times per month, never to extreme intoxication) to 12 (almost every day, usually to extreme intoxication); individuals who reported not using a given substance were assigned as 0. Age at first use for each substance was reported in years.

Data Analysis

Several statistical techniques were used to examine the differences among the violence perpetration severity groups. Comparisons between groups were conducted using chi- square analyses (for categorical level variables) and analyses of variance (ANOVA; for variables that were interval level or higher). Following significant ANOVA F tests, planned comparisons were conducted using the Least Significant Difference test (LSD).

RESULTS

Table 1 provides the prevalence of specific violent behaviors reported by each of the violence perpetration severity groups. Between group differences as well as means or per- centages and standard deviations for the variables reflecting sociodemographics, violence perpetration, and victimization are presented in Table 2; data for indicators of CJ involve- ment, lawbreaking, and substance use are provided in Table 3.

Descriptive Findings

On average, the men in this sample were 43 years of age and most were African American (84.5%). Approximately 18.0% of the men reported being married or living with their female partners, more than half reported being employed, and almost 68.0% had a high school diploma/GED or more education.

As a whole, about 37% of the sample reported perpetrating at least one act of physical IPV and more than half of the men reported engaging in physical violence against a nonin- timate. Victimization of the participants themselves appeared common as well. Thirty-nine percent of the men acknowledged being physically abused as children, whereas 26% and 68% of the men, respectively, reported experiencing physical violence from an intimate or nonintimate partner as an adult.

Overall, the participants appeared to have consistent and significant involvement in the CJ system and lawbreaking. More than a third of the sample reported serving time in juvenile detention, experiencing their first adult arrest by the age of 23 years, having an average of 22 lifetime arrests, and engaging in six lawbreaking behaviors. Respondents reported participating in a variety of specific lawbreaking behaviors; driving while intoxi- cated, stealing (as well as receiving, buying, and selling stolen goods), damaging property, and selling, distributing, and making illegal drugs were the most common lawbreaking behaviors. Alcohol was the most frequently used substance (mean 5 4.807; SD 5 4.142) followed by crack (mean 5 3.291; SD 5 3.747) and marijuana (mean 5 1.905; SD 5 3.157). Post-hoc examination indicated that the correlations among all the substance use variables were significant (p # .01) and moderate in strength, suggesting some overlap in use of these substances. The highest correlation among these variables was between alco- hol and crack, the substances most commonly associated with violence (data not included in Table 3; correlations: alcohol and crack 5 .481; alcohol and marijuana 5 .257; crack and marijuana 5 .352; Boles & Miotto, 2003).

Between Group Differences

No between group differences were identified for any of the sociodemographic character- istics. However, the violence perpetration severity groups were differentiated by distinct patterns of violence, victimization, CJ involvement and lawbreaking, as well as substance use. Not surprisingly, given the grouping criteria, men in the Extreme Violence Severity group had significantly higher rates of physical violence perpetration against both intimate and nonintimate partners than men in the Moderate Violence Severity group. Almost twice as many men in the Extreme Violence Severity group reported engaging in physical IPV as compared to men in the Moderate Violence Severity group (65% compared to 36%) and almost all the men in the Extreme Violence Severity group (97%) reported physical violence against a nonintimate.

Similar trends were observed across the indicators of victimization. Men in the Extreme Violence Severity group reported experiencing significantly more psychological victim- ization as children and more frequent physical childhood abuse than did their peers in either the No Violence or Moderate Violence Severity groups. Almost half of the men in the Extreme Violence Severity group reported experiencing physical IPV and 89% reported being physically abused by a nonintimate.

The groups were further distinguished by significant differences across all the major indicators of CJ involvement and lawbreaking as well as substance use. As with violence and victimization, there was a pattern of increasing prevalence of risk across the groups. Men in the Extreme Violence Severity group reported having earlier involvement in the CJ system and lawbreaking as evidenced by indicators assessing whether they had been in juvenile detention, age of first arrest, and frequency of lawbreaking. In regard to substance use, significant differences were observed for the pattern of marijuana and crack use; men in the Extreme Violence Severity group reported higher frequency of use for both of these substances.

DISCUSSION

This is the first known study to examine the heterogeneity of violence perpetration (IPV and nonintimate partner) among drug-involved men residing in the community. Based on established protocols (Logan & Leukefeld, 2000), three subgroups were identified: No Violence, Moderate Violence Severity, and Extreme Violence Severity. These groups dem- onstrate differential relationships with respect to violence perpetration, victimization, CJ involvement and lawbreaking, and substance use. Findings have implications for policy and practice with drug-involved men, particularly in the CJ arena, as well as for future research.

Although drug-involved men may be a high risk group for violence perpetration gener- ally, the identification of subgroups demonstrated that there is significant variation in use of violence. In fact, almost one-fifth of the men in this sample reported no acts of vio- lence perpetration in their lives against any type of victim. Notwithstanding this, physical violence against a nonintimate partner was the most common and frequently engaged in type of violence measured. Almost all the men in the Extreme Violence Severity group and more than half the men in the Moderate Violence Severity group reported engaging in physical aggression against a nonintimate. Although not as prevalent, violence against an intimate partner was still pervasive. About 65% of the men in the Extreme Violence Severity group and 36% of men in the Moderate Violence Severity group report physical IPV. Taken together, data regarding nonintimate and IPV suggest that there is a relation- ship between the use of physical violence across victim type, particularly for the most violent subgroup of men. Furthermore, these findings strongly suggest that it is a small percentage of drug-involved men who participate in the most severe forms of violence and the highest frequency of violent acts. These suggestions, however, are tempered by the need for further research to better understand the specific nature of the violence (i.e., type of physical violence; level of lethality) that is used, the specific contexts in which it occurs (e.g., in response to a perceived threat; concurrent with substance use), and how these fac- tors vary as a function of victim characteristics (i.e., intimate partner; nonintimate partner).

The examination of between group differences demonstrates how distinctions in levels of violence perpetration affected behavior and functioning in the areas of victimization, CJ involvement and lawbreaking, and substance use. To begin with, in regard to their own experiences of victimization, the findings document that violent victimization appears to be relatively common among drug-involved men, generally, and that it varies as a function of an individual's own level of violence perpetration. Notwithstanding this finding, it is notable that data from the sample as a whole, and by subgroups, indicate that levels of childhood physical victimization fall below those found in the general population (NVAWS; Tjaden & Thoennes, 2000). Data from the NVAWS indicate that almost 54% of men report being physically assaulted by an adult caretaker as children; in contrast, 39% of the sample as a whole and 51% of the Extreme Violence Severity group, the subgroup reporting the highest rate of childhood physical victimization, reported this type of abuse (Tjaden & Thoennes, 2000). This finding is distinct from those found among similar samples of drug-involved women, who typically report levels of childhood physical abuse higher than the general population (e.g., Golder & Logan, 2010, 2011). In contrast, rates of childhood sexual abuse in this study were generally higher than the national average; for the sample as a whole, approximately 3.4% of the men reported being victimized sexually as children compared to approximately 2% of men nationally (Tjaden & Thoennes, 2000). Perhaps, contrary to expectations, men in the No Violence group reported the highest levels of child- hood sexual assault (7.1%), followed by men in the Extreme Violence Severity (5.4%) and Moderate Violence Severity (1.2%) groups, respectively. The relationship between childhood sexual victimization and adult violence needs to be explored further.

In regards to IPV, 37.0% of the sample as a whole reported experiencing physical victimization by an intimate partner sometime in their lives. In comparison, data from the NVAWS analyzed by Coker and colleagues (2002) found that 23.3% of men reported a lifetime experience of any type of IPV (physical, sexual, and/or psychological), whereas 6.0% reported experiencing physical IPV alone. Similarly, in a comprehensive review of the literature on partner violence against men, estimates of the prevalence of any IPV var- ied between 0.6% and 32.0% among heterosexual men (Nowinski & Bowen, 2012). Thus, as a whole, men in this sample report experiencing physical IPV at rates higher than those found in other samples. Examining the subgroups, we see that men in the No Violence group reported rates of physical IPV somewhat lower than national averages, whereas the other groups had rates substantially higher than those found in the NVAWS for physical IPV alone (approximately 24.0% and 49.0%, respectively, among the Moderate Violence Severity and Extreme Violence Severity groups). These findings are consistent with prior research that finds childhood physical abuse to be the strongest risk factor for IPV among men (Coker et al., 2002).

The findings regarding IPV perpetration and victimization underscore the dyadic con- text of partner violence and suggest that both partners, male and female, may be simul- taneously perpetrating violence as well as being victimized. In fact, research has found that women report similar, and potentially higher levels of physical violence perpetration, than men (Archer, 2000; Morse, 1995; Straus, 1999; Straus & Gelles, 1986). Swan and Snow (2002) conducted novel research that empirically identified typologies of women's violence in intimate relationships; the identified typologies were classified as victim type (where the woman's partner committed more severe violence and coercion against her), aggressor type (where the woman committed more violence or more severe violence than her partner), or mixed type (where one partner was more violent than the other but the other partner exhibiting more controlling behaviors). This research highlighted the fact that even when women were the aggressors in a relationship, they usually experienced significant violence from their male partners. Results from Swan and Snow as well as this study underscore the importance of examining male violence against their intimate partners within the context of the dyadic relationship; such research may be necessary for gaining a more complete and dynamic understanding of IPV that ultimately leads to more effective prevention and intervention strategies (Swan & Snow, 2002).

Finally, in regard to nonintimate physical victimization, for the entire sample and for the subgroups that perpetrated violence, the percentage of men who reported being physi- cally victimized appeared elevated in comparison to the general population. For example, data from NVAWS analyzed by Desai, Arias, Thompson, and Basile (2002) indicated that 45% of men in the general population report ever being physically victimized in their lives by any type of aggressor. As a whole, the findings regarding violence and victimization indicate that increased levels of violence perpetration are associated with increased levels of physical victimization across the lifespan.

Continuing this trend, CJ involvement and lawbreaking, as well as substance use, were highest among men in the Extreme Violence Severity group. Specifically, men in the Extreme Violence Severity group reported significantly more and earlier engagement across their lives in the CJ system than men in either of the other subgroups. This group also had the greatest involvement in individual lawbreaking behaviors, involvement in a wider variety of lawbreaking behaviors, and exhibited higher levels of marijuana and crack use than men in the other groups. These results appear to be consistent with research that finds that childhood victimization is associated with increased engagement in the CJ system (Weeks & Widom, 1998; Widom et al., 2006) and that IPV is associated with sub- stance use (Coker et al., 2002).

Perhaps, somewhat unexpectedly, given their different victimization profiles, men in the No Violence and Moderate Violence Severity groups had very similar levels of illicit drug use (all three groups had statistically equivalent levels of alcohol consumption). This finding suggests that there may be a threshold operating such that only higher levels of victimization produce the highest levels of substance use within this sample of drug- involved men. This assertion is consistent with the "saturation effect" hypothesis offered by Widom and colleagues (Widom et al., 1995; Widom et al., 2006) to explain the relation- ship between childhood victimization and subsequent adult substance use in men; how- ever, further research is needed to fully explore this issue. Aside from these similarities in substance use, the No Violence and Moderate Violence Severity groups differed in regard to CJ involvement. Men in the Moderate Violence Severity group had greater involvement in the CJ system and engaged in a wider variety of lawbreaking behaviors than did men in the No Violence group.

The identification of patterns of violence perpetration across any type of victim and sub- sequent differences between the subgroups is instructive in the development of targeted pre- vention and intervention strategies that seek to reduce violence, justice system involvement (and associated lawbreaking), as well as substance use among drug-involved men. Evidence- based principles of practice indicate that effective interventions must target criminogenic needs, the potentially malleable factors that are most closely associated with continued involvement in the justice system (for a review, see Golder et al., 2005). Thus, substance use appears to be a risk factor (i.e., criminogenic need) that should be targeted for intervention to reduce further involvement in the CJ system and engagement in lawbreaking.

As such, there is a great deal of research identifying the attributes of effective pro- gramming for drug- and crime-involved individuals. For incarcerated men, research has demonstrated the effectiveness of the continuum of care model for drug-involved offenders from prison-based therapeutic communities (TCs) to community-based TCs, and finally outpatient, community-based follow-up services (Golder et al., 2005). More broadly, for men residing in the community (e.g., on probation or parole), effective interventions and programming are characterized by the following attributes:

* Cognitive-behavioral models of treatment that focus on social and cognitive skill training (Golder et al., 2005; Pearson & Lipton, 1999; Pearson, Lipton, Cleland, & Yee, 2002). In particular, interventions that contain components that target criminal thinking "errors," teach self-monitoring exercises (e.g., journals, "thought logs"), and provide direct feedback from staff and other participants that identifies criminal thinking and related behaviors are effective (Field, 1986; Center for Substance Abuse Treatment, 2005; Wanberg & Milkman, 1998).

* Regular, timely monitoring and close supervision in the form of weekly urinalysis and the provision of clinical progress reports to the supervising authorities. In addition, the use of clear and immediate consequences/feedback for behavior. For example, rewarding compliant behav- ior with reductions in reporting requirements and penalizing negative behavior with program termination and potential legal sanctions (Marlowe, 2003).

* Treatment of sufficient duration and intensity to exert a measurable effect on behavior. Research demonstrates that increased time in substance abuse treatment is associated with significant reductions in recidivism (Lipton, Pearson, Cleland, & Yee, 2003a, 2003b; Marlowe, 2003; McLellan, Lewis, O'Brien, & Kleber, 2000; Simpson, Joe, & Brown, 1997). In fact, data indicate that 6-12 months of treatment may be the "dose threshold" for observing lasting reductions in drug use (Marlowe, 2003). Furthermore, the intensity of the intervention should be commiserate with the risk such that greater intensity of programming/intervention is neces- sary to reduce recidivism among individuals at higher risk of reoffending (i.e., perpetrating violence; continued substance use; engaging in lawbreaking behaviors; Andrews et al., 1990).

These evidence-based principles of effective practice suggest that men in the Extreme Violence Severity group should receive the most intensive and comprehensive interven- tion efforts. Interestingly, Holtzworth-Munroe and Meehan's (2004) research identifying batterer typologies found that the men in their study that were most similar to our Extreme Violence Severity group (i.e., the generally violent/antisocial group) appeared to experi- ence the most challenges regarding participation in traditional batters interventions (e.g., not entering treatment; dropping out of treatment; resuming violence after treatment). Consistent with the principles of evidence-based practice with CJ populations identified earlier, Holzworth-Munroe and Meehan suggest that cognitive-behavioral approaches and intensive, rehabilitation-oriented CJ strategies may be the most effective in creating behavior change in this particular group (Gendreau, Cullen, & Bonta, 1994).

Lastly, the ability to address basic issues of living (e.g., food, shelter, clothing) is criti- cal in the design of any intervention for this populations, regardless of the level of violence perpetration. Sociodemographic data indicate high levels of undereducation and unemploy- ment among these men as a whole. For example, almost a third of the men in this study reported that they had less than a high school diploma or GED. In comparison, national data indicate that among all men 25 years and older, 84% are at least high school graduates (this figure is approximated 87% for White men and 80% for Black men; Crissey, 2009). Research attempting to isolate the most critical and relevant risk factors associated with criminal behavior has identified personal education and vocational achievement as being significantly associated with recidivism (Andrews, 2001). Thus, intervention components that address education, employment, and basic necessities of life are essential elements of any programming for this population.

This research has several limitations. Men who participated in this study do not con- stitute a randomly selected sample, thus the data in this study may not be representative of all drug-involved (or crack-involved) men. This study relies heavily on respondents' self-reports of socially undesirable behaviors. Self-report of violence and victimization is "thought to produce a relatively accurate but also incomplete assessment" (Heckert & Gondolf, 1997; Swan & Snow, 2002). Some of our assessments used short-item mea- sures to capture experiences of violence and victimization; assessments with more items may capture more variability in these experiences. Similarly, research on substance use and other risk behaviors indicates that both the validity and reliability of self-report data is good to excellent (Darke, 1998; Fincham, 1992; Rouse, Kozel, & Richards, 1985). Because of the cross-sectional nature of the present data, studies that employ longitudinal designs are needed before causal conclusions can or should be drawn.

The purpose of this study was to identify subgroups of drug-involved men based on the severity of violence perpetration and to examine between group differences across the domains of violence perpetration, victimization, CJ involvement and lawbreaking, and substance use. Results suggest that important differences exist in the severity of violence perpetration among men in this high-risk population. Furthermore, the three subgroups evidenced unique patterns of victimization, CJ involvement and lawbreak- ing, and substance use. This research provided further information about the intersec- tion of these factors among drug-involved men and provided the data necessary to develop tailored intervention strategies designed to meet the needs of different sub- groups of men so that substance use and CJ involvement can be reduced or prevented altogether.

NOTES

1. In considering why childhood victimization does not appear to be a risk factor for adult substance use in men, Widom and colleagues have developed a hypothesis they refer to as the "satu- ration effect" (Widom, Ireland, & Glynn, 1995; Widom, et al., 2006). According to the saturation effect, because men in general have a higher base risk for substance use (i.e., they are already fully saturated with risk for substance use), the experience of childhood victimization, in and of itself, does not exert an additional "demonstrable effect" resulting in higher levels of substance use (Widom et al., 1995).

2. Among the participants in this study, 85.8% reported that they had not had a same sex partner ever in their lives; only 3.4% (n 5 5) of the total sample reported that they had a same sex sexual partner in the past 2 years.

3. As indicated by the breadth of items used to assess engagement in violence (Table 1), violent behavior is broadly operationalized, encompassing different behaviors; it is likely that there is vari- ance in the etiology of these behaviors. For example, some items are assessing controlling behaviors while others are assessing physical or sexual violence. Likewise, certain behaviors may be more correlated with IPV than non-IPV. To further disentangle the nature of violence in this population, an exploratory factor analysis was conducted on the 16 items assessing violence perpetration. The items were submitted to a series of principal component analyses (PCA) using principle axis factoring (via SPSS version 21.0; Faul & Zyl, 2004). The Kaiser criterium of eigenvalues greater than 1 was used to estimate the initial number of factors. The identified factor matrixes were rotated to simple structures with the varimax procedure. The initial PCA extracted six factors, three of which had eigenvalues greater than 1. The next iteration specified a three factor solution, accounting for 31.53% of the total variance. Items 1 through 7, 10, and 11 loaded on Factor 1 (eigenvalue 5 2.65; variance explained 5 16.60%; alpha 5 .698); items 12 and 13 loaded on Factor 2 (eigenvalue 5 1.47; vari- ance explained 5 9.20%; alpha 5 .595); and items 14 through 16 loaded on Factor 3 (eigenvalue 5 .917; variance explained 5 5.73%; alpha 5 .624). Because one of the factors in this solution had an eigenvalue less than 1, a two-factor solution was also evaluated; however, the three-factor solution was assessed to provide a better empirical and substantive interpretation of the data. The three-factor solution can be interpreted as follows: Factor 1, controlling and physically violent behaviors; Factor 2, armed violence; and Factor 3, sexually violent behaviors.

4. Items 10 through 16 in Table 1 were weighted by multiplying an affirmative response (1) by the following numbers: 2, 5, 8, 6, 5, 5, 5, respectively (Straus, 1990; Straus et al., 2003). Affirmative responses to the remaining questions were left unchanged (i.e., equal to 1).

5. The range for this variable was limited to 100; four individuals reported several lifetime arrests greater than 100.

REFERENCES

Afifi, T., MacMillan, H., Cox, B., Asmundson, G., Stein, M., & Sareen, J. (2009). Mental health cor- relates of intimate partner violence in marital relationships in a nationally representative sample of males and females. Journal of Interpersonal Violence, 24(8), 1398-1417.

Andrews, D. A. (2001). Principles of effective correctional programs. In L. L. Motiuk & R. C. Serin (Eds.), Compendium 2000 on effective correctional programming (pp. 9-17). Ottawa, Canada: Correctional Service Canada.

Andrews, D. A., Zinger, I., Hoge, R. D., Bonta, J., Gendreau, P., & Cullen, F. T. (1990). Does cor- rectional treatment work? A clinically-relevant and psychologically-informed meta-analysis. Criminology, 28, 369-404.

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

Bennett, L., & Williams, O. (2003). Substance abuse and men who batter: Issues in theory and prac- tice. Violence Against Women, 9(5), 558-575.

Boles, S. M., & Miotto, K. (2003). Substance abuse and violence: A review of the literature. Aggression and Violent Behavior, 8, 155-174.

Center for Substance Abuse Treatment. (2005). Substance abuse treatment for adults in the criminal justice system. Treatment improvement protocol (TIP) Series, No. 44. Rockville, MD: Substance Abuse and Mental Health Services Administration. Retrieved from http://www.ncbi.nlm.nih .gov/books/NBK64124/

Coker, A. L., Davis, K., Arias, I., Desai, S., Sanderson, M., Brandt, H., & Smith, P. (2002). Physical and mental health effects of intimate partner violence for men and women. American Journal of Preventive Medicine, 23(4), 260-268.

Crissey, S. (2009). Educational attainment in the United States: 2007. Current Population Reports. Washington, DC: U.S. Census Bureau.

Darke, S. (1998). Self-report among injecting drug users: A review. Drug and Alcohol Dependence, 51, 253-263.

Desai, S., Arias, I., Thompson, M., & Basile, K. (2002). Childhood victimization and subsequent adult revictimization assessed in a nationally representative sample of women and men. Violence & Victims, 17(6), 639-653.

El-Bassel, N., Gilbert, L., Wu, E., Chang, M., & Fontdevila, J. (2007). Perpetration of intimate part- ner violence among men in methadone treatment programs in New York City. American Journal of Public Health, 97(7), 1230-1232.

Fagan, J., & Brown, A. (1994). Violence between spouses and intimates: Physical aggression between women and men in intimate relationships. In A. Reiss & J. Roth (Eds.), Understanding and prevent- ing violence: Social influences (Vol. 3, pp. 115-292). Washington, DC: National Academy Press.

Faul, A., & Zyl, M. A. V. (2004). Constructing and validating a specific, multi-item assessment or evaluation tool. In A. R. Roberts & R. Y. Kenneth (Eds.), Desk reference of evidence based prac- tice in health care and human services (pp. 564-581). New York, NY: Oxford University Press.

Feucht, T., & Gfroerer, J. (2011). Mental and substance use disorders among adult men on proba- tion or parole: Some success against a persistent challenge. SAMHSA Data Review. Rockville, MD: Substance Abuse and Mental Health Services Administration, National Institute of Justice.

Field, G. (1986). Psychological deficits and treatment needs of chronic criminality. Federal Probation, 50(4), 60-66.

Fincham, F. (1992). Assessing attributions in marriage: The relationship attribution measure. Journal of Personality and Social Psychology, 62(3), 457-468.

Gendreau, P., Cullen, F. T., & Bonta, J. (1994). Intensive rehabilitation supervision: The next genera- tion in community corrections? Federal Probation, 58, 72-78.

Golder, S., Ivanoff, A., Cloud, R., Besel, K., McKiernan, P., Bratt, E., & Bledsoe, L. K. (2005). Evidence based practice with adults in jails and prisons: Strategies, practices, and future direc- tions. Best Practices in Mental Health: An International Journal, 1(2), 100-132.

Golder, S., & Logan, T. (2010). Lifetime victimization and psychological distress: Cluster profiles of out of treatment drug-involved women. Violence & Victims, 25(1), 62-83.

Golder, S., & Logan, T. (2011). Cumulative victimization, psychological distress, and high risk behavior among substance-involved women. Violence & Victims, 26(4), 477-495.

Goldstein, P. J. (1985). The drugs/violence nexus: A tripartite conceptual framework. Journal of Drug Issues, 15, 493-506.

Harlow, C. (1999). Prior abuse reported by inmates and probationers. Washington, DC: Bureau of Justice Statistics.

Heckert, A., & Gondolf, E. (1997, July). Agreement of men's and women's reports of woman assault during batterer program follow-up. Paper presented at the Fifth International Family Violence Research Conference, Durham, NH.

Holtzworth-Munroe, A. (2000). A typology of men who are violent toward their female partners: Making sense of the heterogeneity in husband violence. Current Directions in Psychological Science, 9(4), 140-143.

Holtzworth-Munroe, A., & Meehan, J. (2004). Typologies of men who are maritally violent: Scientific and clinical implications. Journal of Interpersonal Violence, 19(12), 1369-1389.

Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychological Bulletin, 116(3), 476-497.

Inciardi, J., & Pottieger, A. (1994). Crack cocaine use and street crime. Journal of Drug Issues, 24, 273-292.

Lawson, D. M., Weber, D., Beckner, H. M., Robinson, L., Marsh, N., & Cool, A. (2003). Men who use violence: Intimate violence versus non-intimate violence profiles. Violence & Victims, 18(3), 259-277.

Lipton, D., Pearson, F., Cleland, C., & Yee, D. (2003a). The effectiveness of cognitive-behavioural treatment methods on offender recidivism: Meta-analytic outcomes from the CDATE pro- grammes. In J. McGuire (Ed.), Offender rehabilitation and treatment: Effective programmes and policies to reduce re-offending (pp. 79-112). London, United Kingdom: Wiley.

Lipton, D., Pearson, F., Cleland, C., & Yee, D. (2003b). The effects of therapeutic communities and milieu therapy on recidivism: Meta-analytic findings from the correctional drug abuse treatment effectiveness (CDATE) study. In J. McGuire (Ed.), Offender rehabilitation and treat- ment: Effective programmes and policies to reduce re-offending (pp. 39-78). London, United Kingdom: Wiley.

Logan, T., & Leukefeld, C. (2000). Violence and HIV risk behavior among male and female crack users. Journal of Drug Issues, 30(2), 261-282.

Logan, T., Walker, R., & Leukefeld, C. (2001). Intimate partner and non-intimate violence his- tory among drug-using, incarcerated men. International Journal of Offender Therapy and Comparative Criminology, 45(2), 228-243.

Marlowe, D. (2003). Integrating substance abuse treatment and criminal justice supervision. Addiction Science & Clinical Practice, 2(1), 4-14.

Maxfield, M. G., & Widom, C. S. (1996). The cycle of violence: Revisited 6 years later. Archives of Pediatric and Adolescent Medicine, 150, 390-395.

McClellan, D., Farabee, D., & Crouch, B. (1997). Early victimization, drug use, and criminality: A comparison of male and female prisoners. Criminal Justice and Behavior , 24 (4), 455-476.

McLellan, A. T., Lewis, D. C., O'Brien, C. P., & Kleber, H. D. (2000). Drug dependence, a chronic medical illness: Implications for treatment, insurance, and outcomes evaluation. Journal of the American Medical Association, 284(13), 1689-1695.

Morse, B. J. (1995). Beyond the Conflict Tactics Scale: Assessing gender differences in partner violence. Violence and Victims, 10, 251-272.

Nowinski, S., & Bowen, E. (2012). Partner violence against heterosexual and gay men: Prevalence and correlates. Aggression and Violent Behavior, 17, 36-52.

Pearson, F., & Lipton, D. S. (1999). A meta-analytic review of the effectiveness of corrections-based treatments for drug abuse. Prison Journal, 79(4), 384-410.

Pearson, F., Lipton, D., Cleland, C., & Yee, D. (2002). The effects of behavioral/cognitive-behavioral programs on recidivism. Crime & Delinquency, 48(3), 476-496.

Rouse, B., Kozel, N., & Richards, L. (1985). Self-report methods of estimating drug use: Meeting current challenges to validity (NIDA Research Monograph No. 57). Lexington, KY: National Institute of Drug Abuse.

Sabol, W., & Couture, H. (2008). Prison inmates at midyear 2007. Washington, DC: Bureau of Justice Statistics.

Simpson, D. D., Joe, G. W., & Brown, B. S. (1997). Treatment retention and follow-up outcomes in the Drug Abuse Treatment Outcome Study (DATOS). Psychology of Addictive Behaviors, 11(4), 294-307.

Smith, C., & Thornberry, T. P. (1995). The relationship between childhood maltreatment and adoles- cent involvement in delinquency. Criminology, 33, 451-477.

Stouthamer-Loeber, M., Loeber, R., Homish, D. L., & Wei, E. (2001). Maltreatment of boys and the development of disruptive and delinquent behavior. Developmental Psychopathology, 13, 941-955.

Straus, M. A. (1990). Measuring intrafamily conflict and violence: The Conflict Tactics Scale. In M. Straus & J. Gelles (Eds.), Physical violence in american families: Risk factors and adaptation to violence (pp. 29-48). New Brunswick, NJ: Transaction Publications.

Straus, M. A. (1999). The controversy over domestic violence by women: A methodological, theo- retical, and sociology of science analysis. In X. B. Arriaga & S. Oskamp (Eds.), Violence in intimate relationships (pp. 17-44). Thousand Oaks, CA: Sage.

Straus, M. A., & Gelles, J. (1986). Societal change and change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and the Family, 48, 465-479.

Straus, M. A., Hamby, S. L., & Warren, W. L. (2003). The Conflict Tactics Scales handbook (Vol. 19). Los Angeles, CA: Western Psychological Services.

Swan, S., & Snow, D. (2002). A typology of women's use of violence in intimate relationships. Violence Against Women, 8(3), 286-319.

Tjaden, P., & Thoennes, N. (2000). Full report of the prevalence, incidence, and consequences of violence against women: Findings from the national violence against women survey. Washington, DC: U.S. Department of Justice.

Wanberg, K. W., & Milkman, H. B. (1998). Criminal conduct and substance abuse treatment: Strategies for self-improvement and change. Thousand Oaks, CA: Sage.

Weeks, R., & Widom, C. S. (1998). Self-reports of early childhood victimization among incarcerated adult male felons. Journal of Interpersonal Violence, 13(3), 346-361.

White, H. R., & Gorman, D. M. (2000). Dynamics of the drug-crime relationship. In G. LaFree (Ed.), Criminal justice 2000 The nature of crime: Continuity and change (pp. 151-218). Washington, DC: U.S. Department of Justice.

Widom, C. S., & Ames, M. A. (1994). Criminal consequences of childhood sexual victimization. Child Abuse & Neglect, 18(4), 303-318.

Widom, C. S., Ireland, T., & Glynn, P. J. (1995). Alcohol abuse in abused and neglected children followed-up: Are they at increased risk? Journal of Studies on Alcohol, 56, 207-217.

Widom, C. S., Marmorstein, N., & White, H. R. (2006). Childhood victimization and illicit drug use in middle adulthood. Psychology of Addictive Behaviors, 20(4), 394-403.

Wolff, N., & Shi, J. (2012). Childhood and adult trauma experiences of incarcerated persons and their relationship to adult behavioral health problems and treatment. International Journal of Environmental Research and Public Health, 9, 1908-1926.

Zingraff, M., Leiter, J., Myers, K., & Johnsen, M. (1993). Child maltreatment and youthful problem behavior. Criminology, 31, 173-202.

Seana Golder, PhD

University of Louisville, Louisville, KY

TK Logan, PhD

University of Kentucky, Lexington, KY

Correspondence regarding this article should be directed to Seana Golder, PhD, University of Louisville, Kent School of Social Work, Louisville, KY 40292. E-mail: [email protected]