An Evaluation of Intimate Partner Violence Intervention With Incarcerated Offenders

Abstract

The following study is an evaluation of the Moderate Intensity Family Violence Prevention Program (MIFVPP). The sample consisted of 298 male federal offenders who participated in the MIFVPP while incarcerated or on release within the community. Participants were assessed pre-, mid-, and postprogram using an assessment battery consisting of self-report questionnaires and facilitator-rated evaluation scales. Results of the study found uniform and significant (p < .001) improvement for pre and post program change in the self-report questionnaires and in the facilitator ratings. A positive improvement in motivation, whether assessed by the participant or facilitator, was associated with improvement in program outcomes and significant within, between, and interaction effects were found when participant program performance over time was compared among grouped postprogram ratings of motivation. The implication for the efficacy of addressing offender motivation to change in intimate partner violence (IPV) interventions is discussed.

Keywords

intimate partner violence offender program family violence

Intimate partner violence (IPV) remains an ongoing public health and public safety concern. A survey of 10 countries presented at the World Health Organization forum in 2005 found that 15% to 71% of women reported being physically assaulted by an intimate partner at some point in their lives (Garcia-Moreno, Heise, Jansen, Ellsberg, & Watts, 2006). In the United States, surveys indicate that 1.3 million women annually are physically assaulted by an intimate partner (Tjaden & Thoennes, 2000) and that 40% to 50% of all murdered women in that surveyed year were killed by an intimate partner (National Institute of Justice, 2000). According to the web site for the American Institute on Domestic Violence (2010), health related costs from IPV assaults exceeds US$5.8 billion annually in the United States, while costs associated with loss of productivity and wages are estimated at US$1.8 billion each year. These aforementioned rates are reflective of the increasing number of IPV perpetrators incarcerated within the North American criminal justice systems (Dutton & Hart, 1992; Kropp & Lee, 2004; White, Gondolf, Robertson, Goodwin, & Caraveo, 2002)

Until recently, treatment programs offered within correctional institutions for male perpetrators of IPV have been neglected within the research literature (White et al., 2002). Within Canada, the Correctional Service of Canada (CSC) has developed specialized family violence programming to treat perpetrators of IPV serving federal sentences. To our knowledge, it is the only IPV prevention program in North America provided to incarcerated offenders that delivers 100 hours of intervention. The purpose of the present study is to evaluate the effectiveness of the IPV program through measuring change in the treatment targets and motivation levels of offenders in this nationally implemented IPV prevention program.

IPV Intervention Targets

To date, the most utilized program to treat IPV perpetrators is the Domestic Violence Intervention Program, commonly referred to as the Duluth Model, developed by Pence and Paymar (1993). This model focuses on treating perpetrators of IPV from a feminist theoretical perspective. The focus of the intervention is on stopping the violence, not on fixing or ending interpersonal relationships (Duluth model.org.). The proponents of the Duluth model argue that it has been widely successful in offering greater victim protection and reducing repeat acts of violence in many different communities. According to Cunningham et al. (1998), the Duluth program is delivered in a group format with the objective to have the male participants accept that they believe in male superiority and recognize they use power and control tactics (as defined by the Duluth Power and Control wheel) to dominate their partners.

Other theorists such as Stith, Smith, Penn, Ward, and Tritt (2004) argue that IPV programs need to include a multifactor treatment target framework. The authors propose that “[t]hese multifactor frameworks suggest that partner violence is not simply caused by an individual’s patriarchal belief system or psychological dysfunction, but rather result from the interaction between various characteristics of the individual and their environment” (pp. 67-68). Stith and colleagues proposed that addressing multiple factors relating to IPV would improve treatment outcome. The Ecological Nested Model developed by Dutton (Stith et al. 2004) is concerned with the interaction of the macrosystem (the societal system that sanctions violence), the exosystem (presence or absence of social structures), the microsystem (the interaction between the couple themselves), and the ontogenic level (the perpetrators’ life history and psychological make up). This model works from the premise that “violence is not committed because of ‘sex role beliefs,’ but formation into an intimate abuser begins developmentally much sooner, from a much broader array of psychosocial risk” (Dutton & Corvo, 2006, p. 458). The focus of this model is the interaction between different levels of the system.

In addition to a multi-risk factor framework for intervention, there is a growing body of research that demonstrates treatment is more beneficial for those who are motivated to change in comparison to those who are unmotivated (Bowen & Gilchrist, 2006; Kistenmacher & Weiss 2009; Rosenfeld, 1992). A number of literature reviews indicate that there are high attrition rates in IPV treatment programs with some researchers reporting that 40% to 50% of perpetrators drop out prior to completion (Bowen & Gilchrist, 2006; Davis & Taylor, 1999; Eckhardt, Holtzworth-Munroe, Norlander, Sibley, & Cahill, 2008; Gondolf, 2000; Hamberger & Hastings, 1988; Saunders, 2008). Research also indicates that program completers reoffend less often than those that do not attend at all or drop out (Babcock & Steiner, 1999; Bowen & Gilchrist, 2004; Cortoni, Nunes, & Latendresse, 2006; Dobash, Dobash, Cavanagh, & Lewis, 1999; Eckhardt et al., 2008; Hanson & Bussiere, 1998; Marshall, Thornton, Marshall, Fernandez, & Mann, 2001). Therefore lowering attrition rates may improve treatment outcome and reduce limitations in research methods and results.

Many perpetrators of IPV are mandated to participate in IPV programs and intrinsic motivation is not frequently present and therefore resistance to treatment occurs (Daniels & Murphy, 1997). This may also explain high attrition rates in these programs (Levesque, Velicer, Castle, & Greene, 2008). Identifying the degree of treatment readiness and motivating perpetrators to participate are therefore key elements for program success. Prochaska and DiClemente (1982) developed the Transtheoretical Model of Change (TTM) to categorize participants according to five Stages of Change: precontemplation, contemplation, preparation/determination, action, and maintenance. Miller and Rollnick (2002) developed a therapeutic style of working with the ambivalent or resistant client (precontemplators/contemplators) to help them engage in the process of change and subsequently remain in treatment. This style is termed Motivational Interviewing (MI) and utilizes an active directive style, yet with a client-centered approach that explores the clients’ ambivalence about changing and enhances intrinsic motivation. Therapeutic alliance and working with the clients’ at their current stage of change is the predominant focus. Outcome studies that have utilized MI in their programs appear to have encouraging results and it is having a positive effect on reducing attrition rates (Chambers, Eccelston, Day, Ward, & Howells, 2008; Cortoni et al., 2006; Joe, Simpson, & Broome, 1998; Marshall et al., 2001). Support for the use of the TTM in IPV treatment stems from results that indicate participants were more engaged in the treatment when they were in the action stage of change and therefore less likely to drop out (Levesque, Gelles, & Velicer, 2000). Researchers such as Sartin, Hansen, and Huss (2006) and Hamberger and Hastings (1988) have theorized that by matching an offenders’ TTM stage to treatment may potentially decrease attrition rates and improve on post treatment outcomes and suggest the use of a motivational enhancement component to address this. A review of the literature concluded that the programs for IPV have “curiously neglected” motivational issues (Day, Chung, O’Leary, & Carson, 2009, p. 208). The program evaluated within the current study has incorporated both TTM and MI into its interviewing, testing, and daily sessions.

The Moderate Intensity Family Violence Prevention Program

The Moderate Intensity Family Violence Prevention Program (MIFVPP) for incarcerated male offenders in the CSC was developed to address the treatment needs of offenders who have perpetrated IPV. This program incorporates the Risk, Need, and Responsivity (RNR) principles (Andrews, Bonta, & Hoge, 1990; Andrews, Bonta & Wormith, 2006) and matches intensity (hours of intervention) of services to the offender’s assessed level of risk to reoffend. The RNR principles propose that higher risk offenders should receive higher levels of intervention (risk principle), targets for intervention should be those modifiable areas (criminogenic needs) related to the criminal behavior (need principle), and delivery of intervention should take into account specific offender characteristics that may influence intervention efficacy (responsivity principle)

The MIFVPP was accredited through the International Accreditation Committee in keeping with “what works best” for the treatment of offenders (Andrews & Dowden, 2007). The program consists of 29, 3-hr group sessions, with at least three individual counseling sessions. Each group has a maximum of 12 participants and is delivered by two facilitators, who were specially trained correctional program delivery officers, with an effort to have the team gender balanced. Every facilitator is provided extensive training by the program developers or accredited trainers and is subject to stringent quality review.

The MIFVPP is comprised of six modules to address identified treatment targets related to reducing IPV. These modules include Motivational Enhancement, Awareness and Education (incorporates components of the Duluth Model), Managing Thoughts and Emotions Related to Abuse (identifying, disputing and replacing cognitive distortions), Social Skills, Self-Management/Relapse Prevention Plan, and Healthy Relationships. The teaching of the skills incorporates adult learning principles and follows cognitive-behavioral therapy strategies. It incorporates the ABC model developed by Ellis (2001), and enhances skill development by engaging in role-plays and systematic steps for skills related to deficits the men have. After each module there is an in-group closed book quiz to test for content knowledge. Remedial sessions are provided for participants who struggled with the material as determined by self-report, facilitator identification, or by quiz results. Completion of the program requires 100% attendance, (sometimes in the form of makeup sessions), completion of all homework assignments, and a completed realistic relapse prevention plan.

Ongoing evaluations for program content knowledge are conducted by way of quizzes administered at the end of each of the first five modules, personalized homework assignments, and the development of a relapse prevention plan. Throughout the program there are personalized role plays for skill development, autobiographies focused on the offenders’ development of abusive behavior, establishing offence pathways and identifying high risk situations, including both thought patterns and a relapse prevention plan.

Most studies in the IPV literature are focused on recidivism. Generally, these studies are plagued with high attrition rates, poor controls, lack of structured programming, differing definitions of recidivism, small sample sizes, and inconsistent or limited follow-up periods (Saunders, 2008). There are few studies that examine the effectiveness of intervention using pre- postdesign (Saunders, 2008), and none that we could identify that treated incarcerated perpetrators of IPV. Most of the IPV literature evaluates community-based programs and/or assesses voluntary versus court-ordered offenders or completers versus noncompleters within a community-based program (Feder & Wilson, 2005; Saunders, 2008). This study will review participant change utilizing pre-post assessment measures that focus on risk factors, assess program content knowledge, skill acquisition, and application. It was hypothesized that offenders would show improvement as measured by self-report, facilitator ratings, and responses to domestic violence vignettes and that those offenders who were more motivated to change would show greater improvement than those who were not motivated.

Method

Participants

The participants were 298 federally incarcerated male offenders who participated in the MIFVPP. In Canada, adult offenders serving two or more years enter federal custody regardless of the type of crime. The average age of participants was 36.6 years (SD = 9.4) and the majority were not married or partnered (71.8%) at the time of the program. Ethnic origin was as follows: White 61.4%, Aboriginal 22.4%, African American 5.5%, and other racial minority 10.7%. Most participants had children (80.8%). At the commencement of the program, participants were asked to report how many times they had assaulted their partner. The results varied considerably: never assaulted partner (9.5%), 1 assault (34.2%), 2 assaults (23.9%), 3 assaults (9.2%), 4 assaults (3.9%), 5 to 10 assaults (11.3%), 10 to 20 assaults (2.1%), 21 or more assaults (1.4%). Thirteen participants chose to use verbal estimates from “a lot,” “frequently,” “often,” and so forth. Participants were also asked to provide an estimate of the period of time the abuse occurred. Ninety percent responded with a timeframe. Of these 30.2% reported that the abuse lasted for less than a month, 18.5% reported that the abuse lasted from 1 month to 1 year, 37.3% said the abuse lasted between 1 to 10 years, and 4.4% reported the abuse lasted for more than 10 years. The attrition rate for this program was 15.8% in total.

Measures

Interpersonal Relationship Scale (IRS)

The IRS is a 27-item self-report measure that assesses six aspects of jealousy: Threat to Exclusive Relationships, Self-Deprecation/Envy, Dependency, Sexual Possessiveness, Competition and Vindictiveness, and Distrust (Hupka & Rusch, 2001). Responses are made on a 6-point scale (ranging from 1 = strongly agree to 6 = strongly disagree) with higher scores indicating lower levels of jealousy. Previous evaluation of the IRS found that on program completion, offenders in both the High and Moderate programs reported better functioning (Kropp & Lee, 2004).

Aggression Questionnaire–Revised (AQ-R)

The AQ-R is a 38-item questionnaire that contains three subscales assessing specific aspects of anger such as awareness, understanding, self-regulation, and self-management (Correctional Service Canada, 2001). It was originally developed for use within an internal anger and emotions intervention program. Responses are made on a 5-point scale (ranging from 0 = not at all true of me to 4 = extremely true of me). Increase in scores from pre- posttesting indicates improvement. Previous evaluation of the AQ-R indicated that change scores on the three subscales were associated with significant reductions in violent reoffending.

Abusive Relationship Inventory (ARI)

The ARI contains 4 subscales: Rationales for Hitting, Need for Control, Legal Entitlement, and Batterer’s Myths (Boer, Kroner, Wong, & Cadsky, 1993). Responses are on a 7-point scale (ranging from 1 = strongly disagree to 7 = strongly agree). Lower scores on the subscales indicate more positive functioning.

Facilitator Structured Ratings (FSR)

As part of the program’s evaluation procedure, 14 anchored ratings of participant behavior were completed by both facilitators independently, with differences negotiated to consensus, on three separate occasions (early-, mid-, and postprogram; Correctional Service Canada, 2001). These 14 ratings covered Program Treatment Targets (11 ratings: acceptance of responsibility for abusive/violent behavior, acknowledges use of power and control tactics, shows empathy for victims, extent of skills development, recognizes cognitive distortions/beliefs and neutralizations, minimizes consequences, understands lifestyle dynamics, understands abusiveness pattern, identify relapse prevention concepts, discloses personal information, overall quality of plan on release) related to violence against a partner, Motivation to Change (1 rating), Overall Participation (1 rating), and Overall Program Performance (1 rating). The latter two ratings are reported for the midprogram and postprogram only. Responses for each rating are anchored along a 5-point scale ranging from −2 to +2 where each score on the continuum has an accompanying description.

Domestic violence vignettes (DVV)

The DVV employs a structured interview and vignettes to evaluate participant responses to potential risk situations relevant to family violence (Correctional Service Canada, 2001). Two versions of five vignettes cover the areas of Jealousy, Employment and Finances, Substance Abuse, Rejection, and Sexual issues. One version is administered preprogram and the second postprogram. Skills are assessed in the following domains: Behavioral Response, Interpretation, and Perspective Taking. Responses are on a 3-point scale (ranging from 0 = ineffective, demonstrates hostile attitudes, and/or no recognition of partner’s thoughts and feelings to 2 = good use of skills, assess partner as equal and good recognition of partner’s thoughts and feelings).

Offenders’ self-rated readiness to change (OSRC)

The offenders’ self-report readiness to change was measured both before the program commenced and after completion through their forced-choice verbal response to the statement “My violence against my partner(s) is” (1) not a problem, (2) is a problem, but I’m not sure what to do about it, (3) is a problem and I intend to take steps to deal with it, or (4) is a problem for me and I am currently dealing with it by making change in me and my life (Correctional Service Canada, 2001). Prior to the forced-choice response, “violence” was verbally defined for the offender as

trying to control your partner physically, threatening to hit or throw something at your partner, throwing something at your partner, pushing, grabbing, or shoving your partner, slapping your partner, kicking, biting, or hitting your partner with a fist, choking or strangling your partner, physically forcing your partner to have sex, beating your partner up, threatening your partner with a knife or gun, using a knife or gun on your partner.

The rationale for defining violence stems from many offenders endorsing severe cognitive distortions around violence that some physical altercations may not be considered by them to be violence.

Procedure

Referral criteria

The participants were screened into the MIFVPP based on their level of estimated risk for domestic violence being within the moderate range according to the Spousal Assault Risk Assessment (SARA; Kropp, Hart, Webster, & Eaves, 1999) and on the presence of one prior incident of abuse against an intimate partner. It should be noted that the incident of abuse need not have resulted in a conviction and could be identified through official documents such as police reports or offender self-report. Participants could decline the program but did so with the knowledge that it may affect their release opportunities. Those who did participate gave informed consent. The MIFVPP was delivered at minimum, medium, and maximum security institutions as well as within the community for offenders on various forms of release.

Assessment process

The participants were assessed at various points throughout the program. Preprogram (within 30 days prior to group commencement) and postprogram (within 10 days following group completion) assessment included the IRS, AQ-R, ARI, DVV, and OSRC. The FSR ratings occurred early into the program, at the midpoint of the program (Module 3, second session), and postprogram. In addition, five quizzes on the content of the program were administered following each of the first five modules. Scoring of the quiz responses was through structured anchored ratings.

Analysis

The strategy for analyzing the data was threefold. First we examined the pre- and postdifferences (t tests) in the assessment instruments and the facilitator ratings. Next we investigated the relationship of motivation and change by comparing motivated versus unmotivated participants postprogram (t tests) and by correlating change in motivation with change in assessment measures. Finally, we employed a mixed (within-between) design to compare the change in program performance over time and between groups of offenders. The within participant change (repeated measures) was measured by the FSR ratings at points early-, mid-, and postprogram. The between groups comparisons were the levels of motivation both as rated by the facilitator and by the participant.

Results

Prepost Program Differences

Descriptive statistics for both pre- and postadministrations of the three self-report instruments are reported in Table 1. We only report the total scores; however, all of the subscales for each instrument were statistically different at the p <.001 level. The results show a uniform and significant change in the direction of improvement. The ARI demonstrates improvement with decreases in scores, whereas the IRS and AQ-R demonstrate improvement when the score increases. Pre-program and postprogram scores of the DVVs are reported in Table 1. The three domains of Behavioral Response, Interpretation, and Perspective Taking were scored from 0 to 2 for each of the five vignettes. This produced a possible range of scores from 0 to 10 for each of the three domains. All of the pair-wise comparisons were significant at the p < .001 level. However, the mean scores would indicate that even following completion of the program the average participant scored little better than 50%.

Table 1.

Descriptive Statistics of the Self-Report Instruments and Domestic Violence Vignettes

	Pre-Program			Post Program
	Mean	SD	α	Mean	SD	α
Interpersonal Relationship Scale	119.5	22.8	.93	133.6	17.9	.93
Aggression Questionnaire–Revised	138.8	14.5	.90	151.6	13.6	.90
Abusive Relationships Inventory	54.1	23.3	.89	41.7	13.8	.94
Domestic violence vignettes
Behavioral response	3.3	2.2	—	6.3	2.5	—
Interpretation	3.3	2.1	—	6.0	2.6	—
Perspective taking	3.3	2.1	—	5.5	2.5	—
Offenders’ self-rated readiness to change	2.9	1.1	—	3.3	1.1	—

Note: All pair-wise comparisons are statistically significant at the p < .001 level.

All FSRs showed improvement at the p < .001 level (see Table 2). The Program Treatment Targets (mean scores of 11 ratings) showed consistent improvement from early program (M = −9.2, SD = 6.9), to midprogram (M = −2.3, SD = 7.4) to postprogram (M = 4.3, SD = 8.3). Comparisons between early and midprogram were significantly different, t(262) = 18.2, p < .001, as were differences between mid- and postprogram, t(262) = 22.7, p < .001. Facilitator-rated Overall Participation showed significant improvement between mid- (M = .21, SD = .79) and postprogram (M = .54, SD = .89), t(277) = 8.9, p < .001, as did facilitator-rated Overall Program Performance mid- (M = .00, SD = .77) and postprogram (M = .42, SD = .86), t(277) = 10.0, p < .001. Recall that the ratings were anchored from −2 to +2, thus for most areas the average score did not move into the positive adjustment range until the final rating. Furthermore, the range of ratings would indicate that the deficits at the beginning of the program were greater than the positive adjustment at the end of the program.

Table 2.

Facilitator Structured Ratings

	Early Mean (SD)	Mid Mean (SD)	Post Mean (SD)
1. Acceptance of responsibility	−.56 (.93)	−.03 (.92)	.50 (.94)
2. Acknowledges use of power and control	−.69 (.99)	.08 (.89)	.62 (.94)
3. Shows empathy to victims	−.76 (.95)	−.28 (.88)	.21 (.93)
4. Skills development	−.67 (.85)	−.11 (.74)	.36 (.72)
5. Recognizes cognitive distortions	−1.23 (.76)	−.39 (.89)	.22 (.97)
6. Minimizes consequences	−.91 (.84)	−.36 (.87)	.14 (.98)
7. Understands lifestyle dynamics	−.66 (.87)	−0.1 (.86)	.68 (.91)
8. Understands abusive patterns	−.93 (.80)	−.19 (.92)	.50 (.96)
9. Identify relapse prevention concepts	−1.14 (.76)	−.42 (.79)	.35 (.90)
10. Discloses personal information	−.32 (.87)	.14 (.88)	.62 (.93)
11. Quality of release plan	−1.24 (.76)	−.54 (.78)	.22 (.86)
12. Motivation to change	−.53 (.85)	−.07 (.86)	.41 (.95)
13. Overall participation in program	—	.21 (.79)	.52 (.87)
14. Overall program performance	—	0.0 (.77)	.41 (.85)

Note: All pairwise comparisons are at the p < .001 level.

Motivation to Change and Pre-Post Program Differences

The participants’ self-reported readiness to change was measured by the OSRC before the program commenced (M = 2.9, SD = 1.1) and after the program was completed (M = 3.3, SD = 1.1). Mean changes in these participant ratings were in the positive direction, t(261) = 5.0, p < .001. Facilitator rated Motivation to Change was measured at three points and showed significant improvement from early (M = −.53, SD = .86), to mid- (M = −.07, SD = .86) to post program (M = .42, SD = .96). Comparisons between early and midprogram were significantly different, t(277) = 9.3, p < .001, as were differences between mid- and post program, t(277) = 11.4, p < .001.

Our next step was to compare Pre-Post change in the IRS, AQ-R, ARI, and DVV between participants classified as “not ready” or “ready” to change as rated by the participant and the facilitator at the end of the program (Table 3). The results show that, with the exception of the ARI, there was significantly greater improvement (pre-post change) on all measures for those who were rated as “ready to change,” this was true regardless of who made the rating. In addition, we compared the content quiz performance for these same two groups and again the results showed those rated as ready to change performed significantly better than those who did not regardless if the ratings were made by offender or facilitator.

Table 3.

Change in Self-Report, Facilitator-Rated Performance, and Quiz Performance by Motivation

	Offender-Rated Motivation (Post program)			Facilitator-Rated Motivation (Post program)
	Not Ready (n = 92)	Ready to Change (n = 170)	p	Not Ready (n = 153)	Ready to Change (n = 137)	p
Pre- Post change
Self-report measures
IRS	10.9	16.5	*	11.7	16.6	*
AQ-R	10.2	14.5	*	10.9	14.7	*
ARI	−9.4	−14.2	ns	−10.8	−14.0	ns
DV vignettes
Behavioral response	1.8	3.5	***	2.4	3.6	***
Interpretation	1.9	2.9	**	2.1	3.2	***
Perspective taking	1.6	2.3	*	1.5	2.8	***
Content knowledge
Quiz 1	9.9	11.1	**	9.6	11.9	***
Quiz 2	13.9	16.1	***	14.6	16.3	***
Quiz 3	4.5	5.9	***	4.8	6.3	***
Quiz 4	5.5	7.0	***	5.8	7.3	***
Quiz 5	3.6	4.6	***	3.7	4.8	***

Note: IRS = Interpersonal Relationship Scale, AQ-R = Aggression Questionnaire–Revised, ARI = Abusive Relationships Inventory.

p < .05. **p < .01. ***p < .001.

As already reported, pre- postprogram changes were evident among all of the variables and end of program motivation was related to program performance. The next step in our analyses was to examine the relationship of change in motivation with change in the assessment measures using both self-report and facilitator ratings. Table 4 reports the correlations among these variables for both participant and facilitator-rated change in motivation. With the exception of the self-report measures, change in motivation, whether assessed by the participant or the facilitator, resulted in a significant relationship with facilitator ratings of performance on the vignette scenarios and treatment target improvement. Indicating that positive improvement in motivation was associated with improvement in program outcomes.¹

Table 4.

Correlations Between Change in Motivation and Change in Study Measures

Change in Self-Report and Performance	Facilitator Identified Change in Motivation	Offender Identified Change in Motivation
Self-report measures
IRS	.07	.09
AQ-R	−.10	−.13*
ARI	.08	.13*
DV vignettes
Behavioral response	.24***	.26***
Interpretation	.23***	.26***
Perspective taking	.26***	.18**
Facilitator structured ratings
Early to midratings change	.35***	.15*
Mid- to postratings change	.32***	.21**

Note: IRS = Interpersonal Relationship Scale, AQ-R = Aggression Questionnaire–Revised, ARI = Abusive Relationships Inventory.

p < .05. **p < .01. ***p < .001.

Finally, we employed a mixed within-between design to examine the relationship of participant program performance over time (early, mid-, and post program) between three groups of participants according to their post program ratings of motivation (Table 5). The results indicated that there were significant within, between, and interaction effects. While participants improved over time, those who were rated as motivated at the end of the program had improved at a much significantly higher rate. This finding was the same regardless if motivation was participant or facilitator rated. Those participants in the highest motivation group appeared to have started the program with deficits but not to the same extent as those participants in the lower motivation groups. Thus suggesting that motivation and the extent of initial deficits may have been related. Moreover, as displayed within Table 5, we examined those 140 participants who were rated by the facilitators as unmotivated at the beginning of the program as a subsample by applying the same within-between analysis. Once again there were significant within, between, and interaction effects. Importantly, differences in initial deficits at the beginning of the program were not the same, yet there were clear indications that those participants who ended the program motivated improved at a faster rate in comparison to participants whose motivation remained unchanged.

Table 5.

Within Participant Performance Changes by Post Program Readiness to Change

	Early	Mid	Post	Effect	F	p
Offender postprogram self-rated motivation
No problem or not sure	−12.1	−7.7	−2.8	Within	479	<.001
Recognizes problem, expresses intent	−9.7	−4.4	3.8	Between	26	<.001
Recognizes problem, efforts to change	−8.1	−0.1	6.8	Interaction	10.9	<.001
Facilitator rated postprogram motivation
Unmotivated or inconsistent	−13.7	−9.8	−6.6	Within	752	<.001
Motivated	−10.6	−4.4	1.9	Between	112	<.001
Well motivated and consistent	−6.4	2.2	10.2	Interaction	37	<.001
Facilitator-rated postprogram motivation for participants who were unmotivated early in the program (n = 140)
Unmotivated or inconsistent	−15.2	−10.9	−7.3	Within	750	<.001
Motivated	−12.6	−5.4	1.4	Between	59	<.001
Well motivated and consistent	−12.4	−1.6	8.2	Interaction	47	<.001

Discussion

Our results demonstrate that offenders significantly improved in skill acquisition and application over the course of the program. Based on the pre-post test scoring, there was a positive change in attitudes, a reduction in feelings of jealousy, anger, and dependency, an increase in their acceptance of responsibility, an improvement in their ability to dispute their cognitive distortions regarding their violence, and an improvement in their ability to engage in perspective taking and dealing with conflict. Although the change was significant, the participants did not acquire all the skills provided in the program based on facilitator ratings of program treatment targets. Based on these ratings, offenders had greater deficits (negative ratings) at the beginning of the program than corresponding positive ratings at program completion.

Overall, the motivation of the participants improved over the course of the program. Positive change in motivation, whether facilitator rated or participant rated, was associated with improved outcomes. Furthermore, there was a significant interaction of program improvement over time between groups with varying levels of motivation at the end of the program. This held true for those participants who started the program unmotivated. These findings are consistent with Bowen and Gilchrist (2004) who argue for focusing on perpetrators readiness to change and MI in the treatment of IPV. It became evident during our analysis that motivation is a dynamic construct (McMurran, 2002; Miller & Rollnick, 2002; Prochaska & DiClemente, 1982).

Based on the results of this study, it would appear that efforts to improve motivation are a requirement for program engagement and offender change. Results of our study suggest that offenders who are motivated learn more, engage more, and do so at a faster rate in comparison to unmotivated offenders. Other researchers remarking on this have made claims that enhancing motivation is a necessary requirement for full program/intervention engagement from the participant (Prochaska & Levesque, 2002). Given the high attrition rate in IPV programming generally (Olver, Stockdale, & Wormith, 2011), it would appear that developing strategies for motivating and retaining offenders early in the program can result in their becoming motivated later in the program with the accompanying improvement in treatment target outcomes.

Of note was that our sample does not fit the typical profile of a perpetrator who completes programs in the community. Typically the completer is more educated, more likely to be employed, married, White, and less likely to have a criminal record (Babcock, Green, & Robie, 2004). Despite this, 85% of participants completed the program. We observed that the initial facilitator ratings of motivation to change for the 15% who did not complete (M = −.85) was statistically lower than those who did complete the program (M = −.53), t(351) = 2.6, p < .01, but not as low as the overall group median of −1 for those who completed the program. We also looked at the mean motivation to change for those offenders who ended the program “well motivated and consistent” but started unmotivated (see Table 5) and discovered that the mean of this group was −1.1. While the identified dropouts were among the lesser motivated participants our results would suggest that improving methods to maintain them within the program may result in positive outcomes for some.

We speculate that external motivators such as improved opportunities for early release and the interactive activity a program offers to an incarcerated offender may motivate them to enter the program. Furthermore, within prison there are fewer life circumstances (i.e., employment, housing, finances, relationships) to distract the participant from completion. However, all offenders entered the program under the same set of external motivators and all offenders in our analyses completed the program but it was real improvement in motivation to change during the program that resulted in better program outcomes as measured by facilitator ratings and objective performance on the DVVs. We also speculate that the intentional inclusion of motivational enhancement practices throughout the duration of the program, both in the group as well as in the individual sessions, decreased attrition (Chambers et al., 2008; Marshall et al. 2001; Joe et al., 1998).

However, some limitations of the study are apparent. The first is that the outcome variables were rated by the facilitators therefore introducing a potential bias. This bias, however, was addressed by having both facilitators rate the participant. We further note that facilitators training included interviewing and scoring pre and post materials by both facilitators to ensure interrater reliability. The second limitation relates to the lack of ethnic diversity of the participants who were primarily Caucasian. The third limitation centers on our study’s focus on pre-post program performance as a measure of change and not recidivism. That said, the pre-post analysis is deemed an important first step in the overall program evaluation strategy as it identifies what treatment targets have changed before relating those changes to the outcome of recidivism. A fourth limitation is that this study was completed on incarcerated men who may be externally motivated (for release purposes) to participate in the program. These findings may not generalize to community samples where noncompletion or nonparticipation may not have the same potentially negative consequences.

The results from this study support targeting empirically based risk factors within a “what works best” offender treatment paradigm with an emphasis on motivating offenders to remain in treatment and change their behavior. What seems clear is that motivation, as a key component to learning, is a dynamic factor that can and does change over time for improved treatment results. Whether this affects recidivism remains to be determined but knowledge and skills development are definitely affected by motivation. Furthermore, our study indicates that not all IPV programs need suffer from high attrition rates and identifying perpetrators who are incarcerated for treatment and intervention may be an effective strategy for reducing societal IPV.

Footnotes

Acknowledgements

The authors would like to thank Carolyn Bourgeois for her assistance with data collection.

The views expressed in this article are those of authors and do not necessarily reflect the views or policies of the Correctional Service of Canada.

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

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

Notes

References

Andrews

D. A.

Bonta

Hoge

R. D.

(1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52.

Andrews

D. A.

Bonta

Wormith

J. S.

(2006). The recent past and near future of risk and/or need assessment. Crime and Delinquency, 52, 7-27.

Andrews

D. A.

Dowden

(2007). The risk-need-responsivity model of assessment and human service in prevention and corrections: Crime-prevention jurisprudence. Canadian Journal of Criminology and Criminal Justice, 49, 439-464.

Babcock

J. C.

Green

C. E.

Robie

(2004). Does batterers’ treatment work? A meta-analytic review of domestic violence treatment. Clinical Psychology Review, 23, 1023-1053.

Babcock

J. C.

Steiner

(1999). The relationship between treatment, incarceration, and recidivism of battering: A program evaluation of Seattle’s coordinated community response to domestic violence. Journal of Family Psychology, 13, 46-59.

Boer

D. P.

Kroner

D. G.

Wong

Cadsky

(1993). Assessing the dysfunctional beliefs of spouse assaulters: Development and validation of the Abusive Relationships Inventory. (Unpublished manuscript).

Bowen

Gilchrist

(2004). Do court- and self-referred domestic violence offenders share the same characteristics? A preliminary comparison of motivation to change, locus of control and anger. Legal and Criminological Psychology, 9, 279-294.

Bowen

Gilchrist

(2006). Predicting drop-out of court mandated treatment in a British sample of domestic violence offenders. Psychology, Crime & Law, 12, 573-587.

Chambers

Eccleston

Day

Ward

Howells

(2008). Treatment readiness in violent offenders: The influence of cognitive factors on engagement in violence programs. Aggression and Violent Behavior, 13, 276-284.

10.

Correctional Service Canada. (2001). Assessment manual for the moderate and high intensity family violence prevention programs. Ottawa, ON: Reintegration Programs Division. Revised 2006.

11.

Cortoni

Nunes

Latendresse

(2006). An examination of the effectiveness of the Violence Prevention Program (Research Report No. R-178). Ottawa, ON: Correctional Service of Canada.

12.

Cunningham

Jaffe

P. G.

Baker

Dick

Malla

Mazaheri

Poisson

(1998). Theory-derived explanations of male violence against female partners: Literature update and related implications for treatment and evaluation. London, Ontario, Canada: London Family Court Clinic.

13.

Daniels

J. W.

Murphy

C. M.

(1997). Stages of change in batterers’ treatment. Cognitive and Behavioral Practice, 4, 123-145.

14.

Davis

R. C.

Taylor

B. G.

(1999). Does batterer treatment reduce violence? A synthesis of the literature. Women & Criminal Justice, 10, 69-93.

15.

Day

Chung

O’Leary

Carson

(2009). Programs for men who perpetrate domestic violence: An examination of the issues underlying the effectiveness of intervention programs. Journal of Family Violence, 24, 203-212.

16.

Dobash

R. P.

Dobash

R. E.

Cavanagh

Lewis

(1999). A research evaluation of British programmes for violent men. Journal of Social Policy, 28, 205-233.

17.

Dutton

D. G.

Corvo

(2006). Transforming a flawed policy: A call to revive psychology and science in domestic violence research and practice. Aggression and Violent Behavior, 11, 457-483.

18.

Dutton

D. G.

Hart

S. D.

(1992). Risk markers for family violence in a federally incarcerated population. International Journal of Law and Psychiatry, 15, 101-112.

19.

Eckhardt

Holtzworth-Munroe

Norlander

Sibley

Cahill

(2008). Readiness to change, partner violence subtypes and treatment for partner assault. Violence and Victims, 23, 446-475.

20.

Ellis

(2001). Overcoming destructive beliefs, feelings, and behaviors. Amherst, NY: Prometheus Books.

21.

Feder

Wilson

D.B.

(2005). A meta-analytic review of court-mandated batterer intervention programs: Can courts affect abusers’ behavior? Journal of Experimental Criminology, 1, 239-262.

22.

Garcia-Moreno

Heise

Jansen

H. A. F. M.

Ellsberg

Watts

(2006). Violence against women. Science, 310, 1282-1283.

23.

Gondolf

E. W.

(2000). Mandatory court review and batterer program compliance. Journal of Interpersonal Violence, 15, 428-437.

24.

Hamberger

L. K.

Hastings

J. E.

(1988). Skills training for treatment of spouse abusers: An outcome study. Journal of Family Violence, 3, 121-130.

25.

Hanson

R. K.

Bussiere

M. T.

(1998). Predicting relapse: A meta-analysis of sexual offender recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348-372.

26.

Hupka

R. B.

Rusch

P. A.

(2001). Interpersonal Relationship Scale. In Perlmutter

B.F.

Touliatos

Holden

G. W.

(Eds.), Handbook of family measurement techniques: Instruments and Index: Vol. 3. (pp. 148-150). Thousand Oaks, CA: SAGE.

27.

Joe

G. W.

Simpson

D. D.

Broome

K. M.

(1998). Effects of readiness for drug abuse treatment on client retention and assessment of process. Addiction, 93, 1177-1190.

28.

Kistenmacher

B. R.

Weiss

R. L.

(2009). Motivational interviewing as a mechanism for change in men who batter: A randomized controlled trial. In Murphy

C. M.

Maiuro

R. D.

(Eds.), Motivational interviewing and stages of change in intimate partner violence (pp. 25-42). New York, NY: Springer.

29.

Kropp

R. P.

Hart

S. D.

Webster

C. D.

Eaves

(1999). Spousal assault risk assessment guide: User’s manual. Toronto, Ontario, Canada: Multi-Health Systems.

30.

Kropp

R. P.

Lee

(2004). Evaluation of the Correctional Service of Canada Family Violence Prevention Programs: Final report. Vancouver, Canada: B.C. Institute Against Family Violence.

31.

Levesque

D. A.

Gelles

R. J.

Velicer

W. F.

(2000). Development and validation of a stage of change measure for men in batterer treatment. Cognitive Therapy Research, 24, 175-199.

32.

Levesque

D. A.

Velicer

W. F.

Castle

P. H.

Greene

R. N.

(2008). Resistance among domestic violence offenders: Measurement development and initial validation. Violence Against Women, 14, 158-184.

33.

Marshall

W. L.

Thornton

Marshall

L. E.

Fernandez

Y. M.

Mann

(2001). Treatment of sexual offenders who are in categorical denial: A pilot project. Sexual Abuse: A Journal of Research and Treatment, 13, 205-215.

34.

McMurran

(2002). Motivation to change: Selection criterion or treatment need? In McMurran

(Ed.), Motivating offenders to change (pp. 1-14). Chichester, UK: Wiley.

35.

Miller

W. R.

Rollnick

(2002). Motivational interviewing: Preparing people for change (Vol. 2). New York, NY: Guilford

36.

National Institute of Justice. (2000). U.S. Department of Justice, Office of Justice Programs. www.ojp.usdoj.gov/nij/topics/crime/intimate-parnter-violence/extent.htm .

37.

Olver

M. E.

Stockdale

K. C.

Wormith

J. S.

(2011). A meta-analysis of predictors of offender treatment dropout and its relationship to recidivism. Journal of Consulting and Clinical Psychology, 79, 6-21.

38.

Pence

Paymar

(1993). Education groups for men who batter: The Duluth model. New York, NY: Springer.

39.

Prochaska

J. O.

DiClemente

C. C.

(1982). Transtheoretical therapy. Toward a more integrated model of change. Psychotherapy: Theory, Research & Practice, 19, 276-288.

40.

Prochaska

J. O.

Levesque

D. A.

(2002). Enhancing motivation of offenders at each stage of change and phase of therapy. In McMurran

(Ed.), Motivating offenders to change (pp. 57-74). Chichester, UK: Wiley.

41.

Rosenfeld

B. D.

(1992). Court-ordered treatment of spouse abuse. Clinical Psychology Review, 12, 205-226.

42.

Sartin

R. M

Hansen

D. J.

Huss

M. T.

(2006). Domestic violence treatment response and recidivism: A review and implications for the study of family violence. Aggression and Violent Behavior, 11, 425-440.

43.

Saunders

D. G.

(2008). Group interventions for men who batter: A summary of program descriptions and research. Violence and Victims, 23, 156-172.

44.

Stith

Smith

Penn

C. E.

Ward

Tritt

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

45.

Tjaden

Thoennes

(2000). Full report of the prevalence, incidence and consequences of violence against women. Washington, DC: U.S. Department of Justice.

46.

White

R. J.

Gondolf

E. W.

Robertson

D. U.

Goodwin

B. J.

Caraveo

L. E.

(2002). Extent and characteristics of woman batterers among federal inmates. International Journal of Offender Therapy and Comparative Criminology, 46, 412-426.