Attitudinal Change in Participants of Partner Assault Response (PAR) Programs: A Pilot Project


The current study describes the creation and examination of self-report measures of attitudes and knowledge relevant to investigation of men's [1] progress through Ontario's Partner Assault Response programs.

Four measures were created and are recommended for further research and use. These include a:

A 10-item assessment of men's attitudes towards intervention that was created in the current study is not recommended for use as an indicator of program efficacy, but could be used to examine moderating effects of men's approach to intervention.

A 17-item measure of men's knowledge of abusive behaviour was created and may be considered for evaluation of program efficacy if perfect or near perfect knowledge of abuse is expected of PAR program participants.

Additional research is needed to determine if the created measures of attitude change are related to meaningful change in men's abusive behaviour towards their intimate partners.

Executive Summary


Specialized intervention programs have become an increasingly popular component of criminal justice and community-based services for men who have abused their intimate partners. In Ontario, these specialized programs are called Partner Assault Response (PAR) programs and are administered along with specialized Domestic Violence Courts. In 2003, approximately 7000 Ontario men received intervention through PAR programs.

Despite their popularity, there is currently little evidence that intervention programs for abusive men lead to reductions in men's assault of their intimate partners. Recent reviews of the literature suggest that, at best, these programs lead to a small reduction in rates of subsequent assault relative to non-intervention. Other possible impacts of programs for abusive men, such as impact on communities and on the quality of women's lives more broadly (e.g., feelings of safety and well-being), have been less well investigated.

In light of these results, further research is critically needed to guide the improvement of programs for men who have been abusive in their intimate relationships. In particular, research is needed to identify the factors most likely to promote change in men's abusive behaviour.

One promising area for investigation is change in men's attitudes. Three domains of men's attitudes - attitudes towards abuse, attitudes towards women, and attitudes towards intervention - show particular promise on the basis of research linking these attitudes to the development and change in abusive behaviour.

Men's knowledge is targeted by PAR intervention programs as an additional mechanism of change. Two areas of education are emphasized - men's knowledge of the range of abusive behaviours and of the cognitions, or self-talk, that supports healthy / or and abusive behaviour in intimate relationships.

The purpose of the current study was to develop self-report measures of attitudes and knowledge and to conduct pilot examination of these measures using a small sample of men attending a PAR program. Analyses compared the pre-intervention attitudes of men referred to PAR programs through Early Intervention Court, Coordinated Prosecution, and voluntary referral streams and examined change in men's attitudes over the course of PAR intervention. It was expected that this pilot work would set the foundation for an examination of whether or not there was significant change in men's attitudes over the course of PAR intervention.


Men were sequentially recruited into this study between January and March of 2004. These men completed a demographic information form and the following assessment measures that were developed for the current study:

At the end of intervention, counsellors were asked to review men's progress in group and feedback from the men's intimate partners to judge whether they were "likely," "very likely", or "unlikely" to avoid abusing their intimate partner in the future.


Pre-intervention data was gathered on 41 men. All men were referred through the PAR program via Coordinated Prosecution services (63%), Early Intervention Court (17%), and voluntary referral (20%). Consistent with other samples, there were numerous indicators that these clients were at-risk for difficulties in a variety of life domains. Approximately 25% of respondents reported that they were unemployed, 39% reported that they earned under $10,000 per year, 24% reported a history of past assaults and 13% reported drinking daily.

Men were also assessed after they completed the PAR program. Post-group assessment was available for 14 men. Post-assessment was not available for the remaining men because they either dropped out of the program prior to completion (13 men) or were lost to follow-up for a variety of reasons (14 men). Men who completed the pre and post-group assessment were similar in characteristics to men who were lost to follow-up.

Summary of Principle Findings

Factor analysis of the 78-item Abuse-Related Attitudes Assessment measure resulted in the creation of five attitude subscales:

  1. Sexist attitudes towards women, 10-items, for example "Most women fail to appreciate fully all that men do for them".
  2. Distrust of, and lack of readiness for, intervention, 10-items, for example "The only purpose of this program is to blame men for their problems".
  3. Lack of personal responsibility for abuse and its effects, 10-items, for example "My behaviour has made my partner angry but has had no lasting negative effects on her".
  4. Partner blaming, 10-items, for example "My behaviour is not nearly as bad as my partner makes it sound".
  5. Denial of expected relationship difficulties, 9-items, for example "I have never been annoyed when my partner expresses ideas very different from my own".

For the first four scales, scores are created by taking the mean of men's responses across items on a four-point agreement scale varying from "strongly agree" to "strongly disagree." For the fifth scale, extreme responses suggestive of full denial of relationship difficulties were summed across items.

Consideration of men's attitudes in these five domains pre-intervention suggested that most men had problematic attitudes in three domains - lack of personal responsibility, partner blaming, and denial of expected relationship difficulties. In contrast, most men reported non-sexist attitudes towards women and relatively positive attitudes towards intervention. There were no significant differences in the pre-intervention attitudes of men referred by the Coordinated Prosecution, Early Intervention, or voluntary processes.

Examination of men's attitudes over time suggested that at least two of the created measures were sensitive to the impact of intervention.   Those were "lack of personal responsibility for abuse and its effects" and "denial of expected relationship difficulties".

Men's responses on the Attitudes Toward Referral Incident scale were closely associated with their general attitudes towards personal responsibility for abusive behaviour.

Consideration of men's responses on the abusive behaviour knowledge scale revealed that, prior to intervention, men were accurate in the identification of about 80% of abusive behaviours. Men generally recognized that behaviours such as yelling, hitting, and forcing sex were hurtful and controlling. They were less consistent in recognizing that behaviours such as unilateral financial decision making, monitoring a partner's time or activities and threatening a partner were also abusive. Analysis of a small subsample of men suggested that over the course of intervention, men's knowledge of abusive behaviours improved to a degree closely approaching significance.

Prior to beginning intervention, men were incorrect in identifying the potential value (or danger) of one third of target thoughts. Men were most likely to recognize the value of cognitions around efficacy, such as "I don't need to defend myself, I can hear her out" and much less likely to recognize the potential danger of ruminative thoughts (e.g., "I can't believe we are having this discussion again - we just talked about this yesterday") and self-talk around entitlement (e.g., "I should not have to listen to this kind of criticism"). Over the course of intervention, men's knowledge of healthy and unhealthy cognitions did not show significant change.

There were no significant differences in the pre-intervention knowledge of men referred by the Coordinated Prosecution, Early Intervention, or voluntary processes.

Conclusions and Recommendations

There is good potential for the use of attitude measures in the evaluation of PAR programs. In particular, use of: the lack of personal responsibility for abuse and its effects; partner blaming; and denial of expected relationship difficulties subscales is recommended. The distrust of, and lack of readiness for, intervention scale is not recommended as an indicator of program efficacy, but could be used to examine the moderating effects of men's approach to intervention. For example, this scale could be used to test the hypothesis that a certain level of trust and readiness for intervention is necessary for there to be any meaningful impact of PAR intervention.

Change in knowledge, particularly knowledge about abuse-supporting thoughts, also has potential in evaluating PAR programs. If change in knowledge is investigated, the created measure of healthy and unhealthy cognitions is recommended.

The intervention needs of men attending PAR programs voluntarily, though EIC referral and through Coordinated Prosecution are not substantially different. Thus, similar attitude and knowledge measures should be useful across all referral groups.

Additional research is critically needed on the expected links between attitudes and knowledge. This work is necessary to ensure that noted changes in attitudes are predictive of changes in men's abusive behaviour and not spurious results of group participation.

Given the potential impact of abusive behaviour on victims of abuse, studies of the efficacy of PAR programs must continue to include a measure of change in behaviour, as well as measures of change in attitudes and knowledge. Inclusion of a behavioural measure, such as victim reports of continued assault, are particularly important in studies seeking to make critical decisions about service models used in the PAR system.

[1] Women were not included in this sample because the majority of PAR programs serve men.