Research Evaluation of Programmes for Abusers Rebecca Emerson Dobash Russell P. Dobash Criminology, School of Law University of Manchester

1
Research Evaluation of Programmes for Abusers
Rebecca Emerson DobashRussell P.
DobashCriminology, School of LawUniversity of
ManchesterTen years of the Abuser Programme
Practiced in ViennaVienna, Austria December
2009
2
Overview

• The transformative project
• Risk factors and violence
• Interventions for victims abusers
• Evaluations of abuser programmes
• Types- randomised quasi-experimental designs
• Results- from quasi-experimental studies
• Ongoing issues, Ongoing evaluations
• Conclusions
• evidence based knowledge policy - interventions

3
The Transformative Project

• The goal is changing violent men, improving the
  safety of women developing effective
  interventions for abusers and victims
• Three Arenas of Change
• Individual
• Beliefs behaviours
• Institutional
• Policies practices
• Cultural
• Popular beliefs practices

4
The Transformative Project

• Goal of Change
• Improving Womens Safe, Sense of Well being
  Autonomy
• Eliminating the Constellation of Violence

5
Responses to Violence Abuse

• Sanctuaries for Women
• Support
• Assistance
• Information
• Safety
• Provision, Protection, Participation Prevention
• Sanctions/Interventions for Men
• Control
• Surveillance
• Re-education
• Responsibility
• Accountability
• Positive Role for Justice System (Symbolic Real)

6
Intervention

• Effective responses must include
• -Sanctuaries for women
• -Clearly focused interventions for men
• -Comprehensive community approach

7
Abuser Programmes
8
Risk Factors - Offending Escalating/Lethal IPV

• General Criminogenic
• Unemployed
• Criminal career
• Substance abuse
• Education deficits
• Problematic social networks
• Poor emotional management
• Poor thinking skills
• Mental health problems

• Specific problems (IPV)
• Tenuous relationships (cohabiting, dating)
• Contested relationships - Prolonged conflict
• Intense Possessiveness/jealousy
• Separation/attempts to leave
• Persistent/intense harassment
• Sexual Violence
• Use of instruments/weapons
• Violence to murder victim and pervious partners
• Specialising in IPViolence
• Ordinary guys

9
Transforming Violent Men

• Why Include Men?
• What Needs Changing? Beliefs, Cognitions
  Behaviour
• Self-oriented/ narcissistic
• Objectification of woman/victim
• Violence is purposeful functional
• Violence is legitimised
• Responsibility
• - rejected and/or deflected
• Consequences
• - denied, minimised

10
Programme Content

• 1. Focus on violent/offending behaviour
• 2. Focus on attitudes and beliefs
• 3. Develop knowledge/ skills/ strategies to avoid
  further violence
• 4. Accountability - monitoring individuals
  progress
• 5. Accountability - monitoring the programme
  itself

11
Research Evaluation of Abuser Programmes
12
Do Men Change?Questions to ask?????

• Three Questions of Research Evaluation
• Effectiveness of Criminal Justice Sanctions
• Effectiveness of Abuser Programmes
• Sustainability of any change after intervention

13
Research Evaluation

• Non-equivalent control group design
• CJS Mens Programmes
• Other CJ (fines, prob, etc)
• Time 1 Men (n122) Women (n134)
• Men-Programme Group (n), Other CJ group
  (n)
• Women-Programme Group (n), Other CJ group
  (n)
• including 95 couples
• 3 Time periods
• intervention, 3mos, 12mos
• Interviews and postal Questionnaires
• Response rates
• men women
• Time 2 76 80
• Time 3 51 59

14
MEASURES

• INDEXES
• Violence Assessment Index
• Injury Assessment Indices
• Controlling Behaviour Index
• Quality of Life Index (Women)
• Quality of Life Index (Men)

15
Evaluations of Abuser Programmestypes, strengths
and limitations

• Process evaluations studies what is done, why
  complete, participants programme providers
  views programme integrity
• Randomised designs show little or no effect
• Random assignment of individuals to different
  interventions (compare experimental control
  groups and theoretically deals with all
  significant factors)
• Problems- ethics of random assignment, informed
  consent, maintaining design, little on nature of
  violence and why intervention may/may not work,
  little attention to context of violence
  intervention
• Quasi-experimental/naturalistic designsshow some
  effect
• Compare real life interventions (e.g. programmes
  probation), can study context and
  multi-dimensional assessments, easier to maintain
  design
• Problems requires comparison of groups
  outcomes, requires large samples, complex
  statistics, impossible to rule out selection
  effect
• Meta-Analysis show small effect
• Combination of many studies depends on quality
  of those studies

16
Quasi-experimental Research Design
(Viol.Men.Study)

• Control Group Design
• Comparative Longitudinal
• Two Naturally occurring groups
• Criminal Justice Interventions- Other CJ
• Court Mandated Abuser Programmes (the first in UK
  Europe)
• -CHANGE LothianDomesticViolenceProbationProject
• Pre Post Tests at 3 Time Periods
• Time 1 at intervention
• interviews with 122 abusers 132 women partners
• Time 2 after 3 mos - follow-up--postal
  questionnaire
• Time 3 after 12 mos - follow-up--postal
  questionnaire
• Baseline Assessments-
• 5 Indexes
• (violence, injuries, controlling behaviour,
  quality of life (women men)
• Data Analysis
• Assess change(s) in
• violence, injuries, controlling behaviour,
  quality of life
• findings focused on womens reports
• Selection bias - Post-hoc matching

17
Quasi-experimental Research Design
(Viol.Men.Study)

• Control Group Design
• Comparative Longitudinal
• Two Naturally occurring groups
• Criminal Justice Interventions- Other CJ
• Court Mandated Abuser Programmes (the first in UK
  Europe)
• -CHANGE LothianDomesticViolenceProbationProject
• Pre Post Tests at 3 Time Periods
• Time 1 at intervention
• interviews with 122 abusers 132 women partners
• Time 2 after 3 mos - follow-up--postal
  questionnaire
• Time 3 after 12 mos - follow-up--postal
  questionnaire
• Baseline Assessments-
• 5 Indexes
• (violence, injuries, controlling behaviour,
  quality of life (women men)
• Data Analysis
• Assess change(s) in
• violence, injuries, controlling behaviour,
  quality of life
• findings focused on womens reports
• Selection bias - Post-hoc matching

18
Some results from quasi-experimental studies

• Abuser Programmes more effective than other
  interventions
• Reductions in Violence prevalence, frequency and
  severity
• Reductions in Constellation of Abuse
• -across a range of controlling and intimidating
  behaviours
• Improvements in Quality of Life
• -men women feel happier and women feel safer
• Importance of Dosage US (length of programme)
• Importance of Context/System
• -police, courts, probation, additional sanctions,
  victim support
• -consistent messages and actions
• Repeat assaulters, difficult to identify but
  higher risk when UK unemployed, younger,
  alcohol problems, non-state sanctioned
  relationship, criminal careers, intense
  constellation of abuse (Dobash et al.)
• USA alcohol problems, severe previous assault
  criminal career, constellation of abuse (small
  with severe mental disorder and no difference in
  personality types) (Gondolf, et. al.)
• - best predictors are womens judgements mens
  drunkenness

19
Evaluations of Abuser Programmes - research
issues -

• Self-assessment by programme staff
• Programme integrity not assessed
• No outcome measures
• No control or comparison group
• Little consideration of offender characteristics
• Psychological, behavioural, ethnic
  characteristics, voluntary or court mandated,
  IMPORTANCE OF STAKE IN CONFORMITY FACTORS
• No consideration of dosage/length programme
• Small sample size, high attrition at follow-up
• Selection bias
• Short follow-ups sustainability
• Poor or singular outcome measures
• Only use arrests or self-reports of offenders
  not reports of partners

20
Ongoing Issues, Ongoing Evaluations

• Limits of Research
• ethics, methods, pragmatics and resources
• Research design
• selection bias, sample sizes, generalisability,
  what to assess/compare, use of drop-outs,
  attrition rates, arrests vs. womens assessments
  
• People changing
• difficult to achieve, difficult to evaluate
• Public policy is incremental based on evidence,
  debates and informed judgements
• Overall, there are benefits of abuser programmes
• for perpetrators, victims, communities society

21
Violence at Times 1,2 3 (womens reports)
22
Three Stories of Change

• 1. Men who cannot or will not change despite the
  intervention
• 2. Men who engage in limited change maintained
  under the watchful eye of the enforcers of law
  and the threat of increasing sanctions
• 3. Men who change their violent behaviour and
  supporting attitudes and become the regulators of
  their own behaviour

23
Why Men Change (eight stage process)

• Change is Possible
• -away from impossibility/undesirability of change
  to view as real prospect
• 2. Motivation to Change
• -must want to change
• 3. Why Change
• -recognition of cost and benefits to self and
  expand to include the other
• 4. What Changes
• -from a view self as object -acted upon
• -to self as subject - making decisions -
  (taking responsibility for actions)
• 5. Gen. Mechanisms of Change
• -external constraints to internal controls
  (from surveillance to self control)
• 6. New Discourse
• -from accept viol.,deny, min./blame to notions
  of rights,respect, responsibility
• 7. The Medium of Change
• -learning,talking,listening
• 8. Specific Elements of Change
• -cognitive and behaviour skills on mens
  programmes

24
Ongoing Issues, Ongoing Evaluations

• Limits of Research
• ethics, methods, pragmatics and resources
• Research design
• selection bias, sample sizes, generalisability,
  what to assess/compare, use of drop-outs,
  attrition rates, arrests vs. womens assessments
  
• People changing
• difficult to achieve, difficult to evaluate
• Public policy is incremental based on evidence,
  debates and informed judgements
• Overall, there are benefits of abuser programmes
• for perpetrators, victims, communities society

25
Books- Intimate Partner Violence by Dobash et al.
26
Abuser Programme Evaluation Selected Publications

• 1999, Dobash et al., A Research Evaluation of
  British Programmes for Violent Men, Journal of
  Social Policy, 28205-233.
• 2000, Dobash et al., Changing Violent Men.
  London Sage.
• 2000, Dobash Dobash, Evaluating Criminal
  Justice Interventions for Domestic Violence,
  Crime Delinquency, 46252-270.
• 2005, Dobash Dobash, Abuser Programmes
  Violence Against Women. In Smeenk and Malsch,
  Family Violence and Police Response, London
  Ashgate.

27
Intimate Partner Murder Selected Publications

• 2004, Dobash et al., Not an Ordinary Killer, Just
  an Ordinary Guy When Men Murder an Intimate
  Partner, Violence Against Wives An International
  Journal, 10577-605.
• 2007, Dobash et al., Lethal and Non-Lethal
  Violence Against an Intimate Female Partner.
  Violence Against Women, 13, 4329-353.
• 2007, Cavanagh, Dobash Dobash, The Murder of
  Children by Fathers in the Context of Child
  Abuse, Child Abuse and Neglect , 31731-746.
• 2009, Dobash et.al., Out of the Blue Men who
  murder an intimate partner, Feminist Criminology,
  4194-225.
• 2010, Dobash Dobash, What were they thinking?
  Men who murder an intimate partner, Violence
  Against Women (in press)

28
Intimate Partner Violence Selected Publications

• 1979. Dobash Dobash, Violence Against Wives,
  New York Free Press.
• 1992, Dobash Dobash, Women, Violence and Social
  Change, London Routledge.
• 1992, Schlesinger, Dobash, Women Viewing
  Violence, London.
• 1992, Dobash, et.al., The myth of symmetry in
  family violence. Social Problems, 39,1,71-91.
• 1998, Dobash Dobash, eds., Rethinking Violence
  Against Women London Sage.
• 1998, Dobash, et.al., Separate and Intersecting
  Realities, Violence Against Women a comparison
  of mens and womens accounts of violence against
  women, Violence Against Women, 4,4,382-414.
• 2001, Cavanagh, et al. Remedial work mens
  strategic responses to their violence against
  intimate female partners, Sociology,
  35,3,395-714,
• 2004, Dobash Dobash, Womens Violence Against
  an Intimate Male Partner Working on a Puzzle,
  British Jr of Criminology, vol.44,324-349.