DCJ COG GROUPS: CLIENTS EXITING IN 2004 FIRST EPISODE

1
DCJ COG GROUPS CLIENTS EXITING IN 2004 FIRST
EPISODE

• Charlene Rhyne, PhD
• Research and Evaluation Unit
• May 2006

2
Overall Sample (n300)

• 80 Male 20 Female
• 96 H/M 4 L/Ltd
• 69 Caucasian
• 23 African-American
• 3 Native American, Hispanic and Asian

• Average
• Pre-arrest .43
• Post-arrest .44
• Home Visit .19
• Office Visit 3.92
• Phone Call 1.50
• Face-2-Face 4.49

3
Three Group Analysis
4
Exit Type Group Demographics
5
Findings

• Lower percentage of African-Americans in
  completer group
• Higher percentage of person-to-person crimes in
  completer group
• Higher percentage of status crimes in
  non-completer group

6
Exit Type Group Demographics
significant differences between completers
and non-completers at the .05 level
7
Findings

• Significant differences in expected direction in
  Office Visits, Face-to-Face, Length of Stay and
  Total Group Hours
• Significant differences in pre/post arrest in
  comparison group only
• Significant differences in Refer2start with
  Completers entering later than Non-completers

8
Pre/post Arrest Patterns
9
Arrest Change by Direction

• Decrease group 61 fewer arrests overall
• Increase group 66 more arrests overall
• No change group 138 had no arrests 13 had one
  arrest

10
One Year Pre/Post Arrest Change Group
Demographics
11
Findings

• Higher percentage of males in the increase group
• Higher percentage of PPO referrals in the
  increase group

12
Averages by Post-arrest pattern
significant differences between decrease and
increase groups at the .05 level
13
Findings

• Significant differences in LOS in the expected
  direction
• Decrease group had significant difference in
  pre/post arrest changes
• Increase group had significant difference in
  pre/post arrest changes
• No change group had significantly more total COG
  group hours

14
DRC Dosage and Post-Arrest Pattern
Significant differences between decrease and
increase group at the .05 level
15
COG Dosage by Post-arrest Pattern
Significant differences between decrease and
increase groups at the .05 level
16
Correlations and Effect Size

• Difference between statistical significance and
  effect size
• Correlation between PPO, DRC and Cog dosage and
  post-arrest
• No significant correlations with the exception of
  Basic1.
• Effect size is small.

17
Program Targets for Improvement

• Manualize cog curriculum such that it can be
  replicated
• The following programs have shown to be
  effective
• Reasoning and Rehabilitation (RR)
• Moral Reconation Therapy (MRT)
• Thinking for a Change (T4C)
• Identify and implement a pre/post measure of
  change

18
Program Targets for Improvement

• Standardize number of cog hours/group required
  for success
• For successful completers,
• Basic 1 Basic 3 hours completed ranged from 3
  13.50
• Summit hours completed ranged from 7.50 13.50
• Women hours completed ranged from 10.50 13.50
• Remember, cog works best in conjunction with
  other programming