Idea Transcript
ALAMEDA COUNTY PROBATION DEPARTMENT
Donald H. Blevins, Chief Probation Officer Presents
What Works and What Doesn’t in Reducing Recidivism: The Principles of Effective Intervention: Edward J. Latessa, Ph.D. Center for Criminal Justice Research Division of Criminal Justice University of Cincinnati www.uc.edu/criminaljustice
TERMS:
What Works and What Doesn’t in Reducing Recidivism: The Principles of Effective Intervention:
• Evidence: Refers to results from controlled studies, involving distinguishing between experimental groups and control or comparison groups.
P Presented t d by: b Edward J. Latessa, Ph.D. Center for Criminal Justice Research Division of Criminal Justice University of Cincinnati www.uc.edu/criminaljustice
• Risk: Refers to risk of reoffending. Recidivism rates are compared over a standard and specified follow-up period.
Misapplication of Research “XXX Study Says”
FROM THE EARLIEST REVIEWS:
If you believe every study we wouldn’t eat anything (but we would drink a lot of red wine!)
• Not a single reviewer of studies of the effects of official punishment (custody, mandatory arrests, probation, increased surveillance, etc.) has found consistent evidence of reduced recidivism. recidivism
• Looking at one study can be a mistake • Need to examine a body of research • Three ways researchers summarize research: 1. Traditional literature review 2. Ballot Counting 3. Meta-Analysis
• At least 40% and up to 60% of the studies of correctional treatment services reported reduced recidivism rates relative to various comparison conditions, in every published review.
• Meta-analysis is now the favored approach to conducting a sytematic review of research studies. Meta-analysis provides a quantitative review of a body of literature
Criminal Sanctions vs Treatment for Youthful Offenders
Criminal Sanctions versus Treatment Number of studies=54
Mean Phi
Number of studies=175
0.14
0.2
0.12
0.15
0.1
Reduced Recidivism
0.08
0.1
0.15 0.05
0.02
0 Increased Recidivism
-0.05
0.06 0.04
0
-0.07
-0.02 -0.04
-0.1
Criminal Sanctions
CS -.07 (Number of Studies=30)
Treatment .15 (Number of Studies=124)
Yes
-0.02
Treatment 0.13
Source: Dowden and Andrews (1999), What Works in Young Offender Treatment: A Meta Analysis. Forum on Correctional Research.
1
People Who Appear to be Resistant to Punishment • The Evidence Indicates that Treatment is More Effective in Reducing Recidivism than Punishment – But Not All Treatment Programs Are Equally Effective
• Psychopathic risk takers • Those under the influence of a substance • Those with a history of being punished
Type of Treatment and Effect Sizes for Youthful Offenders
Behavioral vs. NonBehavioral
0.25
0.35 0.29
0.3
0.2
Reduced 0.25 Recidivism 0.2
0.15 Reductions in Recidivism
0.15 0.1
0.1
0.07 0.05
Increased 0.05 Recidivism 0
Nonbehavioral (N=83)
0
Behavioral (N=41)
Non-Behavioral Effect Size
Andrews, D.A. 1994. An Overview of Treatment Effectiveness. Research and Clinical Principles, Department of Psychology, Carleton University. The N refers to the number of studies.
Behavioral 0.24
Source: Dowden and Andrews (1999), What Works in Young Offender Treatment: A Meta Analysis. Forum on Correctional Research.
Community Based versus Institutional Programs: Results from Meta-Analyses of Programs Based on Principles of Effective Treatment
Meta-Analysis of Treatment for Females by Dowden and Andrews Mostly Female
0.04
Only Female
0.4
0.4
0.36
0.35
0.34
0.35
0.25
0.2
0.17
Reduction in Recidivism
0 25 0.25 0.18
0.15
0.2 0.15
0.17
0.1
0.1 0.05
0.35
0.3
0.3
0.1 0.02
0.03
0.03
0.05
0 Inappropriate
Weak
Promising
Behavioral
0 Community Based
Dowden, C., and D. Andrews (1999). What Works for Female Offenders: A Meta-Analytic Review. Crime and Delinquency, Vol. 45 No. 4.
Institutional
Source: Gendreau, P., French, S.A., and A. Taylor (2002). What Works (What Doesn’t Work) Revised 2002. Invited Submission to the International Community Corrections Association Monograph Series Project.
2
Factors Correlated With Risk
Another important body of knowledge to understand is the research on risk factors
Mean r
What are the risk factors correlated with criminal conduct?
# of studies
Lower class origins
0.06
97
Personal distress/psychopathology
0.08
226
Educational/Vocational achievement
0 12 0.12
129
Parental/Family Factors
0.18
334
Temperament/misconduct/personality
0.21
621
Antisocial attitudes/associates
0.22
168
Note: A re-analysis of Gendreau, Andrews, Goggin & Chanteloupe (1992) by Andrews & Bonta (1994)
Meta-Analysis of Risk Factors by Simourd Mean Adjusted r Risk Factor
Adjusted R
Correlates of Criminal Conduct and Gender by Simourd and Andrews
#Studies
Factor
Male
Female
Lower social class
.05
38
Lower class origins
.04(58)
.03(12)
Personal distress/psychopathy
.07 07
34
Personal distress/psychopathology
.09(157) 09(157)
.08(19) 08(19) .13(7)
Family structure/parental problems
.07
28
Personal education/vocational achievement
.11(96)
Minor personality variables
.12
18
Parental/family factors
.16(180)
.16(43)
Poor parent-child relations
.20
82
Temperament/misconduct/personality
.18(461)
.23(38)
Antisocial attitudes/associates
.21(113)
.23(12)
Personal educational/vocational achievement
.28
68
Temperament/misconduct/self control
.38
90
Antisocial attitudes/associates
.48
106
Source: Simourd, L. (1993) Correlates of Delinquency: A Look at Gender Differences. Forum on Correctional Research. 6:26-31
Simourd, L., and D.A. Andrews, 1994. Correlates of Delinquency: A Look at Gender Differences. Forum on Corrections Research, Vol. 6: 26-31
Meta-Analysis of the Predictors of Female Delinquency by Hubbard and Pratt (2001)
Simourd and Andrews Mean Adjusted r by Gender
Factor Risk Factor
Females
Males
Lower social class
.07
.06
Personal Distress/psychopathy
.10
.09
Family structure/parental problems
.07
.09
Minor personality variables
.18
.22
Poor parent-child relations
.20
.22
Personal educational/vocational achievement
.24
.23
Temperament or misconduct problems
.35
.36
Antisocial attitudes/peers
.39
.40
Mean Effect Size
Socio-economic status
.03
Anxiety
.06
Self Image
.13
Family Relationships
.17
Antisocial Attitudes/Beliefs
.18
Physical or Sexual Abuse
.21
Antisocial Personality
.21
School Relationships
.25
History of Antisocial Behavior
.48
Antisocial Peers
.53
Source: Simourd, L., and D.A. Andrews (1994) Correlates of Delinquency: A Look at Gender Differences. Forum on Correctional Research. 6:26-31
3
Major Set of Risk/Need Factors • Research by Andrews, Gendreau and others has led to the identification of some major risk/need factors
1. Antisocial/prociminal attitudes, g values, beliefs and cognitiveemotional states
Identifying Procriminal Attitudes, Values & Beliefs Procriminal sentiments are what people think, not how people think; they comprise the content of thought, not the skills of thinking.
Neutralization & Minimizations Offenders often neutralize their behavior. Neutralizations are a set of verbalizations which function to say that in particular situations, it is “OK” to violate the law
What to listen for:
Neutralization Techniques include:
• Negative expression about the law
• Denial of Responsibility: Criminal acts are due to factors beyond the control of the individual, thus, the individual is guilt free to act.
institutions values, values rules, rules & • Negative expression about conventional institutions, procedures; including authority • Negative expressions about self-management of behavior; including problem solving ability • Negative attitudes toward self and one’s ability to achieve through conventional means • Lack of empathy and sensitivity toward others
Major set Risk/needs continued: 2. Procriminal associates and isolation from anticriminal others
• Denial of Injury: Admits responsibility for the act, but minimizes the extent of harm or denies any harm • Denial of the Victim: Reverses the role of offender & victim & blames the victim • “System Bashing”: Those who disapprove of the offender’s acts are defined as immoral, hypocritical, or criminal themselves. • Appeal to Higher Loyalties: “Live by a different code” – the demands of larger society are sacrificed for the demands of more immediate loyalties.
Major set Risk/Needs continued: 3. Temperamental and personality factors conducive to criminal activity including: – – – – – – –
Psychopathy Weak Socialization Impulsivity Restless Aggressive Energy Egocentrism Below Average Verbal intelligence A Taste For Risk Weak Problem-Solving/Self-Regulation Skills
4
Psychopathy Checklist (Hare Psychopathy) • • • • • • • • • • • • • • • • • • • •
Glib/superficial charm Grandiose sense of self Stimulation seeking Pathological lying Conning/manipulation Lack of remorse/guilt Shallow affect Callousness/lack empathy Parasitic lifestyle Poor behavioral control Promiscuous sexual behavior Early behavioral problems Lack of realistic goals Impulsivity Irresponsibility Not accepting responsibility Many marital relationships Juvenile delinquency Conditional release revoked Criminal versatility
Personality Factors • Studies are finding that personality is related to criminal behavior • Criminals are more likely to be characterized as negative ti or hostile h til in i interpersonal i t l relationships, unempathetic, and lacking in self control • Personality is most likely working in tandem with other risk factors such as peers and attitudes
History of Antisocial Behavior Major set of Risk/Need factors continued:
4. A history of antisocial behavior: – Evident from a young age – In a variety of settings – Involving a number and variety of different acts
Lifecourse studies indicate that: • By age 12, up to 40% of later serious offenders have committed their first criminal act • B By age 14 14, up tto 85% have h committed itt d th their i fi firstt criminal act • Variety of settings including home, school, streets • Escalating behavior
Major set of Risk/Needs Continued:
Major set of Risk/Needs continued:
5. Family factors that include criminality and a variety of psychological problems in the familyy of origin g including: g
6. Low levels of personal educational, vocational or financial achievement
– – –
Low levels of affection, caring and cohesiveness Poor parental supervision and discipline practices Out right neglect and abuse
5
Principles of Effective Intervention
This research has led to the identification of some principles p p
• Risk Principle – target higher risk offenders (WHO) • Need Principle – target criminogenic risk/need factors ( (WHAT) ) • Treatment Principle – use behavioral approaches (HOW) • Fidelity Principle – implement program as designed (HOW WELL)
25 Change In Recidivism Rates
Risk Principle
The Risk Principle & Correctional Intervention Results from Meta Analysis
• Target those offender with higher probability of recidivism • Provide most intensive treatment to higher risk offenders • Intensive treatment for lower risk offender can increase recidivism
20
19
15 10 5 -4 0 -5 High Risk
Low Risk
Dowden & Andrews, 1999
The Risk Principle and Lower Risk Youth: Results from Ohio Study of over 10,000 Youth
60
60 48
50 Recidiv vism Rates
Recent Study of Intensive Rehabilitation Supervision in Canada
50
40
40
40 27
30
30
16
20 9
30 20 10
10
0
0 Commitment Commnity
Conviction Residential
N-10,594, Community N= 8,582, Residential N= 500, Institutional N= 1,502
Institutional
High Risk Low Risk
Treatment
Non-Treatment
31.6 32.3
51.1 14.5
Bonta, J et al., 2000. A Quasi-Experimental Evaluation of an Intensive Rehabilitation Supervision Program., Vol. 27 No 3:312-329. Criminal Justice and Behavior
6
RECENT STUDY OF COMMUNITY CORRECTIONAL PROGRAMS IN OHIO
Reanalysis of RAND ISP Data
Percentage Rearrested based on Level of Risk & Appropriateness of Supervision & Services 60
• Largest study of community based correctional treatment facilities ever done
50 40
• Total of 13,221 offenders – 37 Halfway Houses and 15 Communityy Based Correctional Facilities ((CBCFs)) were included in the study.
30 20
• Two-year follow-up conducted on all offenders 10
• Recidivism measures included new arrests & incarceration in a state penal institution
0 Arrested - Appropriate Arrested Not Appropriate
High
Low
32.6 53.4
19.1 24.6
Hanley, D. (2002). Risk Differentiation and Intensive Supervision: A Meaningful Union? Doctoral Dissertation, University of Cincinnati.
• We also examined program characteristics
Experimental Groups:
• 3,737 offenders released from prison in FY 99 and placed in one of 37 Halfway Houses in Ohio
Determination of Risk
• 3,629 offenders direct sentenced to one of 15 CBCFs
• Each offender was given a risk score based on 14 items that predicted outcome.
Control Group: • 5,855 offenders released from prison onto parole supervision during the same time period
• This allowed us to compare low risk offenders who were placed in a program to low risk offenders that were not, high risk to high risk, and so forth.
• Offenders were matched based on offense level & county of sentence
Treatment Effect For Any Incarceration: Low Risk Offenders
Treatment Effects for Low/Moderate Risk Offenders 15
8
10
0 0
-2 -2 -2 -2 -5
-7 -7 -10
-11-11 -11 -15 15
-15
-20
-21 -21 -21 -21 -25 -30
-29
-35
-36 -40
-6
-5 -5
-4 -4 -4
-1
1
1
2
3
3
5
9
6
9
10
Probability of R Reincarceration
5
4
5
0
1
1
1
2
3 3
4
4
5
5
6
10 10
11 11
7
0 -5
-4 -7
-10
-12 -15
-3
-2
-1 -1
-6
-10 -10 -11 11
-14 -14
-20
-19 -25
-23 -23
-30 -35
-36 -40
7
Treatment Effect For Any Incarceration: High Risk Offenders
Treatment Effects for Moderate Risk Offenders 20
18
40
32
19
10
8
5 0
1
1 1 1
2
3
4 4
5 5
5
6
6 6
8
9
9
13 13
Probability of R Reincarceration
Probability of R Reincarceration
12 10
6
0 -2
5 -5 -7
-1
-6
-10 -15
-14
-20
-13
-20
21 20
12 12 12 10
5 2
3
6
7
8 8
13 13 13
22
24
25
27
15
9 10 10
3 3
0
-2 -2 -6 -6
-10
-14 -15 -20
-18 -30
-25 -30
34
30 30
15
-26 -28
-34 -40
Targeting Criminogenic Need: Results from MetaAnalyses
Need Principle By assessing and targeting criminogenic needs for change, agencies can reduce the probability of recidivism
0.35 0.3
Criminogenic
Non-Criminogenic
• • • • •
• • • • •
Anti social attitudes Anti social friends Substance abuse Lack of empathy Impulsive behavior
Reduction in Recidivism
Anxietyy Low self esteem Creative abilities Medical needs Physical conditioning
0.25 0.2 0.15 0.1 0.05
Increase in Recidivism
0 -0.05
Target 1-3 more noncriminogenic needs
Target at least 4-6 more criminogenic needs
Source: Gendreau, P., French, S.A., and A.Taylor (2002). What Works (What Doesn’t Work) Revised 2002. Invited Submission to the International Community Corrections Association Monograph Series Project
Needs Targeted & Correlation with Effect Size for Youthful Offenders 0.4
Treatment Principle
0.3 Reduced Recidivism
0.2
The most effective interventions are behavioral:
0.1 0 Increased Recidivism
• Focus on current factors that influence behavior
-0.1 01 Bond Anti Social Peers
Target Self-Esteem
Vague Emotional Problems
Respect Anti Social Thinking
Physical Activity
Criminogenic Needs
Effect Size
Fear of Punishment
-0.2
-0.18
-0.12
-0.09
-0.06
-0.05
-0.03
0.36
• Action oriented • Offender behavior is appropriately reinforced
Source: Dowden and Andrews, (1999). What Works in Young Offender Treatment: A Meta Analysis. Forum on Correctional Research. Correctional Services of Canada
8
Relationship between Treatment Model and Treatment Effect for Residential Programs 9
6 4 2 0 -2 -4
Cognitive Behavioral
Other
-6 -8 -10
% Change in Reccidivism
% Change in Recid divism
10 8
Why practice? Relationship between Treatment Activities and Treatment Effect for Residential Programs 15 13 11 9 7 5 3 1 -1 -3 -5
13
6 2
-4 Practice
Role Playing -9
Yes
No
Effective programs have certain characteristics: • • • • • •
Are based on research & sound theory Have leadership Assess offenders using risk &need assessment instruments Target crime producing behaviors Use effective treatment models Vary treatment & services based on risk, needs, & responsivity factors Disrupt criminal networks Have qualified, experienced, dedicated & educated staff Provide aftercare Evaluate what they do Are stable & have sufficient resources &support
• • • • •
Some so called “theories” we have come across •
“Been there done that theory”
•
“Offenders lack creativity theory”
•
“Offenders need to get back to nature theory”
•
“Offenders lack discipline theory”
•
“Offenders lack organizational skills theory”
•
“Offenders have low self-esteem theory”
•
“Offenders need to change their diet theory”
•
“Treat them as babies & dress them in diapers theory”
•
“We just want them to be happy theory”
•
“Offenders (females) need to learn to put on makeup & dress better theory”
•
“Male offenders need to get in touch with their feminine side theory”
• Many correctional intervention programs are based on tradition, custom, & imitation rather than scientific evidence of effectiveness
Effective Programs are Based on Theory and Research • Program development includes extensive literature review • There is theoretical foundation to the program and its components • The interventions are linked to criminogenic needs • The staff understands the interventions, why they are being used, and how to apply them
9
Some Goals of Assessment
The Evolution of Classification • First generation – “Gut Feelings”
1.
To identify risk of recidivism
2.
To identify appropriate offenders for programs
3.
To identify criminogenic needs
4.
To identify factors that can affect program success
5.
To provide risk & need levels that will facilitate development of case plan
6.
To facilitate reassessment of offender to determine which risk & need factors have changed Ideally a process will be utilized that allows for all of these goals to be accomplished.
Factors in the Burgess Scale:
• Second generation – Primarily historic or static predictors, e.g. Burgess Scale • Third generation – Incorporates dynamic and static factors, e.g. Level of Service Inventory
Classification & Assessment of Offenders
General type of Offense (e.g. fraud, robbery, sex, homicide) Parental & Marital Status (parents living, offender married) Criminal Type (first timer, occasional, habitual, professional) Social Type (e.g. farm boy, gangster, hobo, drunkard, Ner’do well) Community Factor (where resided) Statement of Trial Judge & Prosecutor (recommend or protest leniency)
Previous Record Work Record (e.g. no work record, casual, regular work) Punishment Record in Prison Months Served Prior to Parole Intelligence Rating Age when Paroled Psychiatric Prognosis Psychiatric personality type (Egocentric, Socially Inadequate, Unstable)
• Primary measures have been identified • Best predictors of criminal behavior: Static factors – past criminal behavior Dynamic factors – crime producing needs
• Best assessment method is the actuarial (statistical) approach • Best practices allow for risk management and risk reduction through effective treatment • Latest generation of instruments allow for measurement of change in offender
Comparison of Clinical vs. Statistical Prediction of Recidivism
Comparison of Clinical vs. Statistical Prediction of Sex Offenders
0.24 0.22
0.5
0.2 r value
0.18 r value
0.16
0.4
0 14 0.14 0.3
0.12
Clinical Statistical
0.1 0.2
0.08 0.06
0.1
0.04 0.02
0
0 General Recidivism
Clinical
Statistical
0.08
0.22
Goggin, C.E. (1994). Clinical versus Actuarial Prediction: A Meta-analysis. Unpublished manuscript. University of New Brunswick, Saint John, New Brunswick.
Clinical Statistical
Sexual Recidivism
0.1 0.46
Nonsex violent recidivism
0.06 0.46
General Recidivism
0.14 0.42
Hanson, R. K. and M. T. Bussiere, 1998. Predicting Relapse: A Meta Analysis of Sexual Offender Recidivism Studies. Journal of Consulting and Clinical Psychology, 66: 348-362.
10
Youthful LSI: Assessment and Reassessment
Statewide LSI-R Adult Offender Data for Community Corrections
Total Score Initial = 19.4
LSI-1
Reassess = 16.98
LSI-2
25
3.5 3
20
2.5 15
2 1.5
Initial Discharge
10
1 0.5
5
Attitudes
Person
Leisure
SA
Peers
Educ
Family
Offenses
0
0 State Wide (19,344) Initial Discharge
Residential/WR (4,728)
Day Reporting (2,056)
Home Detention (6,909)
19.44
18.76
21.4
18.15
14.95
15.28
14.4
14.87
Total of 84 youth have been reassessed
Effective programs assess offenders: • Program has screening criteria • Offenders are assessed on all major risk, need & responsivity factors • Assessment process is objective and standardized
Responsivity refers to learning style and characteristics of the offender, which can effect their engagement in treatment
• Levels of risk, need & responsivity are determined by assessment process • Instruments are normed and validated
Responsivity areas to assess can include • • • • • •
Motivation to change Anxiety/psychopathy Levels of psychological development Maturity Cognitive functioning Mental disorders
Responsivity Factors General Population • Anxiety • Self esteem • Depression • Mental illness • Age • Gender • Race/Ethnicity
• • • •
Offenders Poor social skills Inadequate problem solving Concrete oriented thinking Poor verbal skills
11
Maximizing the Assessment Process
Some Common Problems with Offender Assessment
• View assessment as a process not a “one time” activity • Develop a flexible process that expands as needed • Standardize process and instruments • Make sure the assessment is accurate & correct errors • Make sure staff correctly interpret the results • Develop p case supervision p & treatment p plan from the assessment results • Audit assessments on a regular basis • Train and retrain staff • Assign offenders to programs/groups based on assessment information • Share information with service providers • Reassess offenders periodically • Collect data & analyze
¾ Assess offenders but process ignores important factors ¾ Assess offenders but don’t distinguish levels (high, moderate, low) ¾ Assess offenders then don’t don t use it – everyone gets the same treatment ¾ Make errors and don’t correct ¾ Don’t assess offenders at all ¾ Do not adequately train staff in use or interpretation ¾ Assessment instruments are not validated or normed
• Periodically validate instruments with your population
Most Effective Behavioral Models • Structured social learning where new skills and behavioral are modeled • Cognitive behavioral approaches that target criminogenic risk factors • Family based approaches that train family on appropriate techniques
The Four Principles of Cognitive Intervention 1. Thinking affects behavior 2. Antisocial, distorted, unproductive irrational thinking causes antisocial and unproductive behavior 3. Thinking can be influenced
Social Learning Refers to several processes through which individuals acquire attitudes, behavior, or knowledge from the persons around them. Both modeling and instrumental conditioning appear to play a role in such learning
Recent Meta-Analysis of Cognitive Behavioral Treatment for Offenders by Landenberger & Lipsey (2004)* • Reviewed 79 studies: 26 random samples 27 matched samples 26 convenience samples • Found that on average CBT reduced recidivism by 35%
4. We can change how we feel and behave by changing what we think
12
Factors Not significant: • • • • • •
Type of research design Setting - prison versus community Juvenile versus adult Minorities or females Total hours Brand name
Significant Findings (effects were stronger if): • • • • • •
Some Examples of Cognitive Behavioral Correctional Curriculums
Effects based on Cognitive targets: • • • •
Cognitive restructuring Anger control Interpersonal problem solving Substance Abuse
.18 .35 .26 26 .22
*Landenberger, N, and M. Lispey (2004). Cognitive Behavioral Treatment for Offenders. Paper presented at the American Society of Criminology annual meeting, Nashville, TN.
Cognitive Behavioral Approaches Based on Social Learning Theory
At least 2 sessions per week (.25) Smaller groups (.20) Implementation p monitored ((.25)) Staff trained on CBT (.27) Higher proportion of treatment completers (.25) Higher risk offenders (.38)
•
Controlling Anger and Learning to Manage It (CALM and CALMER)
•
Aggression Replacement Therapy (ART)
•
Criminal Conduct and Substance Abuse Treatment (Strategies for Self-Improvement and Change) (SSC)
•
Thinking for a Change (T4C)
•
Choices, Changes & Challenges
•
Persistently Violent Curriculum
•
Corrective Thinking/Truthought
•
Reasoning and Rehabilitation
•
Moral Reconation Therapy
•
Drug Abuse Treatment Program (FBOP)
•
Moving On (Female Offenders)
Treatment should be Behavioral in Nature • Use rewards and punishers effectively
Cognitive Restructuring (What we think content)
Cognitive Skills Development
• Train, practice, rehearse offenders in prosocial alternatives
(How we think: process)
Behavioral Strategies
• Completion criteria should be based on acquisition of prosocial skills
(Reinforcement and modeling prosocial behavior
13
For a new behavior to occur one must: • Have a strong positive intention to perform the behavior • Have the skills necessary to carry out the behavior, and • Be in an environment that is free of constraints such that the behavior can occur
Effective Modeling • Demonstrate behavior • Specify the rewards for behaving this way - What do most people gain in the short & long term? - What can the person expect to gain? • Provide reinforcement each and every time the person behaves in the desired way
Maximizing the Effectiveness of Rewards • Reinforcement is most effective when it comes immediately after the behavior, however, this is not always practical. You can make a promise of delayed reinforcer (IOU). • Remember, vary reinforcers since they will lose potency over time • Natural reinforcers should be used frequently (since they are likely to be received outside the program) • Consistency is very important • Rewards should outnumber punishers by 4-1. • Build rewards into program structure and train staff on use
Skill Development • Demonstrate, rehearse, practice prosocial alternatives • Increase difficulty • Completion based on acquisition of new prosocial skills
One way is to structure groups around the quarter rule • First ¼ spent reviewing what they learned last time • Second ¼ demonstrate new skill • Third ¼ practice new skill • Fourth ¼ make practice more difficult
Punishers • Designed to extinguish inappropriate behavior • Most effective are response cost (i.e. (i e losing privileges, and disapproval)
14
Maximizing the Effectiveness of Punishers
What Doesn’t Work with Offenders?
• • • •
Escape should be impossible Should be applied immediately Should be applied at maximum intensity Should be applied after every occurrence of deviant behavior • Should not be spread out & should be varied • Remember, a punisher only trains a person what not to do: must also teach prosocial alternative • When punishment is inappropriately applied several negative consequences can occur (unwanted emotional reactions, aggression, withdrawal, or increased behavior that is being punished)
Average Effects of Punishing Smarter Programs on Recidivism: Results from Meta Analyses
Ineffective Approaches
2
• Drug prevention classes focused on fear and other emotional appeals • Shaming offenders • Drug education programs • Non-directive, client centered approaches • Bibliotherapy • Freudian approaches • Talking cures • Self-Help programs • Vague unstructured rehabilitation programs • Medical model • Fostering self-regard (self-esteem) • “Punishing smarter” (boot camps, scared straight, etc.)
% Recidivism Reduced
0 -2 -4
% Recidivism Increased
-6 -8 -10 -12 %
Fines
Drug Testing
Restitution
Elec Monit
Scared Straight
ISP
Juv Boot Camps
0.4
0
-3
-3
-4
-6
-11
Sources: Gendreau et al (2000) The Effects of Community Sanctions and Incarceration on Recidivism, FORUM; Aos et al (1999) The Comparative Costs and Benefits of Programs to Reduce Crime, Washington State Institute for Public Policy.
Quality Assurance
Effective programs evaluate what they do: • Quality assurance processes (both internal and external • Assess offenders in meeting target behaviors • Track offender recidivism • Have an evaluator working with the program
1.
2.
Internal – processes to ensure that assessments, services & interventions provided by the program are delivered as designed External – processes to ensure services and interventions provided by outside providers are delivered as designed
Can include: • Case file audits • Video taping groups • Client satisfaction surveys/ exit interviews • Clinical supervision • Program audits • Site visits and observation • Certification process
15
Effects of Quality Programs Delivery for Evidenced Based Programs for Youth Offenders
35
30
30 ony Recidivism 12 Month Felo
Reduced Recidivism
Therapist Competency Ratings and Recidivism
40
20
10
0 Increased Recidivism
-10
29 25
25
22
20
17 14
15 10 5
-20 Functional Family Therapy
Aggression Replacement Therapy
38 -16.7
24 -10.4
Competently Delivered Not Competent
0 Not Competent
Competent
Highly Competent Control Group
Impact of Program Factors Predicting Felony Adjudication for Juvenile Programs
Effect of Program Integrity on Recidivism: Results from Meta Analysis Andrews and Dowden 1999
50 45 Percent Change in Recidivism
Marginal
Source: Outcome Evaluation of Washington State's Research-Based Programs for Juvenile Offenders. January 2004. Washington State Institute for Public Policy.
Source: Outcome Evaluation of Washington State's Research-Based Programs for Juvenile Offenders. January 2004. Washington State Institute for Public Policy.
70 Predicted Recidivism Rates
40 33 28
30 23
22
22
20
20 12 10
7
5
10
9
7
8
0
60
60
53 47
50
43
40
36 31
27
30
22 18
20 11 8
10
7
ed lv
Low
e ag os
e
he rc ea es
D
R
e at
ng ha
ls ua
Program Score 0
r
rs ke or W
an
C
M
ed
el
rs ke or W
od M
vis
u eq
r ito on
d te in
r pe
vo In
Ad
M
Pr
Su
ific ec
ed ain Tr
Sp
0
Average Reduction in Recidivism by Probation/Community Corrections Program Assessment Score
Moderate
High
Program Score 12
Very High
Program Score 24
Changes in Recidivism by Program Factors for Probation Programs
0.2 0.12 0.11
0.15 0.1
0.16 0.1
0.02
0
0.07
0 06 0.06 r-value
r-value e
0.08
0.12
0.05
0.05 0.04
0.04 0.02
-0.05
0.01 0
0
-0.1
-0.15 -0.02
-0.02 -0.03
-0.15 0-19%
20-39%
40-59%
Program Percentage Score
60+%
-0.04 High Risk Sample High Risk Longer Supervision
High Risk More Txt
High Risk More Referrals
16
0.15 0.15
0.1
r-value
0 05 0.05
0.03
0
-0.05
-0.1
Program Integrity—Relationship Between Program Integrity Score And Treatment Effect 25 20 Change In Recid divismRates
Change in Recidivism by 4 Point Factor Score for Probation Programs
15 22
10 5
10 5
0 -5 -10
-19
-15 -0.13
-0.15 0 Factors
1 or 2 Factors
3 Factors
-20 0-30
How systems are applying assessments – A few examples • Courts are using risk tools to assist in making placement and setting conditions • Courts are taking risk level into account before determining how to respond to revocations • Prosecutors use info for plea negotiations • Prosecutors P and d ddefense f attorneys plea l negotiate i in i suchh a way that does not hamstring probation in applying appropriate conditions (open mandate) • Prosecutors set up criteria for diversion for lower risk and does not require over-service • Probation using assessments to determine level of supervision and appropriate referrals • Probation careful to avoid over or under responding
How systems are applying evaluation, fidelity, and quality assurance – A few examples • Courts are asking for program information and performance data • Existing programs are being evaluated/assessed • QA teams are being formed to monitor programs • Contracts/funding requires EBP • Probation/DOC is auditing programs • Training and refresher training/certification
31-59
60-69
70+
How systems are applying programming: A few examples • Courts are giving sole sanctions for low risk offenders • Probation places low risk on group/administrative supervision • Courts are usingg position p for positive p reinforcement (i.e. ( specialty courts) • Probation is putting in place cog programs • Response to violations is structured around risk level • High risk offenders are being given more supervision and services • Make referral/contracts for service providers that use EBP • Pre-treatment programming is used to increase motivation
Lessons Learned ¾Who you put in a program is important – pay attention to risk ¾What you target is important – pay attention to criminogenic needs ¾How you target offender for change is important – use behavioral approaches
17
Important Considerations ¾Offender assessment is the engine that drives effective programs helps you know who & what to target ¾Design programs around empirical research helps you know how to target offenders ¾Program Integrity make a difference Service delivery, disruption of criminal networks, training/supervision of staff, support for program, QA, evaluation
18