Dynamic Risk Assessment for Sexual Offenders on Community Supervision

1
Dynamic Risk Assessment for Sexual Offenders on
Community Supervision

• R. Karl Hanson
• Public Safety Canada
• Presentation at the 9th Conference of the
  International Association for the Treatment of
  Sexual Offenders, September 6-9, 2006, Hamburg,
  Germany

2
Sex Offender Risk Assessment

• Static (unchangeable)
• Prior sexual offences
• Any extrafamilial victims
• Age
• Dynamic (changeable)
• Stable
• Intimacy deficits
• Sexual pre-occupations
• Lifestyle instability
• Acute
• Negative mood
• Drunkenness

3

• STATIC
• Baseline of risk
• STABLE
• Treatment targets
• ACUTE
• Warning when things could go wrong

4
Hanson, R. K., Bussière, M. T. (1998).
Predicting relapse A meta-analysis of sexual
offender recidivism studies. Journal of
Consulting and Clinical Psychology, 66 (2),
348-362.Hanson, R.K., Morton-Bourgon, K.E.
(2005). The characteristics of persistent sexual
offenders A meta-analysis of recidivism studies.
Journal of Consulting and Clinical Psychology,
73, 1154-1163.Hanson, R. K., Morton-Bourgon,
K. (2004). Predictors of sexual recidivism  An
updated meta-analysis. User Report 2004-02.
Ottawa Public Safety and Emergency Preparedness
Canada.www.psepc-sppcc.gc.ca Corrections
Publications

5
Established Risk Factors (potentially dynamic)

• Deviant sexual interests
• Children Paraphilias
• Sexual preoccupations
• Antisocial orientation
• Lifestyle instability, rule violation, APD
• Attitudes tolerant of sexual assault
• Intimacy deficits
• Emotional identification with children
• Lack of stable love relationships

6

• Few studies of acute predictors
• No follow-up studies
• No single factor is strongly related to sexual
  recidivism
• How to combine factors?

7
(No Transcript)
8
Actuarial Risk Scales for Sex Offenders

• Empirically derived factors
• Explicit rules for combining risk factors
• MnSOST (Epperson et al., 1995)
• VRAG/SORAG (Rice Harris, 1997)
• RRASOR (Hanson, 1997)
• Static-99 (Hanson Thornton, 1999)

9
Conceptual Risk Factors

• Structured Professional Guidelines
• Sexual Violence Risk 20 (SVR-20)
• ERASOR

10
Actuarial conceptual factors

• SVR-20 adding the factors
• Stable-2000 (Hanson Harris)
• SARN (United Kingdom)
• J-SOAP (Prentky)
• VRS-SO (Olver Wong)
• Vermont Treatment Needs and Progress Scale
• Level of Service Inventory Revised

11
Sex Offender Risk Assessment Validation Studies
(1975-2004)
12
Which approach works best?
13
Prediction of sexual recidivism
14
Actuarial Conceptual (Sex Recidivism)
15
Initial Deviance Assessment/SARN (Thornton)

• Deviant sexual interests
• Distorted attitudes
• Socioaffective functioning
• Self-management programs
• Measured by questionnaires

16
(No Transcript)
17
Dynamic Supervision Project

• Prospective study (2001 2006)
• Community Supervision Officers
• Multiple jurisdictions
• (Canada Iowa Alaska)
• 997 offenders 41 months follow-up

18
The Dynamic Supervision Project Methodology

• Static factors assessed once
• Static-99
• Stable factors assessed every 6 months
• Stable-2000
• Acute factors assessed every supervision
• Acute-2000

19
STABLE-2000 (Hanson Harris)

• 1) Significant Social Influences
• 2) Intimacy Deficits
• 3) Sexual Self-Regulation
• 4) Attitudes Supportive of Sexual Assault
• 5) Cooperation with Supervision
• 6) General Self-Regulation

20
Acute Factors

• Victim Access
• Emotional Collapse
• Collapse of Social Supports
• Hostility
• Substance Abuse
• Sexual Preoccupations
• Rejection of Supervision
• Unique Factor (optional)

21
Reliability

• Do different officers come up the same score for
  the same offender?
• Yes
• Static-99 ICC .91 (/- one point)
• Stable-2000 ICC .89 (/- two points)
• Acute-2000 ICC .86 (/- two points)

22
Recidivism Rates as of August, 2006 (41 months
follow-up)
23
Predictive accuracy
24
Sexual Recidivism
25
Sex or Sex Breach
26
Any Violent Recidivism
27
Any New Crime
28
Any Including Breach
29
Acute Risk Factors

• Short term risk
• How was the offender since the last time you saw
  them?
• Timing of new offence

30
Acute -2000

• 7,090 ratings
• January, 2001 to May, 2006
• 178 officers
• 751 offenders (range 1 70, median 6)
• Items small to moderately correlated
• (.14 to .50)
• Cronbach alpha .72 (7 items), .74 (8 items)

31
Acute Frequencies

• 1.7 Intervene Now (any 3)
• 13.6 Definite Problem (any 2)
• 64.6 Some Problem (any 1)
• 35.4 No problem (all zero)
• Total score
• mean 1.81 (SD 2.3) range 0 - 21

32
Acute Predictors of Sexual Recidivism

• Victim Access (p lt
  .002)
• Sexual Preoccupation (p lt .01)
• Rejection of Supervision (p lt.10)

33
Sexual Recidivism Next Month?

• If no info 1 out of 440
• If no acute problems 1 out of 1,240
• If lots of problems (score 4) 1 out of 167

34
Acute Prediction of Sexual Recidivism

• ROC .71
• Month Prior
• ROC .76
• Average of all ratings (except month prior)

35
Does Re-Assessment Help? Sexual Recidivism
36
Does Re-assessment Help?Sex or Sexual Breach
Recidivism
37
Concluding Remarks

• Static, Stable, and Acute Factors all predict
  recidivism
• What is the best method for combining static,
  stable and acute information into an overall
  evaluation?
• What is being measured by the Static, Stable and
  Acute factors?

38
Copies/Questions

• Sandra.Hadden_at_psepc-sppcc.gc.ca
• Karl.Hanson_at_psepc-sppcc.gc.ca
• www.psepc-sppcc.gc.caLook under corrections
  publications