Transcript Sexual Deviance

Working with
Sexual Offenders
Assessing and Managing Risk
Robin J. Wilson, Ph.D., ABPP
Clinical Director
The GEO Group / Florida Civil Commitment Center
[email protected]
Defining the
Problem
Stakeholders

victims
citizens
law enforcement
legal and correctional personnel
mental health personnel
the media

offenders





Sexual Offending
The past 10-15 years has been witness to a
flurry of research into the nature and
consequences of sexually offensive behavior.
 Isn’t
it odd that the focus has come so late in the
game?
 There is no doubt that there has been sexual
offending since there were people to be offenders
and others to be victims—thousands and
thousands of years.
 Why has the attention shifted only recently?
Victims
 As
many as 90% of reporting victims
know their offender
 2/3 or more of known offenses occur
in the victim’s own home
 As many as 90% of victims fail to
report their abuse to authorities or
others in a position to help
Is Child Sexual Abuse Really
That Big a Problem?
Statistics show that child sexual abuse occurs at an
alarming rate. As many as one in three girls and one
in five boys will be sexually abused at some point in
their childhood (<18), according to many reliable
studies of child sexual abuse, although most suggest
that these are underestimates. At a minimum, that
means that if you attend a social event (like a
concert, for instance) of 100 people, between 30 and
40 of those in attendance were sexually abused as a
child.
Consequences for Victims
 Maladaptive
sexuality
(either hypersexuality or
hyposexuality)

Prostitution
Promiscuity
Inability to express one’s
sexuality
Genital disfigurement
Sexual addiction






disorders
 Personality disorders
(antisociality, borderline
features)


 Eating


Dissociative disorders
Suicidality and selfharming behaviours
Interpersonal problems
(e.g., trust, loneliness,
inability to link with
others)
Loss of relationships
with significant others
(due to the abuse or
secondary victimization)
Substance abuse
Violence and aggression
Offensive Sexual Behavior
One of the greatest hurdles to defining
sexual deviance is a lack of clarity as to
what actually constitutes offensive
sexual behavior.
What do you consider to be sexually
offensive?
Assessment
Dangerousness
BAD:
“This person is dangerous.”
GOOD:
“If the following risk factors are
present, then there is a high/medium/
low probability that the person will
engage in some specific behavior
within specify period of time that
may place specific victims at risk for
a specific type and severity of harm.”
Assessment

assessment forms the foundation upon
which all subsequent intervention is built
poor assessment = trouble


comprehensive assessment should take
demand characteristics into consideration
assessment is dynamic
Assessment
 risk
assessment includes
consideration of static (historical)
and dynamic (day-to-day) variables
 risk assessment is facilitated by use
of actuarial instruments
Tenets of Good Assessment






assess reliability and validity of information
available
beware of the base-rate problem
look for corroboration between different sources
of information
beware of malingering and deception--there is
likely no such thing as a “true admitter”
use only standardized measures with documented
utility for your particular purposes
avoid speculation unless the evidence is highly
suggestive
Sources of Information
structured interview
 self reports
 collateral contact (family, friends)
 Police reports, prior criminal justice reports
 Court transcripts, Judge’s Reasons for
Sentencing, Pre-Sentence Reports, Victim Impact
Statements
 prior mental health reports, psychological tests,
actuarial measures, phallometrics

Risk Assessment Tools
General and violent recidivism:





Psychopathy Checklist-Revised (PCL-R)
Level of Supervision Inventory-Revised (LSI-R)
Violence Risk Appraisal Guide (VRAG)
HCR-20
Spousal Assault Risk Assessment Guide (SARA)
Risk Assessment Tools
Sexual recidivism:





Sex Offender Risk Appraisal Guide (SORAG)
Sexual Violence Risk-20 (SVR-20)
Rapid Risk Assessment for Sex Offender Recidivism
(RRASOR)
STATIC-99
STABLE / ACUTE
Strengths
 Valid
risk factors
 Explicit rules for combining
factors
 Explicit probability estimates
 Robust across settings & samples
 Easily scored
Weaknesses
Only moderate predictive accuracy
We
always want to do better!
Neglects important factors
Sexual
Deviance (Phallometrics)
All Dynamic Factors
Dynamic Supervision
of Sexual Offenders
Tab 3
DSP
Training Overview
•
•
•
•
2-day agenda
Binder contents
Who to use this with
Questions & Discussions
Why Assess Risk?
• Promoting public safety
• Routine interventions
• Targeting scarce resources
– Officer time
– Treatment
• Exceptional measures
Static, Stable, & Acute Risk Factors
Definitions
• Static – Non-changeable life factors that relate to risk
for sexual recidivism, generally historical in nature
• Stable – Personality characteristics, skill deficits, and
learned behaviours that relate to risk for sexual
recidivism that may be changed through intervention
• Acute – Risk factors of short or unstable temporal
duration that can change rapidly, generally as a result of
environmental or intra-personal conditions
Sex Offender Risk Assessment
Basis for the RRASOR and STATIC-99
STATIC (unchangable) factors
- prior sex offences
- age
- any extrafamilial victims
- any male victims
SONAR
Hanson & Harris
STABLE
• Intimacy Deficits
• Social Influences
• Attitudes
• Sexual Self-Regulation
• General Self-Regulation
ACUTE
• Substance Abuse
• Negative Mood
• Anger/Hostility
• Victim Access
Three Generations of Risk Assessment
Bonta (1996)
• First Generation = “Clinical Judgment”
•
•
•
•
Unstructured, Non-replicable, Personal Discretion
Based on experience and level of knowledge of the literature
Non-standard (even within same institution)
Level of prediction little better than chance
• Second Generation = “Actuarial Assessment”
•
•
•
•
Static, Actuarial, Structured, Replicable, Less open to Interpretation
Based on factors empirically related to recidivism
Standardized assessment, “Static” - Can not measure change
“Moderate” Levels of prediction, ROC’s upper 60’s to lower 70’s
• Third Generation = “Dynamic Assessment”
• Based on factors empirically related to recidivism
• Standardized assessment, Measures change
• Actuarial measure with dynamic factors
Three Linked Research Projects
The First: Meta-analytic Reviews
R. Karl Hanson and Colleagues
Public Safety Canada
• Hanson & Bussière, 1996, 1998
– Static risk factors
• Hanson & Morton-Bourgon, 2004, 2005
– Promising stable risk factors
• Hanson & Morton-Bourgon, 2007, in press
– Risk assessments
Types of Risk Assessment
Type of Evaluation
Theory
Overall
Evaluation
Professional
Judgement
Mechanical
Actuarial
Professional
Judgement
Theory
Mechanical
Factors
Unstructured Clinical
Judgement
?
Empirical-Actuarial
Empirically
Derived
Structured
Professional
Judgement
Mechanical
SVP-20/HCR-20
SARN/STABLE-2000
Prediction of sexual recidivism
Measures Designed for
Sexual Recidivism
d (95% CI)
N (k)
Empirical Actuarial
.67 (.63-.72)
24,089 (81)
Mechanical
.66 (.58-.74)
5,838 (29)
Structured Judgement
.46 (.29-.62)
1,131 (6)
Unstructured
.42 (.32-.51)
6,456 (11)
Three Linked Research Projects
The Second: Dynamic Predictors 1998
• File review study
• Interviewed the supervising officer
 n = 208 community sexual recidivists
 n = 201 community sexual non-recidivists
• Canada-wide study
• Federal parole and Provincial probation
SONAR
Hanson & Harris
STABLE
• Intimacy Deficits
• Social Influences
• Attitudes
• Sexual Self-Regulation
• General Self-Regulation
ACUTE
• Substance Abuse
• Negative Mood
• Anger/Hostility
• Victim Access
Three Linked Research Projects
The Third: Dynamic Supervision 2000
• Follow 1000+ in-community sex offenders
- for a 43 month period
- prospective design
• Multiple jurisdictions
• Continuous intake
- consecutive new cases
• Trained officers submitting data
• American participants = Alaska & Iowa
Dynamic Supervision Project 2000
Methodology
• Static factors assessed once
• Stable factors assessed every 6 months
• Acute factors assessed every supervision
STATIC-99
• Male victims
• Ever Lived With ...
• Non-contact sex
offences
• Unrelated victims
• Stranger victims
• Prior sex offences
(3 points)
•
•
•
•
Current non-sex violence
Prior non-sex violence
4+ sentencing dates
Age 18 - 24.99
Comparison - Stable Factors
STABLE-2000
STABLE-2007
16 Factors in 6 clusters
13 Factors in 5 clusters
1 Sig. Social Influences
1 Sig. Social Influences
2 Intimacy Deficits
2 Intimacy Deficits
3 Sexual Self-regulation
3 General Self-regulation
4 Att. Support. Sex Asslt.
4 Sexual Self-regulation
5 Cooperation with Supn.
5 Cooperation with Supn.
6 General Self-regulation
Comparison - Acute Factors
ACUTE-2000
ACUTE-2007
7 items – 2 factors
Factor One
Sex & Violence
8 items – 1 factor
Factor Two
General re-offence
1 Victim Access
1
Victim Access
Victim Access
2 Hostility
2
Hostility
Hostility
3 Sexual Preoccupations
3
Sexual
Preoccupations
Sexual
Preoccupations
4 Rejection of Supervision
4
Rejection of
Supervision
Rejection of
Supervision
5 Emotional Collapse
5
Emotional Collapse
6 Collapse of Social
Supports
6
Collapse of Social
Supports
7 Substance Abuse
7
Substance Abuse
8 Unique Factor (optional)
Inter-rater Reliability – Intraclass Correlations
“Field Reliability”
Calculated amongst seven (7) raters
MEASURE
RELIABILITY
STATIC-99 (90 cases)
0.92
“First” STABLE-2000 (87 cases)
0.94
“Second” STABLE-2000 (45 cases)
0.93
“Last” ACUTE-2000 (75 cases)
0.92
“Earlier” ACUTE-2000 (63 cases)
0.88
Inter-rater Reliability
on a standardized training exercise
“Percent Correct” on a Training Exercise
Officers trained by RKH & AJRH
N = 213
STATIC-99
= 91 to 95% of officers were within one point
STABLE-2000 = 47 to 67% within one point
= 74 to 99% within three points
ACUTE-2000 = Median of 82% correct (but huge range)
Officers trained by Ontario trainers
N = 45
STATIC-99
= 88 to 91% of officers were within one point
STABLE-2000 = 22 to 88% within one point
= 80 to 100% within three points
ACUTE-2000 = Median of 87% correct (but huge range)
Who Can I Use This Stuff With?
Population
STATIC-99
STABLE-2007
ACUTE-2007
Adult Male Sexual
Offenders



Adult offenders
with 2 to 10 years
offence-free in the
community



Juvenile offenders
aged 16 & 17
With Caution
With Caution
With Caution
Juvenile offenders
less than 16 years



Research use only
Research use only
Research use only
Adult female
offenders
DSP Outcomes –(1)– Sexual Offences
“Overall” versus “Different Types – SOs”
Recidivism
Rate
Test
ROC
Overall
7.2%
(57/793)
STATIC-99
S-99 & STABLE-07
.74
.76
Extrafamilial CM
7.4%
(15/202
STATIC-99
S-99 & STABLE-07
.74
.77
Incest
1.7%
(3/180)
STATIC-99
S-99 & STABLE-07
.48
.58
Rape
9.0%
(24/267)
STATIC-99
S-99 & STABLE-07
.70
.73
DSP Outcomes –(1)– Sexual Offences
“Overall” vs. “Different Types – SOs”
Stuff to think about
• Overall – both tests are moderately helpful
• Not as good for Incest Offenders
– but only “3” recidivists
DSP Outcomes –(2)– Sexual Offences
“Overall” vs. “Exceptional”
Recidivism
Rate
Overall
“Exceptional”
DD & Psych.
7.2%
(57/793)
13.2%
(15/144)
Test
ROC
STATIC-99
.74
S-99 & STABLE-07
STATIC-99
.76
.73
S-99 & STABLE-07
.71
DSP Outcomes –(2)– Sexual Offences
“Overall” vs. “Exceptional”
Stuff to think about
• “Exceptional” = those guys with
psychological/psychiatric problems and those who are
diagnosed as “Developmentally Delayed”
• Higher recidivism rate
• “DD” guys appear to be higher risk than overall average
and they were rated higher on STABLE measures
• Are these guys just easier to catch?
• STABLE needs – with these two groups there seems to
be something of a ceiling effect as intellectual and
psychological needs are confused with criminogenic
needs
DSP Outcomes –(3)– Sexual Offences
Aboriginal vs. Non-Aboriginal
Recidivism
Rate
NonAboriginal
Aboriginal
6.6%
(41/622)
10.3%
(16/155)
Test
ROC
STATIC-99
.76
S-99 & STABLE-07
STATIC-99
.80
.65
S-99 & STABLE-07
.61
DSP Outcomes –(3)– Sexual Offences
Aboriginal vs. Non-Aboriginal
Stuff to think about
• Aboriginal data really “noisy”
• With some of the Aboriginal data there
were data control issues – some untrained or
insufficiently trained officers collecting data
• Suspect that there may well be some
confounding of non-criminogenic needs
with criminogenic needs (STABLE
measures)
DSP Outcomes –(4)– Sexual Offences
“Overall” vs. “Conscientious”
Recidivism
Rate
Overall
Conscientious
7.2%
(57/793)
6.8%
(23/336)
Test
ROC
STATIC-99
.74
S-99 & STABLE-07
STATIC-99
.76
.81
S-99 & STABLE-07
.84
DSP Outcomes –(4)– Sexual Offences
“Overall” vs. “Conscientious”
Stuff to think about
“Conscientious” officers – those who provided all
requested data
• Points out need for good training
• Points out need for “management buy-in”
• Findings say – “It works pretty well if you take it
seriously”
• Officers have to be careful and consistent
Analysis of Stable Change
• Little change over the 6 month period
• Amount of change unrelated to recidivism
• Only weak evidence that most recent assessment more
accurate the prior assessment
• Few offenders would have received effective treatment
• Recommend Stable re-assessment every year
• Results of the STABLE-2000 were used to refine the
measure to create STATIC-2007
Validity of STABLE-2000 Items
• Most items predict most outcomes in linear
progression
• Exceptions:
–
–
–
–
–
–
Lovers/intimate partners (non-linear)
Emotional identification with children
Sexual entitlement
Rape attitudes
Child molester attitudes
Deviant sexual interests
STABLE-2007
• Dropped the three attitude items
• Slight revision to three other items
– Emotional identification with children only for child
molesters
– Lovers/intimate partners becomes “Capacity for
Relationship Stability”
– Prior sex victims considered in scoring of Deviant
sexual preferences
• Simple addition used for STABLE-2007
(See STABLE-2007 Tally Sheet)
STABLE-2007 adds predictive power
Recidivism
Rate
Test
ROC
7.2%
(57/793)
9.7%
(77/793)
STATIC-99
S-99 & STABLE-07
STATIC-99
S-99 & STABLE-07
.74
.76
.69
.73
13.7%
(109/793)
STATIC-99
S-99 & STABLE-07
.71
.72
Any Crime
19.3%
(153/793)
STATIC-99
S-99 & STABLE-07
.70
.70
Any Crime
plus breaches
29.2%
(232/793)
STATIC-99
S-99 & STABLE-07
.69
.70
Sexual
Sexual plus
breaches
Violent
Acute Variables
• Factor analysis and associations with
recidivism suggested a main factor and a
sub-factor of sex/violent items
Acute Factors
Victim Access
Sexual Preoccupations
Hostility
Rejection of Supervision
Emotional Collapse
Collapse of Social Supports
Substance Abuse
Are Acute Factors Acute?
• Most recent acute ratings predicted
recidivism
• Average of acute ratings predicted better
than most recent acute rating
• Average of acute ratings over the 4th to 6th
month prior predicted as well as average for
first to 3rd month
• Average of last 6 months predicted slightly
better than average of months 7-12
Relative Risk Increases for
Static/Stable/Acute Current Risk
“Generally” – as a heuristic – risk levels
• Double from Low to Moderate
• Double again from Moderate to High
– See “Chart Pack” Tab 09 – Chart # 10
• Rules for Combining Static/Stable/Acute Factors
– See “Chart Pack” Tab 09 – Chart # 9
How good are officers at adjusting
structured risk assessments?
• STATIC-99 (4 categories)
• Static override
– “Are there any exceptional circumstances that
support an override of the static risk assessment?”
(yes up, yes down, no)
• STABLE–2007 (13 items, 3 risk categories)
• N = 507 with these three measures
Prediction of Recidivism (ROC)
Recidivism
Type
Static-99 Static-99 +
alone
override
Sexual
.77
.75
Static-99 +
Stable-2007
.81
Sexual + sex
Breaches
.71
.69
.78
Any violent
.74
.71
.77
Study
Measure
Actuarial Adjusted
Type of
Recidivism
Hanson
(2007)
Static-99
1.04
.95
Sexual
Static-99
.91
.78
Violent
Gore
(2007)
MnSOST-R
.50
.31
Sexual
Vrana
(2008)
LSI-OR
.90
.54
Violent
LSI-OR
.98
.64
Any
.87
.64
Mean
General Recommendations
for Risk Assessment
• Use an explicit list of empirically validated
risk factors determined in advance
• Use an empirically validated method of
combining the risk factors into an overall
evaluation
• Estimate the risk for an individual offender
based on the group he or she most closely
resembles
General Recommendations
for Risk Assessment
• Build-in methods for quality control
• At least one year is needed before reevaluating stable factors (pending deliberate
interventions)
• Use offence history, enduring psychological
characteristics, and current behaviour to
evaluate risk
A Directed Program of Research
Hanson & Bussière
Meta-analysis
1996
RRASOR, 1997
Hanson & Harris
Dynamic Predictors
1998
SONAR, 2000
STATIC-99
Hanson & Harris
Dynamic Supervision
2001 - 2007
STABLE2000
STABLE2007
&
&
ACUTE2000
ACUTE2007
STATIC-2002
Dynamic
Supervision
of Sexual
Offenders
2007
Tab 4
STATIC-99
Hanson & Thornton, 1999
Static Risk Factors
• Don’t change (On the whole)
• Allow you to gauge the long-term level of
risk for sexual recidivism
• Allows you to determine an appropriate
level of supervision and treatment for the
individual (Andrews & Bonta, 2006)
STATIC-99
An actuarial risk tool used
for the prediction of sexual and violent
recidivism among adult male sexual offenders
Three Sections
1) Background, development and validity
2) Scoring
3) Interpretation
JÄSENTYNYT RISKIARVIO – 99.02
JRA/STAATTINEN 99
Riskitekijä
1) Aikaisemmat seksuaalirikokset
0
Pisteet
1
2
2) Tuntematon
3) Ei-sukslainen
0
0
1
1
4) Miespuolinen uhri
5) Ilman kosketusta
6) Yksinäinen/naimaton
0
0
0
1
1
1
7) 18-24 vuoden ikäinen vapautuessa
8) Ei-seksuaalinen väkivalta laskentatuomiossa
9) Ei-seksuaalinen väkivalta aiemmissa
0
0
0
1
1
1
10) Yli neljä tuomiota
0
1
3
Risk Factor
Young
Single
Index non-sexual
violence
Prior non-sexual
violence
Prior Sex Offences
Prior sentencing dates
Convictions for noncontact sex offences
Any Unrelated Victims
Any Stranger Victims
Any Male Victims
Total Score
Codes
Score
Aged 25 or older
Aged 18 – 24.99
Ever lived with lover for at least two
years?
Yes
No
No
Yes
0
1
No
Yes
0
1
0
1
0
1
Charges
Convictions
None
1-2
3-5
6+
3 or less
4 or more
No
Yes
None
1
2-3
4+
0
1
2
3
0
1
0
1
No
Yes
No
Yes
No
Yes
Add up scores from individual risk
factors
0
1
0
1
0
1
Actuarial Risk Scales
1) List of risk factors
2) Explicit method of combining risk factors
3) Explicit probability estimates associated
with each score
Actuarial—Empirical (Sex Recidivism)
d (95% CI)
N (k)
STATIC-99
.67 (.62-.72)
20,010 (63)
RRASOR
.60 (.54-.65)
11,031 (34)
Risk Matrix 2000 Sex
.67 (.56-.77)
2,755 (10)
STATIC-2002
.70 (.59-.81)
3,330 (8)
MnSOST-R
.76 (.65-.87)
4,672 (12)
STATIC-99
• Male victims
• Ever Lived With
• Non-contact sex
offences
• Unrelated victims
• Stranger victims
• Prior sex offences (3
points)
• Current non-sex violence
• Prior non-sex violence
• 4+ sentencing dates
• Age 18-24.99
Created from RRASOR (Hanson, 1997) and SACJ-Min (Thornton)
Samples
Oak Ridge HM Prison
Pinel
Millbrook
344
191
142
531
% Child
Molesters
70
100
49
61
Years
follow-up
4
23
10
16
Sample
Size
STATIC-99
Sexual Reconviction Rates
1
0.8
Pinel
Millbrook
HM Prison
0.6
0.4
0.2
Years after release
NOTE: Untreated Samples
24
21
18
15
12
9
6
3
0
0
STATIC-99
Sexual Reconviction Rates
1
0.8
Low
Medium-Low
Medium-High
High
0.6
0.4
0.2
Years after release
24
21
18
15
12
9
6
3
0
0
STATIC-99 Sexual and Violent
Reconviction Rates
1
0.8
Pinel
Millbrook
HM Prison
0.6
0.4
0.2
Years after release
24
21
18
15
12
9
6
3
0
0
STATIC-99 Sexual and Violent
Reconviction Rates
1
0.8
Low
Medium-Low
Medium-High
High
0.6
0.4
0.2
Years after release
24
21
18
15
12
9
6
3
0
0
Accuracy and Error
1. Inter-rater Reliability
2. Confidence Intervals for Group Estimates
3. Replications
–
–
Rank order
Recidivism rates
4. Variability of individual estimates
Inter-rater Reliability – STATIC-99
Study
Barbaree et al.
(2001)
Hanson (2001b)
Sjöstedt &
Långström (2001)
Harris et al. (2003)
Size Statistic
Reliability
Pearson r –
total scores
% agreementitems
.90
55
Kappa- items
.80
55
Intra-class r –
total score
Kappa – items
.87
Intra-class r total scores
.87
30
55
20
10
.91
.90
Standard Error of Measurement
STATIC-99
SEM = Standard deviation x (1 – reliability)1/2
SEM = 1.97 (1 - .87)1/2 = .71
95% C.I. = 1.86 x .71 = 1.40
Result: 19 times out of 20, the offender’s
true score will be within ± 1.4 points of the
observed score
Variability of Group Estimates
• Confidence Intervals
– Get narrower as sample size increases
– Intervals derived from logistic regression uses
information from full sample (not just specific
score)
Example: 10 year Sexual Recidivism
95% Confidence Intervals
for Logistic Regression Recidivism Estimates
80
70
60
50
40
30
20
10
0
Predicted
Lower C.I.
Upper C.I.
0
1
2
3
4
5
6
7
STATIC-99 Score
8
9
10
Assumptions for
Group Confidence Intervals
• All individuals in each category have the same
probability of recidivism
• All relevant risk factors have been measured
– BUT STATIC-99 does not claim to measure all
relevant risk factors (heterogeneity within groups is
expected)
• Requires assumptions about the similarity between
the individual and the group data
Do Confidence Intervals for
Individuals Make Sense?
• Hart, Michie & Cooke (2007) have claimed
that they can calculate individual
confidence intervals for actuarial
instruments using Wilson’s (1927) method
• These confidence intervals are so large as to
be useless (e.g., 0 to 84%).
Comments on Hart et al.(2007)
• The way the formula is used has no
meaning.
• “The 95% CIS for “individual risk” piles
nonsense on top of meaninglessness . . .
With “1” in place of “n,” the formulae just
don’t mean anything”
(Mossman & Sellke, 2007)
Serious Issue:
How to specify the error associated
with an individual assessment?
• Inter-rater reliability, predictive validity, and
confidence intervals (for group estimates) are
all relevant.
• Requires assumptions about the similarity
between the individual and the group data.
Replications
• Rank Order in different samples
– How well does the STATIC-99 differentiate
between recidivists and non-recidivists in
diverse samples?
• Moderate predictive accuracy
– Average d =.70 (.64-.76)
– 13,288 offenders; 42 samples
STATIC-99 Replications
Authors
Country
Sample
n
d
Sjöstedt &
Långström (2001)
Sweden
Prison
1,303
0.95
Hood et al. (2002)
England
Prison
162
1.11
53
1.05
Beech et al. (2002)
United Kingdom Community
Nunes et al. (2002)
Canada
Community
Pre-trial
258
0.74
Harris et al. (2003)
Canada
Mental
Health &
Prison
396
0.43
McGrath et al.
(2003)
U.S.A.
Prison
172
0.74
STATIC-99 Replications
Authors
Country
Sample
n
d
Bartosh et al.
(2003)
U.S.A.
Prison
186
0.49
Langton (2004)
Canada
Prison
468
0.47
Thomas et al. (2004)
Canada
Community
899
1.00
Ducro et al. (2004)
Belgium
Mental
Health
147
0.58
Netherlands
Mental
Health
121
0.78
Austria
Mixed
81
0.91
De Vogel et al.
(2004)
Rettenberger & Eher
(2006)
STATIC-99 Replications
Authors
Watanabe et al.
(2007)
Country
Sample
n
d
Japan
Child
rapists
Prison
402
0.44
69
0.99
Prison child
molesters
Mental
Health
495
0.82
336
0.40
Prison
Sexual
Homicide
103
90
0.62
0.53
Endrass et al. (2007)
Switzerland
Allan et al. (2007)
New Zealand
Bengtson &
Långström
Denmark
Marghem (2007)
Hill et al. (2008)
Belgium
Germany
New STATIC-99 Recidivism Norms
• Updated Norms project - Ongoing
– 27 datasets collected so far (18 coded and used
to produce new norms)
New Norms:
Sample Descriptives (k = 18)
• Country:
•
•
•
•
•
8 Canadian (n = 2,271)
4 continental Europe (n = 2,416)
4 U.S. (n = 1,028)
1 New Zealand (n = 493)
1 U.K. (n = 198)
• Year of Release (n = 6,114, Info on 16
samples):
• 90% released in 1990 or later
New Norms:
Sample Descriptives (k = 18)
• Offender Type (n = 4,953, Info on 10
samples):
•
•
•
•
53% Child Molester
38% Rapist
5% Other (e.g., exhibitionist, voyeur)
4% Mixed
• Recidivism Criteria
• 10 samples used charges (n = 3,462); 8 convictions
(n = 3,312)
New Norms:
Sample Descriptives (k = 18)
• Location (Info on 16 samples)
• 15 samples from institutions (n = 5,590)
• 3 samples had some community-only offenders (n =
483)
• Sample Type (Info on 12 samples)
• 6 samples pre-selected to be high risk (n = 1,562)
• 5 samples of routine CSC cases (n = 1,249)
• 2 samples with low-risk community cases (n = 477)
New Norms:
Sample Descriptives (k = 18)
• Treatment Status by sample (Info on 9
samples)
• 6 samples mostly treated (75%+, n = 1,414)
• 3 samples mixed (25%-75%, n = 843)
• 0 samples mostly untreated (<25% treated)
New Norms: 5-year Sexual
Survival Analysis, initial n = 6,406
45
40
35
30
25
Original
2008 results
20
15
10
5
0
0
1
2
3
4*
STATIC-99 score
5*
6+*
New Norms: 10-year Sexual
Survival Analysis, initial n = 6,406
50
45
40
35
30
25
20
15
10
5
0
Original
2008 results
0
1
2
3
4*
STATIC-99 score
5
6+
New Norms: 5 year Violent
Survival Analysis, initial n = 6,096
50
45
40
35
30
25
20
15
10
5
0
Original
2008 results
0
1
2
3
4
STATIC-99 score
5
6+
New Norms: 10 year Violent
Survival Analysis, initial n = 6,096
60
50
40
Original
2008 results
30
20
10
0
0
1
2
3
4
STATIC-99 score
5
6+
New Norms: Summary
• Sexual Recidivism
– Recidivism rates significantly lower than original
norms
– Differences are meaningful: E.g., scores of 6+
• 39% versus 26% at five years (survival analysis)
• 45% versus 35% at ten years (survival analysis)
• Violent Recidivism
– Recidivism rates significantly lower than original
norms
• Differences not as apparent in the graphs due to suppression
from increased proportion of rapists in new samples
Recidivism Rates:
Moderator Variables
• STATIC-99 ranks relative risk consistently
across samples
• Observed recidivism rates vary based on
certain moderator variables
– Sexual: Sample type (preselected high risk
samples versus routine CSC samples)
– Violent: Offender type (rapist versus child
molester) and sample type
Moderator Variables
• Necessitates separate recidivism tables
– Sexual and Violent Recidivism: Report both
CSC and preselected high-risk estimates
• List the factors the offender shares with both groups
and make a judgment about which group the
offender likely resembles
– Violent: In addition, report rapist and child
molester estimates, and classify the offender
accordingly
Classifying Offender Type
• Suggested guidelines
– Child molester: Victims 13 years old and less
– Rapist: Victims 18 years old and up
– Victims age 14-17
• Rapist if victims unrelated, predatory, overt force
• Child molester if victims related, grooming,
relationship of trust
– Victims in multiple categories
• Judge predominant victim preference
Reporting Recidivism Estimates:
3 Methods
1)
Survival analysis (original STATIC-99 norms)
•
•
2)
Random fluctuations (e.g., 10 and 15 year sex recidivism
rates for STATIC-99 scores of 0 and 1)
Only use offenders with specific score – noisy, especially
with small sample sizes (scores of 6+ on STATIC-99)
Observed rates for fixed follow-ups
•
3)
Same problems as above, & much reduced sample size
Predicted values from fitted logistic regression curves
(**preferred option**)
•
•
Reduced overall sample size, requires logistic distribution
Corrects random fluctuations by using B1 (the whole
dataset) to smooth the estimates
Example: 10 year Sexual
Recidivism
70
60
50
Survival
Fixed FU
Logistic
40
30
20
10
0
0
1
2
3
4
5
6
Static-99 score
7
8
9
10
10 Year Sexual Recidivism Rates
(from logistic regression estimates)
70
60
50
40
High Risk (n = 735)
30
CSC (n = 342)
20
10
0
0
1
2
3
4
5
6
7
STATIC-99 score
8
9 10
10 Year Sexual Recidivism Rates
(from logistic regression estimates)
70
60
50
40
High Risk
CSC
Original
30
20
10
0
0
1
2
3
4
5
6
7
8
STATIC-99 score
9 10
10 Year Sexual Recidivism Rates
(from logistic regression estimates)
80
70
60
High Risk (n = 790)
50
40
Routine CSC (n =
342)
30
20
10
0
0
1
2
3
4
5
6
7
STATIC-99 score
8
9 10
10 Year Sexual Recidivism Rates
(from logistic regression estimates)
80
70
60
Rapist (n = 427)
50
40
Child Molesters (n
= 580)
30
20
10
0
0
1
2
3
4
5
6
7
STATIC-99 score
8
9 10
Relative Versus Absolute Risk
• Absolute risk: probability of recidivism associated
with specific score
• Relative risk: rank order offenders (e.g.,
percentiles)
– How does this offender look compared to other sex
offenders? What proportion of sex offenders have a
higher score than him?
– Sufficient for most decisions, such as resource
allocation (e.g., treatment or supervision intensity)
– More stable across time and settings than absolute risk
Deriving Percentiles
• Reporting percentiles from our samples
– Problem: Unlikely to be representative of “true”
population of sex offenders
• Solution: Re-weight our samples to
approximate the “true” population of
Canadian adjudicated sex offenders
Percentiles: Method
Our 5 samples of
Canadian sex
offenders (N = 2,398)
– 76.8% Federal
– 10.6 Provincial
– 12.6 Non-Custodial
Adult Court Criminal
Survey Statistics
– Sentences of offenders
whose most serious
charge is sexual
– 11.3% Federal
– 37.4% Provincial
– 51.3% Non-Custodial
Percentiles: Analysis
• Full population has three strata (noncustodial, provincial, federal)
• Estimate of population mean percentile is
sum of mean percentiles for each stratum,
after weighting each stratum based on its
relative distribution in the full population
• See Static-99 estimated percentile tables
(Tab 4c)
Estimated Percentiles: Advantages
• Re-weighting the samples provides
relatively unbiased estimates of population
values
– Estimated percentiles have meaning relative to
“true” population of adjudicated Canadian sex
offenders
Disadvantages of this Method
• Accuracy of our estimation procedures
depends on extent to which Adult Criminal
Court Survey (ACCS) data truly reflects
*real* population of Canadian sex
offenders, and extent to which definition of
“sex offender” is similar in our 5 samples
and the data from the ACCS
• Extent to which Canadian population of sex
offenders approximates other countries is
unknown
Observations from Relative Risk:
Example Using Static-2002
• In population of Canadian adjudicated sex
offenders, proportion of high-risk offenders
(scores of 9+) very small (6%)
• 30% of offenders are in the low-risk
category (scores of 0-2)
– Very low predicted 5-yr recidivism rates
• Up to 2.4% in CSC groups
CSC 5 year Sex Recidivism Rates and
Percentages of Canadian offenders
20
50
18
45
16
40
14
35
12
30
10
25
8
20
6
15
4
10
2
5
0
0
0
1
2
3
4
5
6
7
Static-2002 score
8
9
10
11
12
Percentages
CSC recid (n = 734)
Strengths
•
•
•
•
•
Valid risk factors
Explicit rules for combining factors
Explicit probability estimates
Robust across settings & samples
Easily scored
Weaknesses
• Only moderate predictive accuracy
• We always want to do better!
• Coding rules are confusing
• You need to pay attention to rules
• Neglects important factors
• Sexual Deviance (Phallometrics)
• All Dynamic Factors
STATIC-99
Three Sections
Section # 2
“Scoring the STATIC-99”
Appropriate Populations
• Adult male sexual offenders
– 18 years or older at time of release
– Charged or convicted for an offence that is known to
have a sexual motivation/component
• Victims
– Children
– Non consenting adults
– Other (corpses, animals)
• Recent sex offence conviction/release
Inappropriate Populations
•
•
•
•
Females
Juveniles
Consenting sex between similar age peers
Offenders with 10+ years sex offence-free
in the community
• Not for making statements about possible
guilt
STATIC-99
• Category “A”
– Identifiable victim
• Category “B”
– No identifiable victim
– Indecency without sexual intention
• (e.g. urinating in public)
– Prostitution offences
– “Public toileting”
– Pornography offences
The Origin of the Coding Rules
• Based on definitions used in original data sets
• Now, Validation studies and Coding Rules
(Harris, Phenix, Hanson, & Thornton, 2003)
• Extrapolations
(what we would have scored if we had encountered
such cases)
– Different legal systems
– Rare events
Coding the STATIC-99
• Demographic
– age at assessment/release; relationship history
• Official criminal history
– prior sex offences; index non-sexual violence; prior
non-sexual violence; prior sentencing occasions;
convictions for non-contact sex offences
• All credible information (except polygraph)
– any unrelated victims; any stranger victims; any male
victims
Coding the STATIC-99
• 1 = Yes, problem
• 0 = No, O.K.
• Except Prior Sex Offences (0, 1, 2, 3)
Demographic #1
Young
• Age (18 - 24.99 = 1) (25 or older = 0)
– age when placed at risk
– (current age/age at release)
– when risk to be assessed
• 1 point for being under the age of 25
Demographic #2
Ever Lived With…
• Ever lived with lover for 2 continuous years?
– must be continuous
– prison marriages/lovers don’t count
• 1 point for having never lived for 2 years with the same
lover
#2 Ever Lived With …
Who can you have lived with??? (p 25)
• Legal marriages of < 2 years do not count
• Prison marriages (while guy incarcerated) do not count
– If relationship still there > 2 years after he gets out - does
count
•
•
•
•
Non-human species – do not count
Priests and other celibates – no exemption
Must be a relationship that is legal
Must be age to consent to relationship
– Child “vics” do not count – even if “consensual”
Big Bad Bart
Today is the first day of November. You have been sent by
the Parole Board to do a risk assessment on Bart for
possible release about the middle of February. From your
reading of the file you know that Bart is 24 years old at the
moment and you know that he was born on Christmas day.
You know that Bart and Becky lived together as “man and
wife” for three years before Bart went to prison. However,
it was Becky that made the complaint to the police after he
beat her up just one too many times. Becky was 16 years
old when she finally turned him in. The file indicates that
Bart has never lived long-term with any other woman.
Sexual Offence
• Category A - (The Behaviour)
– Specific Victim, Child or Non-consenting
Adult
– sexual assault, kidnapping, contributing to
juvenile delinquency, invitation to sexual
touching, exhibitionism, voyeurism
• Category B - (The Behaviour)
– No Specific Victim
– prostitution, public toileting, pornography
– Indecency without a sexual motive
Sexual Offence
Remember:
Any “A’s” mean all “B’s” count
(pg 14 - 15)
• Charges for urinating in public, public nudity, or the like,
associated with mental/psychological impairment, are
counted as “B’s”
• Giving Alcohol, Drugs, Noxious Substance, or other
Stupefacient (p 38 & 40)
– Can count as a sexual offence if given with the
intention of making the sexual offence easier
– May also score as NSV (Either Item #3 or #4)
Index Sex Offence
• Most recent sex offence
– conviction, charge, arrest, breach, prison
misconduct for sexual crime.
• May include multiple victims/offences
• Pseudo-recidivism counts as part of index
– did the offender re-offend after the first index
offence was detected?
Identify the Index Offence(s)
Behaviour
Date
1. Exposes self in
public
July, 1990
2. Touches Suzie
Jan, 1992
3. Internet porn
1998- 2000
Conviction
Date
Sentence
1. Indecent Act
Oct, 1990
2 years
probation
2. Sexual
Interference (child)
May, 2000
18 months
Sept, 2001
Recommitted
3. Child
Pornography
4. Drinking
Sept, 2001
4 Parole Violation
Identify the Index Offence(s)
Behaviour
Date
1. Sex assault Jill
Aug, 1995
2. Sex assault Joan
Sept, 1996
3. Sex assault Sue
Dec, 1996
Conviction
Date
Sentence
3. Sexual Assault
(Sue)
March,
1997
3 years
1. Sexual Assault
(Jill)
Sept, 1998
2 years
concurrent
2. Sexual Assault
(Joan)
Jan, 2000
3 years
probation
Sexual versus non-sexual violence
Sexual offence
• Sexual motivation
Non-sexual violent offence
• Motivation does not matter
– Sexual or non-sexual
motivation
• Name of the offence
does not matter
• Charges, convictions
• Name of the offence
indicates violence, but does
not necessarily indicate sex
• Convictions only
#3 Index Non-sexual Violence
• Look at the Name of the Offence
• Count only convictions
• Victim can be the same as victim for sexual
offence or different
• Be aware of pseudo-recidivistic violence
• 1 point for Index Non-sexual violence
#4 Prior Non-sexual Violence
•
•
•
•
Look at the Name of the Offence
Prior to Index Offence
Count only convictions
Victim can be the same as victim for sexual
offence or different
• 1 point for Prior Non-sexual violence
#3 & #4 Non-sexual Violence
Offences that count for both “Prior” and “Index”
• Convictions Only – Adult and Juvenile
• Aggravated Assault, Arson, Assault, Assault CBH, Forcible
Confinement, Kidnapping, Murder, Robbery, Threatening,
Using/Pointing a Weapon/Firearm, Wounding
• Note: Weapons offences would not count unless the weapon
was used in the commission of a violent or sexual offence.
• Note: Do not include – Driving Accidents or Negligence
causing injury or death
• Note: A separate Non-sexual violence conviction is required
for coding Non-sexual Violence
#3 & #4 Non-sexual Violence
If the behavior was sexual, but the offender was
convicted of non-sexual violence, the same
conviction counts as both a sexual offence and a
non-sexual violent offence. Hence, convictions for
Sexual Assault and Forcible Confinement are coded
as two sexual offences and one Non-sexual Violent
offence
#3 & #4 Non-sexual Violence
Date
Charge
July 1999
Forcible
Confinement
Conviction
Forcible
Confinement
Sentence
20 Months & 3
Yrs. Prob.
If you know that the behavior was sexual – this can count
as: One Sexual Offence (Either for “Priors” or an “Index”)
and One Non-sexual Violence
#3 & #4 Non-sexual Violence
Date
Charge
Conviction
Sentence
July 1999 1) Forcible
1) Forcible
20 Months &
3 Yrs. Prob.
Confinement
Confinement
2) Sexual Assault 2) Sexual Assault
As long as you know that the Forcible Confinement was
part of the sexual offence this situation would count as:
Two Sexual Offences and One Non-sexual Violence
(Either for “Priors” or an “Index”)
#3 & #4 Non-sexual Violence
Offences that Do Not count
•
•
•
•
Sexual Assault
Sexual Assault with a Weapon
Aggravated Sexual Assault
Sexual Assault Causing Bodily Harm
• Score these just as sexual offences
#5 Prior Sex Offences
• Look at the Behaviour – Can be pled down and still
count as a sex offence (Underwear stealing)
• Exclude Index offence(s) – Prior to Index
• Count all charges and convictions (“counts
count”)
• Arrests/Prison misconducts/Parole violations
count as one charge
• If change (e.g., plea bargain), count the charges
that go to court
#5 Prior Sex Offences
• P/P/C-R Violations for sexual misbehaviours
count as “one charge”
– Multiples at the same time = “one charge” (p 36)
– Separate occasions – “one charge” per incident (p
36)
• Acquittals – count as “one charge” (p 37)
• Not Guilty – count as “one charge” (p 40)
• Arrests for Sexual Offences - count as “one
charge” (p 37)
#5 Prior Sex Offences
Charges
Convictions
Item Final Score
None
None
0
1-2
1
1
3-5
2-3
2
6 plus
4 plus
3
Code Prior Sex Offences
Behaviour
Date
1. Exposes self in
public
July, 1990
2. Touches Suzie
Jan, 1992
3. Internet porn
1998- 2000
Conviction
Date
Sentence
1. Indecent Act
Oct, 1990
2 years
probation
2. Sexual touching
(child)
May, 2000
18 months
Sept, 2001
Recommitted
3. Child
Pornography
4. Drinking
Sept, 2001
4 Parole Violation
Code Prior Sex Offences
Behaviour
Date
1. Sex assault Jill
Aug, 1995
2. Sex assault Joan
Sept, 1996
3. Sex assault Sue
Dec, 1996
Conviction
Date
Sentence
3. Sexual Assault
(Sue)
March,
1997
3 years
1. Sexual Assault
(Jill)
Sept, 1998
2 years
concurrent
2. Sexual Assault
(Joan)
Jan, 2000
3 years
probation
Code Prior Sex Offences
Date
July 1999
Charge
Forcible
Confinement
Sexual Assault (2
counts)
Feb, 2003 Sexual Touching
(child)
Indecent act
Conviction
Sentence
Forcible
20 Months &
Confinement 3 Yrs. Prob.
Indecent act
18 months
#6 Prior Sentencing Dates
• An appearance in court where the offender
receives a sentence for a crime or cluster of
crimes
• The crime must be sufficiently serious that
incarceration or community supervision are
possible sentencing options
• The actual punishment could be minor
– fines, conditional sentence
Prior Sentencing Dates
These things count:
• Findings of “Not Criminally Responsible” count as a
sentencing date
• Juvenile offences count (if known)
• Juvenile moved to a more secure placement as the result of a
sexual or violent incident
• Suspended Sentences count
• Parole Board (or similar) extends incarceration for a criminal
offence – Must extend total sentence - these count as prior
sentencing dates
• Military Court Martial Judgments for criminal behavior that
includes a sanction
– Purely “Military” charges (Failure of Duty, Not following a Lawful
Order) Do Not Count
#6 Prior Sentencing Dates
•
•
•
•
Do not count the index offence
Prior to the Index offence
Do not count charges or acquittals
Do not count prison misconducts or parole violations
(unless offender has a life sentence)
• Not Criminally Responsible does count
• Juvenile offences count
• Same type of charges within one month considered part
of same spree (one occasion)
#6 Prior Sentencing Dates
Do Not Count,
–
–
–
–
–
–
Overturned on Appeal
Not Guilty
Where fine only possible sanction
Historical Offences (Pseudo-recidivism)
Post-Index Offences
Fail to Appear (only count the date on which a
sentence is received)
– Where Juvenile detention is extended without a
new crime/charges
#6 Prior Sentencing Dates
On Conditional Release – Returned to Custody
– Indeterminate Sentence - “Lifers” – “Dangerous Offenders”
(Canada) – Offenders with already imposed indeterminate
sentences - Guys get “Yanked”
– Returns to custody count as a Sentencing Date
– This requires a high standard of certainty – You have to be
sure that, were the offender not already under criminal justice
sanction, that they would be charged with a new criminal
offence
– Needs to be a new criminal offence
– Returns to custody based on “Technicals” do not count
(“Drinking”, “Failure to Report”, “Presence of Minors” “Association
with Known Felons”)
#7 Convictions
for Non-contact Sex Offences
• Behaviour - Not the name of the offence
(Gross Indecency - exposure to anal sex)
– Only count convictions
– Exhibitionism, voyeurism, obscene phone calls, possession
of child pornography, using the Internet for sexual
purposes, stealing underwear/shoes for sexual purposes
– Attempted contact offences are not counted (e.g.,
invitation to sexual touching)
– Do not count soliciting/prostitution
– None = 0
Any convictions = 1
#7 Convictions
for Non-contact Sex Offences
• “EVER” (Prior or Index)
• “POST” (Creates new Index Offence)
• “Trespass by Night” & “Criminal Trespass”
may count as Voyeurism
• “Disorderly Conduct” for “mooning” may
count – you consider the motivation for the
behavior
• Plea Bargains – If “Pled Down” from a contact
offence – treat as a contact offence
#7 Convictions
for Non-contact Sex Offences
• Internet Crimes
– Is this a new form of crime?
– Or, a modern expression of older crimes?
– We see this as a modern expression of older crimes
– like obscene telephone calls – Hence, these are
scored as Non-contact sexual Offences.
Big Bad Bart’s Criminal Record
Date
Charges
Convictions
Disposition
Aug. 1993 Trespass by Night
(Voyeurism)
Trespass by Night
(Voyeurism)
1 Yr Probation
May 1994 Threatening
(Non-sexual)
Threatening
(Non-sexual)
1 Yr Probation
Jan. 1996 Sexual Assault
Sexual Assault
18 Months &
2 Yrs Probation
June 1998 Poss. Stolen Property
Poss. Stolen Property
90 Days &
1 Yr Probation
Nov. 1999 Sexual Assault
Assault C. B. H.
Sexual Assault
Assault C. B. H.
6 Years
Items #8, #9, & #10
The Three Victim Questions
• Victims from Sex Offences Only
• You need to know the degree of pre-offence
association between the Perpetrator and the Victim
• Animals – Do not count as victims
• “Accidental” victims - Do not count
• Citizens who happen upon
• Officers or workers in the performance of their duties
• Story of 4-year-old boy
» – Rape – Lewd and Lascivious Act on a Minor
Items #8, #9, & #10
The Three Victim Questions
• Not Guilty/Acquitted – You may score victims
based upon “Balance of Probabilities” (Soothill
et al., 1980)
• Conviction – But No Victim - Consensual – but
prohibited by statute
– “Statutory Rape” cases where the contact was
consensual
– Ontario – Charges of Sodomy – If consensual peers
– no victim scored
Items #8, #9, & #10
(p 49)
The Three Victim Questions
• Children Represented In Pornography - not
victims
– Do not count photographic or digital victims
– Real “live” children used to produce pornography
are victims
– You may refer to assumed preferences shown by
pornography in another section of your report
Items #8, #9, & #10
The Three Victim Questions
• Exhibitionism
– If mixed group assume only females unless you
have evidence males were being targeted
– Psychotic Street Person – fountain shower – No
Victims – Use judgment
Items #8, #9, & #10
The Three Victim Questions
• The offenders’ perception of the victim is important
– Transvestite – if the offender thought his victim was a
female, then score as “female”
• Internet “Stings”
– The Intent of the communication is important
• If he thinks he’s sending it to a 13 year old boy – he is – even if it’s
a cop on the other end – you have an identifiable victim
• If he thinks he’s sending it to another adult but it turns out to be a
child posing as an adult on the internet there is no victim
Items #8, #9, & #10
The Three Victim Questions
• Prowl By Night – Voyeurism (p 50)
– Even if male in home – assume female victims only unless
you have reason to believe offender was watching the male
• Victim Not At Home (p 51)
– Masturbate on the bed, Steal underwear and masturbate
– Assume female victims only unless you have reason to
believe offender was attracted to the male
• Sex with Dead Bodies (p 50)
– Victim information counts
– Assess level of pre-offence victim – perpetrator relationship
Items #8, #9, & #10
Polygraph
• Polygraph information was not included in the original
scoring of the STATIC-99 because that data was not
available for the samples of origin
• Never use polygraph information alone to score the
STATIC-99
• Information obtained through disclosure polygraphs
that is then corroborated by outside sources can be
used to score victim questions
• Information obtained through disclosure polygraphs
may be used if the information is sufficient to support
a new criminal investigation
#8 Any Unrelated Victims
•
•
•
•
•
•
Are they too close to marry?
Step-relatives - more than 2 years
Wives count as related
Common-law - more than 2 years = related
Do not count Category “B” victims
Do not count “accidental victims” such as a
police officer or someone who observes the
offence
#8 Any Unrelated Victims
• See “The List” in the coding manual
• Step-relations
– Consider length (two-year rule) and nature of
the pre-offence degree of relationship
– fraternal/paternal/family
• Becoming Unrelated
– Mother and child separated at birth
– Perpetrator has no knowledge he is offending
against a family member
Examples of relationships that would
not be considered “related” for
STATIC-99 scoring
•
•
•
•
Step-relations lasting less than 2 years
Nephew’s wife
Second cousins
Wife’s aunt
If in doubt, consider the psychological relationship prior
to the assault – the offender must have a “family” type
relationship of at least two years duration
#9 Any Stranger Victims
• Has the offender ever committed a sexual
offence against a stranger? (24 hour rule)
• Victim did not know the offender 24 hours
prior to offence (and/or the reverse)
• Do not count “accidental” victims
• 1 point for having a stranger victim
• Note: If stranger - also code unrelated !!
#9 Any Stranger Victims
• It does not take much to be “known” and not a
stranger any more – but does take some
interaction (p 54)
• Where is the Line? At the Corner Store
– The woman who works the register, has sold the
perpetrator cigarettes on three or four occasions and
has spoken to him slightly would not be a stranger
– The woman who stocks the shelves, has seen the
perpetrator in the store on three or four occasions but
has not spoken to him would be a stranger
#9 Any Stranger Victims
• The Internet, telephone and e-mail (p 54)
– The “24 Hour” rule applies
– If the victim and the perpetrator first make contact at 8 p.m.
on a Wednesday the offence must happen or the first faceto-face meeting must take place by 8 p.m. on Thursday
– If they chat back and forth for more than 24 hours they are
no longer strangers
• Becoming a “Stranger Again”
– Victim forgotten completely (over years) Perpetrator thinks
he is assaulting a complete stranger – Score this person as a
stranger victim (p 54)
#10 Any Male Victims
• Do not count
– Pornography offences
– Exposure to a mixed group of children (unless
targeting the males)
– Do not count “accidental” victims
• 1 point for having a male victim
#10 Any Male Victims
• Assault of a Transvestite/Transsexual
– If the perpetrator assumed he was assaulting a female
do not score male victim (p 56)
• Makes male watch or forces him to take part in
the sexual offence
– Score the male as a victim
– If just confines the male – No male victim (p 56)
Big Bad Bart’s Victims
During the summer of 1993 Bart was caught several times trying
to peek in the bedroom or bathroom window of a pretty girl that
went to the same high school. After repeated warnings he was
finally charged with Trespass by Night.
At a 1995 New Year’s Eve party Bart asked a woman he did not
know to dance, she declined. Later, after loudly discussing the
snobby nature of women with several other inebriated males he
reached over the woman’s shoulder from behind and fondled her
breasts while pulling her body against him.
Bart met Becky in Toronto, Becky was 13. As Bart got further
into selling drugs he got an apartment where Becky would stay
and hand over drugs to people he sent over. He beat Becky badly
and raped her after she used too much and told Bart not to touch
her. Becky’s doctor convinced her to talk to police while
recovering in hospital.
STATIC-99 Total Score
•
•
•
•
•
The STATIC-99 total score is the sum of the 10 items.
Scoring range 0 to 12.
Mean (SD) in development sample: 3.15 (1.97)
Median value in Canadian norms: 2
Interpretive ranges (percentages in Canadian Norms)
–
–
–
–
0-1 low (31%)
2-3 low-moderate (42%)
4-5 moderate-high (20%)
6+ high (9%)
Presenting STATIC-99 Scores
• STATIC-99 as part of a complete assessment
• Consideration of factors external to STATIC-99
– Advanced age/ill health
– Criminogenic needs/stable dynamic risk factors
– Completion of credible treatment program
– Stated intentions to reoffend
• Data “Over-ride” caution
Years offence-free
in the community
• If offenders are able to remain in the community
two to ten years without another serious offence,
their chances of sexual recidivism decreases
substantially
• Offence free:
– no new sexual or non-sexual violent offences
– no offences that result in long periods of incarceration
• New tables coming
Presenting STATIC-99 Scores
• Standard Paragraphs
– Background/origin of scale
– Results for offender (score and percentile)
– Summary of research on new norms and
moderator variables
Presenting STATIC-99 Scores
• Standard Paragraphs: Recidivism rates
– Recidivism rates for similar offenders bounded
by routine CSC samples and preselected highrisk samples
– Evaluator must make professional judgment as
to characteristics offender shares with both
groups: where does this offender fall within that
range?
– Clear distinction between reporting actuarial
estimates and making a professional judgment
Presenting STATIC-99 Scores
• Standard Paragraphs
– Other factors considered in evaluation
– Don’t try to “second-guess” or “over-ride”
actuarial scales – go to dynamic assessment
Interpreting the STATIC-99
•
You MAY NOT say in your report that “research
has shown that Mr. X’s estimated recidivism
potential over the next Y years is ?? percent.”
•
You MAY say in your report that “research has
shown that groups of men demonstrating the same
empirical risk factors as Mr. X have been seen to
recidivate at ?? percent over Y years.”
Risk for Female Sexual Offenders
• Low Sexual Recidivism Rates (<5%)
• Recidivism Risk Factors are Unknown
– Cortoni & Hanson (2005) found only 3 sexual
recidivists among 306 female sex offenders
(1% sexual recidivism rate).
– We recommend that you use a risk scale
designed for general or violent recidivism that
has been validated on female offenders (e.g.,
LS/CMI, Andrews & Wormith, 2004)
STABLE
STABLE - 2000
Developed from:
• SONAR (Hanson & Harris, 2000)
• STEP (Beech et al., 2002)
• SRA (Thornton, 2002)
• Explicit, structured risk assessment tool
– 16 Items
– Combined with STATIC-99 into overall risk
• Empirically Informed, but needed validation
STABLE – 2007
5 sections for a total of 13 Items
•
•
•
•
•
Significant Social Influences
Intimacy Deficits
General Self-regulation
Sexual Self-regulation
Co-operation with Supervision
(Please pull Tab 8a STABLE-2007 Tally Sheet)
STABLE - 2007 Scoring
• All available information
• Historical and recent
• STABLE - typical or base line functioning past year and next year
Will the change endure?
90
80
70
60
50
40
30
20
10
0
01
20
99
19
97
19
19
95
Karl
Andrew
• Exceptional
circumstances?
• Major life change?
• Opportunity?
• External pressure?
• Quality of evidence
• New base line?
Return to Base Line
90
80
70
60
50
40
30
20
10
0
01
20
99
19
97
19
19
95
Karl
Andrew
New Base Line
90
80
70
60
50
40
30
20
10
0
01
20
99
19
97
19
19
95
Karl
Andrew
Changes in Scoring
STABLE-2000 to STABLE-2007
• “Attitudes” section gone
Slight changes in three other items,
• “Intimacy Deficits” becomes “Capacity for Relationship
Stability” - (New two-part scoring)
• “Emotional Identification with Children” - Now only scored
for child molesters
• “Deviant Sexual Interests” – Offence history must be taken
into consideration
Also,
• Easier scoring for Significant Social Influences
• Easier Tally Sheet – Easier to calculate final score
• New Empirically-based Category cut-offs
Significant Social Influences
• Name all the people in the offender’s life
who are not paid to be with him. For each
one, is the influence positive, negative, or
neutral?
#positive – #negative = balance
Balance
2 or more
= 0
0, 1
= 1
Any negative = 2
Significant Social Influences
Person’s First
Name
1
2
3
4
Mom
Nature of Relationship and
comments
Provides $ --- Provides place to live
Watches out for him
Thinks we’re picking on him
Tells him he doesn’t need meds
Influence
Pos/Neut/Neg
Neutral ?
Significant Social Influences
Person’s First
Name
1
2
3
4
Mom
Willy
Nature of Relationship and
comments
Influence
Pos/Neut/Neg
Provides $ - Provides place to live
Watches out for him
Thinks we’re picking on him
Tells him he doesn’t need meds
Neutral ?
Old HS friend – No drink/drugs
Takes him to hockey
Gives him pro-social advice
Positive
Significant Social Influences
Person’s First
Name
1
2
3
4
Mom
Willy
Nature of Relationship and
comments
Influence
Pos/Neut/Neg
Provides $ - Provides place to live
Watches out for him
Thinks we’re picking on him
Tells him he doesn’t need meds
Neutral ?
Old HS friend – No drink/drugs
Takes him to hockey
Gives him pro-social advice
Positive
“Jokes around” at work
Jason W.
Wants him to go for beers at the strip club
Possible drug involvement???
Negative
Scoring
Significant Social Influences
• Now – Ignore “Neutrals”
• Total “Positive” and “Negative” influences
separately
• Find Item final score from scoring table on
back of STABLE-2007 Tally Sheet (Tab 08a)
Note: All possible combinations not shown on following page – always use
table on back of Stable-2007 Tally Sheet (Tab 08a)
Scoring Social Influences
Positive
Negative
Final Score
0
0
1
0
1 – or more
1
0
1
1
1
1
1
2 – or more
2
0
2
1
1
2
2
1
2
3 – or more
3
0
0
3
1
0
3
2
1
3
3
1
3
4 – or more
4
0
0
4
1
0
4
2
0
4
3
1
4
4
1
5 – or more
2
2
0
2
2
0
Significant Social Influences
Things to consider
• Does this person provide material support?
• Does this person undermine the offenders controls?
• If the offender went to that person for advice would that
person be likely to give pro-social or anti-social advice?
• Circles of Support and Similar – if the formal structure
dissolved would that (positive) person still go for coffee
with the offender?
• The magic question: If you had a magic wand and could
“zap” that person out of the offender’s life – would the
offender be more or less likely to reoffend?
– More likely to reoffend – they are a positive influence
– Less likely to reoffend – they are a negative influence
Intimacy Deficits
•
•
•
•
•
Capacity for relationship stability
Emotional identification with children
Hostility toward women
Social rejection/loneliness
Lack of concern for others
Capacity for Relationship Stability
A two-part question
“A” Part: Has this offender ever had a two-year
intimate (sexual & “live-in”) relationship with
an appropriate adult partner (STATIC-99, Q#2)
“B” Part: Is the offender currently living with an
intimate partner in a relationship without
obvious problems?
• This relationship can be short, but should be
expected to be reasonably STABLE.
“A” Part:
Capacity for Relationship Stability
• Same question – taken from STATIC-99
• Ever lived with lover for 2 continuous years?
– must be continuous
– prison marriages/lovers don’t count
• Scored as a “Yes” or “No”
• For complete coding rules for “A” Part please see
STATIC-99 Coding Rules – Revised 2003
“A” Part:
Capacity for Relationship Stability
Who can you have lived with???
• Legal marriages of < 2 years do not count
• Prison marriages (while guy incarcerated) do not count
– If relationship still there > 2 years after he gets out - does
count
• Non-human species – do not count
• Priests and other celibates – no exemption
• Must be a relationship that is legal
• Must be age to consent to relationship
– Child “vics” do not count – even if “consensual”
“B” Part:
Capacity for Relationship Stability
0 Current live-in lover/intimate partner - No obvious
problems
1
Living with a lover/intimate partner - but the
relationship is troublesome or problematic
Offender does not believe that it will last
There may be lots of fights
Someone is having an affair
Someone is routinely unfaithful
STABLE dating relationship that does not involve
living together
2
No current lover/intimate partner
Scoring Capacity
for Relationship Stability
“A” Part
“Ever Lived”
“B” Part
“Current”
No
No
No Current (2)
Poor/Stable dating (1)
No
Yes
Yes
Acceptable (0)
No Current (2)
Poor/Stable dating (1)
Yes
Acceptable (0)
Final Score
2
1
0
Emotional Identification with Children
Note: Only score for those with child victims age 13 or less
• Does the offender feel emotionally close to or intimate with
children?
• Sees children as peers or equals
• Relates more easily to children than to adults
• Not parent-child relationship
• Consider not only attitudes and values, but also leisure and work
activities suggestive of a child-oriented lifestyle
• Involved in children’s activities
• Boy-victim child molesters tend to be childlike themselves
• Offender ascribes adult qualities to children (Incest)
Emotional Identification with Children
Note: Only score this item for child molesters with
at least one victim age 13 or less
• Clarification
– This cut-off is empirically based
– Congruent with DSM-IV-TR (pg. 571)
– Teen victims, those who have attained their 14th
birthday at the time of the assault are not counted on
this item
– The victim’s age at the first sexual assault
– Score a “0” or a “N/A”
Emotional Identification with Children
0
No obvious identification with children
1
Immature relationships with adults
May see children as having special qualities
of understanding or communication
Demonstrates some interest in age-inappropriate (childlike)
activities or pastimes
2
Obviously feels more comfortable with children than adults
No adult friends – Has children as friends
Strong child-oriented interests, pastimes or activities
Scoring Example 1 - John
Convicted of “Demand Sexual Touch – Child”, John, age
35, lives alone and has no adult friends. He works as a
janitor at the recreation facility. He generally keeps to
himself, but he does talk with some of the regulars,
including a couple of the pre-teens who come for the
children’s programs. When asked, he says that he likes
children, and that he would rather play ball with the kids
than watch TV on his own. John has never been in a
steady heterosexual relationship. He says that he would
like to date more often, but fears rejection (he is not
physically attractive).
Scoring Example 2 - Fred
Fred was a school principle and was caught
surreptitiously masturbating in the school library. has
been convicted of exhibitionism. He is married and has
two children, ages 8 and 10. His relationship with his
wife is distant, and he has no close friends outside his
family. Fred appears to live through his children. All
of his free time is spent supporting his son’s hockey and
his daughter’s competitive country dancing. Most of his
conversation centres on the special talents of his
children. His emotional highs and lows follow their
successes or failures.
Hostility Toward Women
• A prejudice, making women into a different class
unworthy of trust or respect
• Unable to form warm, constructive relationships
with women
• Believes or endorses sexist attitudes
• Does not consider women as people worthy of
trust and respect
• May have sexual or personal relationships with
women, but these relationships are adversarial and
conflicted
Hostility Toward Women
0 Is comfortable with women and has female friends with whom
he is not sexually involved at this time
Has no female friends but has no conflicts with women
1 Some conflicts with more than one woman in more than one
environment
2 Frequently in conflict with women
• Doesn’t believe that males and females can be “just
friends”
• Believes “Women only good for sex”
• Believes women can not be trusted
• Consistently dismissive of woman’s opinions
Scoring Example 3 - Richard
Richard has been convicted of Forcible Confinement. He
is 28 years old and works on and off as a night-club
bouncer. He likes to party, dance, drink, and socialise
with his male friends from high school. He is out on a
date almost every week, although rarely does he see the
same women for more than a month. He divides women
into those that are “fun” and those that are “boring”. Fun
women drink, flirt and are open to casual sex. Boring
women either don’t pay attention to him, or want a longterm commitment. He does not get into conflicts with
women. When problems arise, he simply ignores them or
ends the relationship.
Social Rejection/Loneliness
• Is the offender able to make friends and feel
close to others (demonstrating secure adult
attachment)?
• Is he lonely, prone to feeling socially rejected?
• Is he emotionally close to friends and family?
• How does he feel over the intermediate term –
his impression of the world
Social Rejection/Loneliness
0 Generally well integrated socially considering their level of
social standing and the process of social upheaval inherent in
having been convicted of a serious sexual criminal offence
1 There are some weak connections with others; some short-term
casual relationships, but no long-term friends. Has no close
relationships with others but does not feel lonely or rejected
(“the loner”).
2 Frequently feels lonely and rejected
No social supports - Poor skills in attracting and maintaining
close personal relationships.
Lack of Concern for Others
• This item does not reflect solely their treatment of their
victims
• Little consideration for the feelings of others
• Acts according to their own self-interest
• Feigns shallow displays of regret, little or no remorse
• Unfeeling, ruthless, or indifferent
• Not just towards their victims or adversaries, but also
towards their in-group
• Possibly has friends, associates and acquaintances, but no
stable, caring relationships
• Quite significant pathology must be present, this condition is
fairly unusual
Lack of Concern for Others
The “self-concept is too narrow and self-absorbed” …
lacking the “awareness of the interconnectedness of
human beings”
A person whose “best self” is focussed almost
exclusively on his or her own self-interest will neither
notice nor care that others have been harmed”
Aaron Lazare
On Apology
(2004) pg. 115
Lack of Concern for Others
0 May be callous/indifferent to some people (e.g.,
adversaries) in specific circumstances, but is
generally emotionally responsive and caring
1 Significantly callous/indifferent in at least one
context (victims, business, not just sex offending), but
shows warmth and concern in some close
relationships (e.g., family, long-term friends)
2 Typically shows little remorse or concern for others
Most interactions are utilitarian, with little warmth or
attachment to others
Scoring Example 5 - Jim
Jim, age 33, works as a construction labourer. He says he
has a few friends, but has not known any for more than 6
months. He was living with his mother, but she recently
evicted him after she caught him pawning small
household articles. He says it was all a big
misunderstanding; he was going to pay her back but now
that she kicked him out he isn’t going to bother. He has
had a number of short-term sexual relationships,
including living with a women for almost a year. Once
she became pregnant, he left and he has had no further
contact with her or the child.
Scoring Example 6 - Ian
Ian lives with his brother. They work together and they
often socialise with other men from the factory. Ian is
generally well liked by his friends and family, but they
describe him as having a mean streak. All of his intimate
female partners have left following beatings. He says he
is glad they are gone. On at least two occasions, he has
been in bar fights in which strangers were left badly
beaten. He shows no remorse for the victims of these or
other crimes, and typically implies that the victims
deserved what they got.
General Self-Regulation
• Impulsive Acts
• Poor Cognitive Problem Solving Skills
• Negative Emotionality/Hostility
Impulsive Acts
• Easily swayed by opportunistic circumstances
• Behavior that has a high likelihood of negative
consequences
• Easily bored, seeks thrills and has little regard for
personal safety or the safety of others
• Impulsive across several settings – not just
represented by his history of sexual offending
Impulsive Acts




reckless driving
substance abuse
“getting into” partying
accepting bets and
dares
 quitting jobs with no
other job in sight
 changing residences
 unsafe work practices
 starting fights with
men much bigger than
himself
Impulsive Acts
0 No problems, or limited only to sexual
misbehaviour
1 Occasional impulsive behavior
Repeated high risk behavior in only one
context (e.g., frequently gambles, but no other
obvious impulsive acts)
2 Frequent impulsive behavior in more than
one setting beyond their sexual offending
Poor Cognitive Problem Solving
• Difficulty accurately identifying and solving
problems
• Proposes unrealistic solutions – (or none at all)
• Unable to choose appropriately between
competing possible options (Always takes the
easiest or the one with the most immediate “payoff”
• Lacks long-term plans
• Fails to recognize the consequences of their
actions
Poor Cognitive Problem Solving
• Problem identification
• Generating alternatives
• Evaluating alternatives
Poor Cognitive Problem Solving
0 Is able to appropriately identify and address
typical life problems
1 Some poorly considered decisions, but open
to correction when difficulties are pointed out
2 Frequently makes poor decisions
Fails to identify obvious life problems
Difficulty generating workable alternatives
Difficulty recognizing negative consequences
of actions even when pointed out to them
Negative Emotionality/Hostility
• A feeling of almost constant grievance is key to this item
• This is not the “blue” guy – this is the guy with “A chip on his
shoulder” – a grudge against the world
• Prone to feeling hostile, victimized, and resentful
• Vulnerable to emotional collapse when stressed
• Although possibly linked to real grievances, the offender’s
emotional response is excessive
• Rather than attempting to cope constructively, the offender
ruminates on the negative events and feelings and may appear
to be “getting into it”
• Your helpful suggestions are dismissed or belittled
• Explosive expressions of emotion, quickly over
Negative Emotionality
0
Occasional expressions of grievance but not beyond that
which would be reasonably expected given the offender’s
life situation
1
Some hostility or resentment beyond what would be
reasonably expected, but offender appears to cope
constructively. Tries to get over it and move on
2
Clings to resentments and ruminates on small life setbacks
Rumination on negative emotions and negative life events,
self-indulgent self-pity, tends to “give up”
Includes irrational feelings of persecution and chronic
suspiciousness
Sexual Self-regulation
• Sex drive/Pre-occupations
• Sex as coping
• Deviant sexual interests
Sex Drive/Pre-occupations
• Recurrent sexual thoughts and behaviour
– ( not directed to a current romantic partner)
• Casual or impersonal sexual activity
• Interference with other pro-social goals
• Perceived as intrusive or excessive by the
offender
• Or just plain excessive
Sex Drive/Pre-occupations
 Masturbation most days (15+ times a month)
 Regular use of prostitutes, strip bars, massage parlours,
phone-sex and phone sex bills
 Sex-oriented internet use, such as sexually explicit sites,
chat rooms - Large amounts of time “surfing the web” for
pornography sites
 Pornography collection (videos, magazines) (or,
parent/baby magazines)
 Cruising for impersonal sex
 A history of multiple sexual partners (e.g., 30 or more)
 Excessive sexual content in typical conversations
 Pre-occupation with own/other’s sex crimes
 Self-report of difficulty controlling sexual impulses
 Any disturbing sexual thoughts
Sex Drive/Pre-occupations
0 No evidence of impersonal sex or sexual preoccupations
1 Some evidence of impersonal sex
Regular use of pornography for sexual
gratification
Some evidence of sexual pre-occupations
2 Clear evidence of any sexual pre-occupations
Some evidence of multiple pre-occupations
Sex as Coping
• Life stress and negative emotions trigger
sexual thoughts or behaviour
• Content may be normal or deviant
• This coping behaviour will be seen in
multiple life domains (in response to work
stress, family stress, interpersonal stress)
• Sexual expression to dissipate anger,
humiliation, or frustration
Sex as Coping
0 No history of using sex as a coping strategy, or, if a
history, during past year has repeatedly experienced
negative life events without resorting to sex as
coping
1 Occasional lapse into sexual fantasy or behaviour
when stressed but not the typical reaction. Has other
coping strategies not including sexual expression
2 Negative emotions or life events typically invoke
sexual thoughts or behaviours
Deviant Sexual Interests
• Sexual interest in people, objects, or activities that
are illegal, inappropriate or highly unusual.
– children, non-consenting adults, voyeurism,
exhibitionism, cross-dressing, coprophilia, and
fetishism
• Assessed by number of sex offence victims,
number of deviant preference victims, self-report
of deviant history or preferences, or the results of
specialised testing (e.g., phallometrics)
Deviant Sexual Interests
“Behavioural History”
• You must consider both frequency and the
unusualness of the behavior
• Sex with pre-pub boy (X1) = 1
• Masturbation in car:
– caught once = 1
– caught more than once = 2
Deviant Sexual Interests
(Please see Table – page 35 Tab 08)
Domain
Number of Sex Offence
Victims
Count/Criteria
0 = Only one victim
1 = 2 to 7 victims
2 = 8+ victims
0 = No deviant victims
Number of Deviant
Preference Victims/Activities 1 = One deviant victim
•Pre-pubescent child victims etc.
2 = Two or more deviant victims
Self-report of deviant history
or preferences (or observed
behaviour)
0 = Endorses only normal fant/prefs
1 = You suspect deviant fant/prefs
2 = Describes or admits to dev. fant/prefs
Results of specialized testing
Not Scored = No evidence testing ever offered
0 = Testing – results show no dev. Prefs
1 = Mixed Results – Possible deviance
2 = Deviant preference shown in testing and
nothing done about it
Score
Deviant Sexual Interests
Scoring Notes:
At the time of Static-99 construction, the age of
consent in Canada for sexual activity was 14 years
old. People who have reached their 14th birthday are
not considered “child” victims.
Physically developed (mature) 12 and 13 year olds are
not considered “deviant” sexual victims (sub-section
two) – they are simply victims. If the victims have a
“mature”, “developed” or “adult” body shape they are
not considered “child” victims.
Scoring Deviant Sexual Interests
Of the four sections – the highest score in any
section is the score for the whole item
Domain
Count/Criteria
Number of Sex Offence Victims
0 = only one victim
1 = 2 to 7 victims
2 = 8+ victims
Number of Deviant Preference
Victims/Activities
•Pre-pubescent child victims etc.
0 = No deviant victims
1 = One deviant victim
2 = Two or more deviant victims
Self-report of deviant history or
preferences (Or observed
behaviour)
0 = Endorses only normal fant/prefs
1 = you suspect deviant fant/prefs
2 = Describes or admits to dev. fant/prefs
Results of specialized testing
Pass = No evidence testing ever offered
0 = Testing – results show no dev. Prefs
1 = Mixed Results – Possible deviance
2 = Deviant preference shown in testing
and nothing done about it
Score
Deviant Sexual Interests
Deviant Sexual Interests in Possible Remission
• An offender who has scored a “1” or a “2” based upon
historical facts can have their Deviant Sexual Interest
score reduced by one point if the following is present:
• The offender is involved in an age appropriate,
consensual, satisfying sexual relationship of at least
one years duration while “at risk” in the community
with the absence of behavioural indicators of Deviant
Sexual Interests for 2 years
• Presence of this relationship requires credible,
independent, collateral confirmation of the
relationship
Scoring Example 7 - Bob
Bob, now age 43, had 15-20 sex partners before marrying
when he was age 24. Bob wanted her to do the things
Susan he saw in porno videos, but she lost interest and left
three years later. By that time, Bob was already involved
with another women. He never remarried, but has almost
continuously pursued casual encounters. He says that he
likes the chase and that he can pull himself out of a funk
by thinking about his past or future conquests. He
describes his current offence of sex with a 15 year old
neighbour as a novelty that he does not want to repeat.
When in prison, he masturbates three times a week.
Scoring Example 8 - Mike
Mike, age 22, has had 4 female and 3 male sexual
partners. He considers himself bisexual. He masturbates
1-2 week, usually thinking about anal sex with
consenting partners of vague (and changing) genders.
His current offence involved tying up a female
acquaintance, forcibly removing her clothes, sexually
assaulting her for several hours, then falling asleep. In
the morning he released her, drove her home and asked
what she was doing next weekend. Once free, she went
directly to the police. He denies any interest in sadomasochistic sex, claims it was all her idea and that it was
the only time he tried it.
Scoring Example 9 - Norman
Born in 1976, Norman was found guilty of sexual assault of a
minor child (boy) in 1997 – he served 30 months incarceration
and was then released to a 10-year intensive supervision order.
Phallometric testing done pre-trial indicated preference for prepubescent boys over adult males and females. During the course
of treatment he admitted to sexual involvement with another male
child (never charged) and stated that at the time of the crimes he
had no idea why he was interested is sex with young boys. He
now believes that this was part of his personal struggle with being
homosexual. While in prison he met a male church volunteer.
Upon release this quickly developed into a relationship and the
men have been living together for five years now. Both indicate
considerable commitment to and satisfaction with the relationship.
Phallometric assessment done in 2008 showed significant arousal
to adult males, no other categories were significant.
Cooperation with Supervision
• Do you feel that the offender is working
with you or working against you?
• Does he see himself as at no risk to reoffend
and place himself in high-risk situations?
• Does not take seriously the conditions of
supervision?
Non-cooperation
• Disengagement:
– just going through the motions, silent/non-disclosing,
keeping secrets, not invested in treatment
• Manipulation:
– trying to “play the system”, trying to be “buddy-buddy”
with you, trying to lie to you and deceive you, asking
for special favours, engaging in the manipulation of
helpers (e.g., playing one off against another)
• No Show:
– often shows up late or at the wrong times, fails to attend
scheduled appointments with you and others
Cooperation with Supervision
0 Offender appears to be working with you
Regular attendance, follows through with
instructions
1 Some problems, but generally cooperative
Occasional missed appointments
2 You perceive the offender to be uncooperative,
deceptive, manipulative or disengaged
Frequently late, missing appointments, no sense that
you know what is going on with him
Think of your normal interview
• Significant Social Influences*
• Who do you hang around with?
• What do you do with Joe?
• Intimacy Deficits*
•
•
•
•
Anybody special in your life?
How do you get along with women? Children?
Would you say you are a loner?
Who do you care most about in the world?
• General Self-regulation*
• Ever play sports? How often in Emerg? Money problems?
Housing problems? Quit jobs a lot?
• What sort of things cause you problems in your life? What do
you do about them?
• What do you think of (the guy who caused the problem)? And
do you meet people like that a lot?
Think of your normal interview
• Sexual Self-regulation
• Sexual outlets? Habits? How often? Pressure? Relationship to
self-esteem and self-perception
• Ever use sexual outlets to change your mood or make yourself
feel better?
• You mentioned a child in your list of friends. You seem to
have a history of people making allegations against you
• Co-operation with Supervision* (Your call)
(You can re-order the STABLE interview to suit yourself)
STABLE - 2007 Total Score
•
•
•
•
•
See STABLE-2007 Tally Sheet (Tab 08a)
12 items for non-child molesters
13 Items for child molesters
Each Item worth 2 points
Sum the 13 Items
Interpretative Ranges
0-3
4 - 11
12 +
Low
Moderate
High
Combining STATIC and STABLE
Empirical Rules
STATIC-99
Risk Category
STABLE-2007
Need Category
Overall
Supervision
Priority
Low
Low
Moderate
High
Low
Low
Moderate-Low
Moderate-Low
Low
Moderate
High
Low
Moderate-Low
Moderate-High
Moderate-High
Low
Moderate
High
Moderate-Low
Moderate-High
High
High
Low
Moderate
High
High
High
Very High
Deviant Sexual Interests
in Possible Remission
• This adjustment (reducing the score by one point)
does not factor into the total score of the
STABLE-2007.
• This adjustment has not been empirically validated
or examined retrospectively on the DSP samples.
• Jurisdictions implementing this sub-item should
collect data to determine its predictive validity.
Treatment
What does all this mean?
• STABLE factors are your best treatment targets
• Use STABLE assessment to inform your treatment and
supervision efforts
• STABLE assessment represents the beginning of
“diagnostic” treatment assessment for sexual offenders
• Does it make sense? - Is it a good use of money to put
everybody through everything
“Child molesters are not a homogeneous
group, so forcing them to address in detail
all aspects of a uniform program
would be unwise."
Marshall, W. L. (2004). Cognitive-behavioural treatment of
child molesters. In R. Karl Hanson, Friedemann Pfafflin, &
Manfred Lutz (Eds.). Sexual Abuse in the Catholic Church:
Scientific and Legal Perspectives: Proceedings of the conference
"Abuse of children and young people by Catholic Priests and
Religious". Citta del Vaticano: Libreria Editrice Vaticana,
Pontificia Academia Pro Vita. Page 98.
Tab 11
DSP
ACUTE
Acute Risk Factors
• Short term risk
• Timing of reoffense
• These factors represent current expressions
of problematic (risky) of risky behaviours
• Note: Research data shows that an average
rating over time (4 mos) performs better
than any individual assessment
• Hence, do ACUTES often and regularly for
best prediction
ACUTE PREDICTORS – Two Factors
Sex/Violence Score
General Recidivism Score
(Four Items)
(All seven items)
•
•
•
•
•
•
•
•
•
•
•
Victim Access
Hostility
Sexual pre-occupation
Rejection of Supervision
Victim Access
Hostility
Sexual pre-occupation
Rejection of Supervision
Emotional Collapse
Collapse of Social Supports
Substance Abuse
[predicts all types of relapse]
Same Behavior – Different Scoring
In the community the same behaviour can result
in scoring differences depending on the
characteristics of the offender.
1. Rapist riding a city bus to work with school children
– most likely a manageable risk
2. Child Molester riding a city bus to work with school
children
– may or may not be a manageable risk
3. Frotteur riding a city bus to work with school children
– most likely an unassumable risk
4. Rapist working the “late shift” at a courier company where
they hire an evening shift of university students
– possibly a manageable risk, if he is “part of the crew”
– most likely an unassumable risk if he’s made the “shift chief”
Scoring Chronic Conditions
• Is it serious enough to be a supervision target?
• To score a “2” there has to be some evidence of a
problem and this problem has to be a supervision
target
• If you have suspicion of a problem you can score a
“1” as a place-holder for one supervision.
• You can score “1” repeatedly if there is some
documented evidence that the problem remains and
it is a topic of supervision.
ACUTE PREDICTORS – Two Factors
Sex/Violence Score
Score
General Recidivism Score
(Sum of four factors)
(Sum of all seven factors)
• Victim Access


Copy Score Over



• Hostility


Copy Score Over





Copy Score Over





Copy Score Over



• Sexual Pre-occupation
• Rejection of Supervision
• Emotional Collapse
• Collapse of Social Supports
• Substance Abuse
Sex/Violence Total
General Recidivism Total
Score
Acute Rating System
•
•
•
•
0 1 2 IN -
No problem
Maybe a problem, not sure
Yes, a concern
Intervene now
Victim Access
• 0 - no problem
• 1 - incidental contact, not repeated/regular
incidental contact that can not be avoided
but no indications of victim approach
• 2 - repeated opportunity, hints of planning
– several different paths, he mostly/always
chooses the most risky one
• IN - clear planning, grooming, stalking,
hiding deliberate contact
Hostility
• 0 - no problems
• 1 - some resentment; harsh words
• 2 - heated confrontations; any physical
aggression, veiled threats, angry
rumination, something is bugging him
and you are aware there is a problem
• IN - direct threats, open plans of retribution
Sexual Preoccupations
• 0 - no problem
• 1 - slight concerns, increased masturbation
• 2 - rumination on sexual issues, sexual
tension, deviant urges, porno/strip
clubs, sex urges when angry or upset
IN - out of control of sexual urges, lots of
impersonal sex
Rejection of Supervision
• 0 - regular reporting, appropriate responses
• 1 - reluctant, missed appointments with others,
you are unsure of what is going on
• 2 - breaching conditions, missed 2
consecutive appointments, manipulative lying,
treatment dropout
• IN - brings weapon, attends drunk, new offences,
driving while suspended, he disappears
Emotional Collapse
• 0 - common misery
• 1 - stressed, but coping (strained)
• 2 - hopeless, helpless, negative emotional
rumination, self-pity, not coping,
paranoia
• IN - suicide risk, acting on paranoid impulses,
not caring what happens to them
Collapse of Social Supports
• 0 - no major changes
• 1 - threats to important relationships, loss
of minor relationships
• 2 - loss of significant social relationship or
social group, gain negative peer group,
initiates or re-joins a dysfunctional relationship
• IN - loss of essential supports, complete
community rejection, pro-paedophilic
clubs
Note: In most cases, the loss of a paid therapist does not count - unless it was a
particularly close and important relationship for the offender.
Substance Abuse
• 0 - no use
• 1 - some drinking, but not problematic
and not prohibited
• 2 - problem use; any prohibited use
• IN - out of control, interference in daily
functioning
Sex & Violence Risk and General Recidivism Risk
Sex/Violence Score
Score
General Recidivism Score
(Sum of four factors)
(Sum of all seven factors)
• Victim Access


Copy Score Over



• Hostility


Copy Score Over





Copy Score Over





Copy Score Over



• Sexual Pre-occupation
• Rejection of Supervision
• Emotional Collapse
• Collapse of Social Supports
• Substance Abuse
Sex/Violence Total
General Recidivism Total
Score
Note: Sexual and Violence risk calculated
using Sex/Violence factors (four).
General recidivism risk calculated using ALL
ACUTE Factors (seven)
Sex/Violence Nominal
Categories
0 = Low
1 = Moderate
2+ = High
General Recidivism Risk
Nominal Categories
0 = Low
1 – 2 = Moderate
3+ = High
Combining STATIC/STABLE and
ACUTE Risk Factors
STATIC- STABLE-2007
Low
ACUTE
SCORE
Low
Moderate
High
Moderate-low or
Moderate-high
Low
Moderate
High
High or Very-high
Low
Moderate
High
CURRENT
Priority
Low
Low
Moderate
Low
Moderate
High
Moderate
High
High
Relative ACUTE Priority
Risk Ratios: Relative Risk for Recidivism within
45 days based upon combined STATIC-99,
STABLE-2007 & ACUTE-2007 Scores
Low
Moderate
High
Sexual Crimes
1.0
2.4
14.9
Any Sexual
1.0
2.3
4.6
Violent or Sexual
1.0
1.8
7.0
Any Crime
1.0
1.9
4.5
Any crime,
Including breaches
1.0
1.6
4.3
Relative Priority of ACUTE Ratings
Implications for Supervision
“Basically”, as a heuristic, men who score
“Moderate” on this “three level” assessment
should receive twice the supervisory priority
as those who score “Low” and those who
score “High” should receive four times (X4)
the supervisory priority as those who score
“Low”.
Tab 9
DSP
Combining STATIC and STABLE - Empirical Rules
STATIC-99
Category
STABLE-2007
Category
Overall
Priority
Low
Low
Moderate
High
Low
Low
Moderate-Low
Moderate-Low
Low
Moderate
High
Low
Moderate-Low
Moderate-High
Moderate-High
Low
Moderate
High
Moderate-Low
Moderate-High
High
High
Low
Moderate
High
High
High
Very High
Number of Offenders in each Category
STABLE-99
Low
Mod-Low
Mod-Hi
High
STABLE-2007
Combined
STATIC and STABLE
(n) [ % ]
[%]
Low
(85) [10.7]
Moderate (109) [13.8]
High
(9) [1.1]
Low
Low
(84) [10.6]
Moderate (203) [25.6]
High
(52) [6.6]
Mod-Low
Low
(12) [1.5]
Moderate (105) [13.3]
High
(49) [6.2]
Mod-Hi
Low
(3) [0.4]
Moderate (32) [4.0]
High
(49) [6.2]
High
N = 792
278
[35.1]
224
[28.3]
157
[19.8]
84
[10.6]
Very High
49
[6.2]
Recidivism Rates for
Combined STATIC/STABLE Categories
Sexual Recidivism
Category
1 Year
2 Years
3 Years
4 Years
Low
0.7
1.1
2.0
2.0
Moderate-Low
2.2
3.7
4.3
4.3
Moderate-High
3.8
6.6
9.2
9.2
High
13.2
17.1
22.0
22.0
Very High
14.3
20.8
23.2
26.0
Overall
4.0
5.9
7.6
7.8
Recidivism Rates for
Combined STATIC/STABLE Categories
Sexual Recidivism (including sexual breaches)
Category
1 Year
2 Years
3 Years
4 Years
Low
1.5
2.2
3.6
3.6
Moderate-Low
4.1
5.5
6.2
6.2
Moderate-High
5.7
10.5
14.1
14.1
High
15.6
20.8
25.8
25.8
Very High
18.4
24.8
27.3
30.6
Overall
5.6
8.2
10.2
10.5
Recidivism Rates for
Combined STATIC/STABLE Categories
Sexual or Violent Recidivism
Category
1 Year
2 Years
3 Years
4 Years
Low
1.5
3.4
4.8
4.8
Moderate-Low
5.4
8.8
10.7
12.6
Moderate-High
7.6
13.8
20.0
20.0
High
16.7
25.6
31.9
31.9
Very High
20.4
31.2
38.7
42.0
Overall
6.6
11.2
14.5
15.3
Recidivism Rates for
Combined STATIC/STABLE Categories
“Any” Criminal Recidivism
Category
1 Year
2 Years
3 Years
4 Years
Low
3.7
6.4
9.2
9.2
Moderate-Low
8.1
11.1
14.8
14.8
Moderate-High
14.6
24.4
30.6
31.9
High
21.4
30.4
36.8
39.0
Very High
26.5
37.3
47.2
47.2
Overall
10.4
15.8
20.4
20.9
Recidivism Rates for
Combined STATIC/STABLE Categories
“Any” Criminal Recidivism (Including breaches)
Category
1 Year
2 Years
3 Years
4 Years
Low
7.3
10.9
15.1
15.1
Moderate-Low
14.4
20.8
25.2
27.3
Moderate-High
24.2
34.6
44.5
47.3
High
31.0
42.5
47.4
47.4
Very High
44.9
57.6
64.5
67.4
Overall
17.6
24.8
30.4
31.7
Combining STATIC/STABLE
and ACUTE Factors
STATIC- STABLE-2007
Combined Category
Low
Moderate-low or
Moderate-high
High or Very-high
ACUTE
SCORE
Low
Moderate
High
Low
Moderate
High
Low
Moderate
High
CURRENT
PRIORITY
Low
Low
Moderate
Low
Moderate
High
Moderate
High
High
Relative Risk for Recidivism (Risk Ratios)
within 45 Days based on Combined STATIC99, STABLE-2007 and ACUTE-2007 Scores
Sexual Crimes
Any Sexual (incl. breaches)
Violent or Sexual
Any Crime
Any Crime (incl. breaches)
Low
Moderate
High
(base rate)
(base rate)
(base rate)
1.0
2.4
14.9
( 0.58)
(1.37)
(8.61)
1.0
2.3
4.6
(2.10)
(4.83)
(9.62)
1.0
1.8
7.0
(2.12)
(3.85)
(14.77)
1.0
1.9
4.5
(3.69)
(6.96)
(16.46)
1.0
1.6
4.3
(8.36)
(13.42)
(35.82)
Treatment &
Intervention
Nothing Works?
Martinson (1974)
Large-scale study of correctional
treatment outcomes
 Could find no clear evidence that efforts
to rehabilitate offenders were “working”
 Led to considerable research into aspects
of treatment/counseling/interventions
that would lead to lower recidivism

Effective Programs
Based on meta-analytic research, Don
Andrews and his colleagues have
suggested four principles of effective
correctional interventions.
Effective Programs
RISK principle



effective programs match the level of treatment
intensity to the level of risk posed by the
offender
high risk = high intensity
mismatching can result in increased risk
Effective Programs
NEED principle



effective programs target identified
criminogenic needs
sex offenders require sex offender specific
treatment programming
other programs may result in some ancillary gain,
but risk for sexual recidivism likely will not be
reduced
Effective Programs
RESPONSIVITY principle

effective programs are those which are
responsive to offender characteristics
 cognitive
abilities
 maturity
 motivation
 mode
of intervention
 scheduling concerns
Effective Programs
PROFESSIONAL DISCRETION



in every effective correctional intervention,
there must be a coordinated plan which takes
risk, need, and responsivity into consideration
someone must be “driving the bus”
sometimes, exceptions to the first three
principles can be justified based on global
perspectives
Promising Targets
changing antisocial attitudes and feelings
 reducing antisocial peer associations
 promoting prosocial associations
 increasing self-control, self-management,
problem-solving skills
 reducing chemical dependencies
 shifting rewards for behavior from criminal to
non-criminal orientation
 develop a plan to deal with risky situations
 confront personal barriers to change

Indicators of Quality
Participation




attendance
engagement in program
completion (mature as opposed to premature
program termination)
quality relationship with service provider


respect, positive attitude
showing change on the intermediate targets
Effective Programs
The consistency of the outcome studies
accentuates the need to move beyond
simple questions as to whether treatment
works (Abracen & Looman, 2004).
 There are a number of significant
questions which have yet to be answered
with reference to sex offender treatment.
 For example, do higher risk clients receive
more treatment programs than lower risk
clients?

Assessment of In-Treatment Change
with Sexual Offenders
We need to ...
 Make
sure that the treatment targets
addressed are actually related to
recidivism
 Need to make sure that targets are actually
being addressed
Nothing Works?
One review of studies relating to the
effectiveness of treatment found that far
more studies reported positive results
(treated group with significantly lower
recidivist rates than untreated) than
inconclusive results.
 Another more recent review found that
19% of the treated offenders re-offended
during an average follow-up period of 6.85
years compared with 27% of the untreated
group.

Risk
Management
Toronto Sun Headlines







Perv Out of Jail Today: Police Consider
Publishing His New Address
Pervert a Risk: Harris
Pedophile “Terrified” to Be Out of Jail
A Community Lives in Fear
Perv’s Neighbours in Scary, Ugly Mood
Neighbours Drive Perv From Home:
Etobicoke Residents Issued Molester 24Hour Ultimatum
Neighbours: We’ve Done Our Jobs
Today’s Situation


Upon release, many sex offenders are
subject to public notification, vilification
and, sometimes, vigilantism.
As a result, some are eventually driven
out of one community into another and,
often, go “underground”.
This does not help.
Sexual Assault
is a Community Issue


The community lives in fear of sex
offenders and responses to dealing
with this fear are varied throughout
history.
At the end of the day, reduced
recidivism is everyone’s business—
offender, victim and community.
Stakeholders







victims
citizens
law enforcement
legal and correctional personnel
mental health personnel
the media
offenders
Risk Management Philosophy
Sex offending results from a complex
interaction of offender specific and
environmental factors which require
competent assessment and, ultimately,
long-term treatment and follow-up.
Risk Management

effective risk management involves the
collaboration of many different service
providers

Clinical staff, TSTs, Medical, Security, Rec
varying the mode of contact allows for
greater monitoring of activities and
attitudes
 greater contact and monitoring increases
the reliability of information leading to
case management and treatment decisions
and initiatives

Information Sharing
team work is critical
 offenders must be apprised of the limits
of confidentiality
 free flow of information is crucial between
all concerned agencies/parties




contact with case manager is facilitated by regular
treatment progress notes and additional contact
as necessary
tendency toward secrecy is minimized or
eliminated
problems can be quickly identified and managed
Holes in the System
However …
 containment
measures are often more helpful for
investigation and prosecution of breaches after the
fact
 other measures are required to increase client
accountability and to prevent further victimization
 no matter how good your Police Service is, officers
cannot be held solely responsible for the totality of
public safety
 community engagement is crucial to ensuring that
there are no more victims
Closing Thoughts
Research has clearly shown that a
collaborative approach which includes
representation from all stakeholders can
assist considerably in enhancing public
safety and offender accountability.
Working together, we can manage the
risk.
Teamwork is the key!!
Contact Information
Robin J. Wilson, Ph.D.
Clinical Director
The GEO Group, Inc.
Florida Civil Commitment Center
13613 SE Highway 70
Arcadia, FL 34266
941 806 9788
[email protected]