Sex Offender Risk Assessment Assessment of Static and Dynamic Risk of Adult Sexual Offenders NJATSA 4/13/12 Rev 1/2013 Presenters: Jackson Tay Bosley, Psy.D. Clinician/Administrator Community/Parole Supervision for Life Program University of Medicine and Dentistry of NJ Merrill Main, Ph.D. Clinical Director Special Treatment Unit Department of Human Services, State of NJ Disk Contents This presentation (early version) STATIC-99 materials (scoring sheet) STABLE and ACUTE 2007 materials Dynamic Risk Factors – various articles Risk Information – various articles Adult, Polygraph, Porn, Megan’s Law, Parole, MIDSA, MnSOST-R, SIR CSL/PSL Program, Treatment Targets The Question Is he (or she) going to do it again? Yes or No (too simple) Defining Risk Probability of reoffending Actuarial Trauma associated with reoffense Probability of recovery Distaste factor Ability to manage risk Nature of offending Legal and situational realities Living conditions Risk Assessment Techniques Clinical Intuition-Professional Judgment (guess) Empirically-Informed Clinical Judgment (educated guess) Purely Actuarial (pure science) Clinically Adjusted Actuarial (science and a little common sense) How do they rate? (chance = 0.0, perfect = 1.0) Clinical judgment r= .10 Prior sex offense .19 Empirically-informed clinical judgment above .10 (?) Purely actuarial .33 to .45 (depends on instrument and sample) Clinically adjusted actuarial using structured instruments measuring dynamic factors (?) Suggestion We have an ethical and moral responsibility to use the best the science has to offer. In court, adjusted actuarial estimates, (with proper caveats) seem to work best. Know the literature and the underlying arguments. Even the best actuarials are only “moderately” predictive. History of SO Risk Assessment Early Period (till late 70’s / early 80’s) Early Research/Theories Clinical Judgment Presentation (remorse, denial, defensiveness) Plethysmography Nick Groth – Men Who Rape (1979) David Finkelhor – 4 Factors (1984) Gene Able – Large sample data (1987) Actuarial Period (mid 90s till now) RRASOR/STATIC-99 and VRAG/SORAG History of SO Risk Assessment (cont.) Men who rape: The psychology of the offender (1979) Nicholas Groth Power rapist (55% of his sample) Anger rapist (40% of his sample) Sadistic rapist (5% of his sample) Re: offenders against children Fixated Child Molesters Regressed Child Molesters History of SO Risk Assessment (cont) 4 Factors- Child sexual abuse: New theory and research (1984) David Finkelhor Emotional congruence – Does SO behavior ‘make sense’ to the offender? Sexual arousal – Arousal to children Blockages – Problems in meeting sexual needs in an adult manner Disinhibition – How does the offender give himself permission to offend? History of SO Risk Assessment (cont.) Large sample study of offenders in treatment (1987) Gene G Abel N = 561 offenders in TN and NY Obtained “certificate of confidentiality” Asked for self-report of other offense behaviors without fearing legal consequence Findings: Crossover behavior Small portion of the sample offended a lot History of SO Risk Assessment (cont.) Science in the field Plethysmograph studies Primacy of Canadian researchers Hanson & Bussiere – Predictors of SO recidivism: A meta-analysis 1994 RRASOR - 1997 Quinsey, Harris, Rice & Courmier – VRAG 1998 Meta-analysis What do the numbers say and where have they led us? Factor Analysis of SO Recidivism Risk Sexual Deviance Sexual preference for pre-pubertal sexual partners (pedophilia) Sexual preference for cues of pain/fear Strength of sexual urges (hypersexuality) Antisociality Enjoyment of illegal activity Impulsivity Criminal value system Actuarials/Risk Assessment Instruments Rapid Risk Assessment for Sexual Offense Recidivism - RRASOR (1997) Minnesota Sex Offender Screening Tool – Revised - MnSOST-R (1998) STATIC-99R (STATIC-2002R is in development) Sex Offender Needs Assessment Rating - SONAR (Stable and Acute-2007) Others Sexual Reoffense CARAT, SORAG, JSORRAT-II, VASOR, Risk for Sexual Violence Protocol (RSVP), Risk Matrix-2000 Violent Reoffense VRAG, LSI-R, PCL-R, SAVRY, SIR Static-99R Age Lived with a lover for at least 2 years Index or prior non-sexual violence Prior sex offenses (charges or convictions) Prior sentencing dates Non-contact sexual offenses Unrelated/stranger or male victims Other Instruments (with some brief overviews) MnSOST-R (now, MnSOST-3) Violence Risk Appraisal Guide (VRAG) Risk for Sexual Violence Protocol (RSVP) Sex Offender Treatment Rating Scale Sex Offender Needs Assessment Rating Risk Matrix-2000 Treatment Progress Scale MnSOST-R Number of sex/sex-related convictions Length of sexual offending history Was the offender under any form of supervision during sex offense Was any sex offense in public Was force or the threat of force ever used Multiple acts on a single victim Number of different age groups victimized (<6, 7-12, 13-15, 16+) Offended against a 13 to 15 year old and offender >5 years older Victim a stranger in any sex offense Evidence of adolescent antisocial behavior Substantial history of drug or alcohol abuse Employment history Discipline history while incarcerated Chemical dependency treatment while incarcerated Sex offender treatment history while incarcerated Age of offender at time of release VRAG items Psychopathy Checklist Score Elementary school maladjustment Age at index offense DSM III personality disorder Separation from parents before age 16 Failure on prior conditional release History of nonviolent offenses Never married DSM III schizophrenia Victim injury in index offense History of alcohol abuse Male victim in index offense RSVP 5 SECTIONS, 22 ITEMS Sexual Violence History (5 items) Psychological Adjustment (5 items) Mental Disorder (5 items) Social Adjustment (4 items) Manageability (3 items) SOTRS Sex Offender Treatment Rating Scale Insight Deviant thoughts Awareness of situational risks Motivation Victim empathy Offense disclosure SONAR Sex Offender Need Assessment Rating STABLE FACTORS Intimacy deficits Negative social influences Attitudes tolerant of sexual offending Sexual self-regulation General self-regulation ACUTE FACTORS Substance abuse Negative mood Anger Victim access Risk Matrix-2000 Three scales: Sexual, Violent, Combined Age 18-24 = 2 points; 25-34 = 1 point; Older = 0 points Sexual Appearances (Convictions) Criminal Appearances Male Victim of Sex Offense Stranger Victim of Sex Offense Single (Never in Marital Type Relationship) Non- Contact Sex Offence Burglary Convictions TREATMENT PROGRESS SCALE -Offense behavior admission -Sexual interests -Sexual behavior -Criminal attitudes -Substance abuse -Mental health stability -Impulsivity -Cooperation with treatment -Employment -Finances -Social influences -Responsibility Acceptance -Sexual attitudes -Sexual risk management -Criminal behavior -Emotion management -Problem solving -Stage of change -Supervision Cooperation -Residence -Adult love relationship -Social involvement Actuarial Item Analysis Actuarial items are chosen based on their empirical link to recidivism (atheoretical) Static Factors - Fixed, easy to code, most researched - priors, age/gender of victim Dynamic Factors – Changeable, harder to code (constructs: empathy, stability, psychopathy), what we target in treatment – indicators of imminent sexual recidivism Static vs Dynamic (?) Psychologically-meaningful risk factors Vulnerabilities Psychological Predispositions Propensities Static factors (events) are proxies for underlying risk-relevant propensities Mann, Hanson & Thornton, 2010 Static Risk Items Age (youth vs old age) Prior criminal behavior Non-sexual crimes Sexual crimes Number of sentencing occasions Supervision violations Sexual offense victim choice Unrelated, Stranger, Male Static Risk Items Relationship history Treatment history (cont.) Completion Failure/termination Substance abuse history Adverse childhood environment Psychological factors/Dx ASPD, psychosis, DD/MR, pedophilia, psychopathy (arguably dynamic factors) Factors associated with recidivism (Hanson & Morton-Bourgon, 2004): Specific interest in boys measured by ppg Deviant sexual preference dx of any paraphilia Sexual Preoccupations Emotional identification with children r = .30 .40 .51 .63 Problems with Research on Static Factors They do not change in response to treatment They are proxies for underlying psychological propensities The data might be clear, but the interpretation might vary Not suitable as targets for intervention Why We Use Static Factors to Predict Recidivism Most research evidence Data gathered as part of background or Intake information Easy to code from data Clear(er) – usually yes or no (or a number) Easier to locate No other factors are as predictively powerful (yet) What are Dynamic Risk Factors Dynamic = changeable Psychological propensities Traits Habits (??) Amenable to treatment intervention Observable, but more difficult to measure Important for supervision purposes Dynamic Risk Items (from Stable-2007) Stable Significant social influences Capacity for relationship stability Emotional identification with children (<13) Hostility toward women General social rejection Lack of concern for others Poor problem-solving skills Dynamic Risk Items Stable (cont.) Negative emotionality Sex Drive/Preoccupation Sex as coping Impulsive Deviant sexual preference Cooperation with supervision (from Stable-2007) Dynamic Risk Items Acute Victim access Hostility Sexual preoccupation Rejection of supervision Emotional collapse Collapse of social supports Substance abuse (from Acute-2007) How predictive validity of the STATIC-99R is affected by the addition of dynamic risk information from Hanson and Helmus, 2009 STATIC and Stable-2007 from Hanson and Helmus, 2009 Static Score 3 year recidivism % Stable-2007 =5 (Mean) =14 (90%) 2 3% 7% 5 7% 18% 7 14% 32% STATIC and Sample Choice from Hanson and Helmus, 2009 Static Score Sample: 2 5 7 5 year recidivism % Routine Risk/Need 5% 12% 11% 25% 19% 38% (Witt & Conroy, 2008) 39 Importance of Static and Dynamic Risk Factors (Hanson, 2002) Context Static Dynamic – Stable Dynamic – Acute ++ + + ++ + Long term sanctions 1. Imposition 2. Release 40 Importance of Static and Dynamic Risk Factors (Hanson, 2002) Context Static Dynamic – Stable Dynamic – Acute ++ + ++ + + ++ Community supervision 1. Placement 2. Revocation /change 41 Other Potential Risk Indicators Plethysmograph results Polygraph results Negative treatment response Inability to manage adult relationships Hostility towards women Callousness/lack of empathy Denial Denial as a Risk Factor Early thinking – Denial means much higher risk Hanson & Bussiere (1997) no effect Nunes et al (2007) higher risk for low-risk offenders Looman (2011) higher risk for higher-risk offenders Denial as a Risk Factor (cont.) Denial is predictive of reoffense for some offenders. Denial at the beginning of treatment is not indicative of higher risk. Denial should not preclude anyone from treatment. Denial is a legitimate treatment target. Denial at the end of treatment might indicate higher risk. Percentage rates of sex offender recidivism (Harris & Hanson, 2004): Type All Rapists Incest Female target CM Male target CM w/out prior offense w/ prior offense Offender over 50yrs 5yr 10yr 14 20 14 21 6 9 9 13 23 28 10 15 25 32 7 11 k=95 15yr 24 24 13 16 35 19 37 13 n=31000 Criminal Recidivism Rates US Dept. of Justice (Bureau of Justice Statistics, 2002) 3 year follow-up Burglary Larceny Auto Theft DUI Sexual Offenses 74% 75% 70% 51% 5.3% n=9691 Supported Risk Factors Mann, Hanson & Thornton, 2010 Sexual preoccupation Deviant sexual interest (children or violence) Offense-supportive attitudes Emotional congruence with children Lack of emotionally intimate relationships with adults Supported Risk Factors Mann, Hanson & Thornton, 2010 Lifestyle impulsivity General self-regulation problems Poor cognitive problem-solving Resistance to rules and supervision Grievance/hostility Negative social influences Promising Risk Factors Mann, Hanson & Thornton, 2010 Hostility towards women Machiavellianism Callousness/lack of concern for others Dysfunctional coping Sexualized coping Externalizing Unsupported Risk Factors (But with some interesting exceptions) Mann, Hanson & Thornton, 2010 Denial Poor self-esteem Major mental illness Except for low risk offenders, high risk offenders, incest offenders, rapists (??) Increases risk of general & violent recid. Loneliness Not Risk Factors Mann, Hanson & Thornton, 2010 Depression Poor social skills Depressed offenders less likely to reoffend Still worthy of treatment I.e., poor dating skills Poor victim empathy Lack of motivation for treatment Assessed pre-treatment Treatment Effect on Recidivism Risk Treatment has a positive effect for reducing recidivism (40% reduction). Failing treatment is a very poor prognosticator (200% increase in recid.). Treatment “works” for those who are invested in the process and “work the program”. Treatment has less effect for those who attend, but are not invested. Juvenile Sexual Offender Risk Assessment Instruments J-SOAP II (Juvenile Sex Offender Assessment Protocol) JSORRAT-II (Juvenile Sexual Offense Recidivism Risk Assessment Tool-II) ERASOR (Estimate of Risk of Adolescent Sexual Offender Recidivism) PFS (Protective Factors Scale) Juvenile Sexual Offender Risk Assessment Instruments (cont.) J-RAT (Juvenile Risk Assessment Tool) SAVRY (Structured Assessment of Violence Risk in Youth) CANS-SD (Child and Adolescent Needs and Strengths-Sexual Development) YLS/CMI (Youth Level of Service/Case Management Inventory) Megan’s Law Tiering Registrant Risk Assessment Scale (adults) Juvenile Risk Assessment Scale (juveniles) These are not purely risk (recidivism) scales, but factor in seriousness of the offense if reoffense occurs. Information is based on all credible information (not just adjudicated offenses). These scales have shown moderate concurrent validity with actuarials. Take Home Message: Baseline risk is best determined by an analysis of the static factors using a structured actuarial assessment tool. Additional analysis of the dynamic factors will determine ongoing risk of reoffense, imminence of reoffense and targets of treatment interventions. Dynamic factors will dictate appropriate legal supervision conditions. Take Home Message: (Cont.) Sexual offender risk assessment is complicated, but there is building scientific consensus on the major issues of interest. Deviant sexual interest is vital to assess. Antisociality (willingness to break society’s rules) is also vital to assess. Keeping up with the science is interesting. Contact Info Jackson Tay Bosley, Psy.D. Clinician Administrator UMDNJ CSL/PSL Program Whittlesey Road, P.O. Box 863 Trenton, NJ 08625 (609) 341-3093 bosleyjt@umdnj.edu Contact Info Merrill Main, Ph.D. Clinical Director DHS Special Treatment Unit PO Box 905 Avenel, NJ. 07001 732-574-2250 x8601 Merrill.Main@dhs.state.nj.us