Handbook of Recidivism Risk / Needs Assessment Tools

JAY P. SINGH, PhD, PhD, is the Founder of the Global Institute of Forensic Research and is affiliated with the Department of Psychiatry and the Wharton School of Business at the University of Pennsylvania. DARYL G. KRONER, PhD, is a Professor at Southern Illinois University and reviewed the effectiveness of offender programs at the Illinois Department of Corrections. J. STEPHEN WORMITH, PhD, is a Professor of Psychology at the University of Saskatchewan, and Director of the Centre for Forensic Behavioural Science and Justice Studies. SARAH L. DESMARAIS, PhD, is an Associate Professor and Coordinator of the Applied Social and Community Psychology Program at North Carolina State University. ZACHARY HAMILTON, PhD, is an Associate Professor of Criminal Justice and Criminology and the Director of the Washington State Institute for Criminal Justice at Washington State University.

• Notes on Contributors xvPreface: Recidivism Risk Assessment in the 21st Century xxiiiJohn MonahanPart I Introduction 11 Performance of Recidivism Risk Assessment Instruments in U.S. Correctional Settings 3Sarah L. Desmarais, Kiersten L. Johnson, and Jay P. SinghCharacteristics of Risk Assessment Instruments 4Characteristics of Samples and Studies 5The Current Review 5Method 6Review Protocol 6Search Strategy 6Results 10Characteristics and Content of Instruments 10Sample and Study Characteristics 12Performance of Recidivism Risk Assessment Instruments 14Discussion 19Limitations 21Conclusions 21Author Note 22References 22Supplemental Table 29Part II Risk/Needs Assessment in North America 312 The CAIS/JAIS Approach to Assessment 33Christopher BairdIntroduction 33History of Development 33Development and Validation of Risk Assessment 35Development of CAIS/JAIS Needs Assessment Instruments 38Development, Synopsis, and Evaluation of CMC/SJS 39Synopsis of CMC Supervision Strategies 41CMC Evaluation Results 44Current Issues and Future Directions 45References 463 Correctional Offender Management Profiles for Alternative Sanctions (COMPAS) 49Tim Brennan and William DieterichIntroduction 49Purposes of COMPAS 49Predictive Risk Assessment 49Explanatory and Needs Assessment 50Avoiding the Various Disconnects between Assessment, Treatment, and Outcomes 50Different versions for Different Organizational Purposes and Target Populations 50Methodological Goals 50Background and History of Development 50Incorporate Theoretical Guidance in Selecting Relevant Scales to Achieve a Comprehensive Coverage of Explanatory Factors 50Incorporate Empirical Guidance from Meta-analytic Research on Predictive Factors 51Incorporate the Strength/Resiliency Perspective 51Incorporate Gender-Responsive Risk and Need Factors 51Dynamic vs. Static Items in COMPAS Risk Assessments 51Built-In Data Validity Tests and Error-Checking 51Flexible Access to External “In-Depth” Diagnostic Assessments 51Advanced Analytical Methods for Risk Prediction 52Multiple Dependent Criterion Variables 52Systematic Re-validation and Calibration to Accommodate Offender Population Changes 52Ease of Use and Efficiency 52Scalability—Agency Control Over Data Needs and Staff Workloads for Specific Processing Decisions 53Integrated Database and Seamless Links from Risk/Needs Assessment to Case Management, Treatment Goals, Treatments Provided, and Outcomes 53Generalizability of Predictive Accuracy Across Regions, Agencies, and Racial/Gender Categories 53Versions of COMPAS 53Youth COMPAS 53Reentry COMPAS 54Women’s COMPAS 54Norming or Standardization Samples 54Settings for COMPAS Assessments (Pretrial, Community Corrections, Prison/Jail) 55Web-based Implementation of Assessments and Subsequent Treatment and Outcome Data 55Predictive and Needs/Explanatory Scales 55Additional Risk Models for Other Criminal Justice Criterion Outcomes 56Measures of Predictive Accuracy 56Independent Validation Studies by External Researchers, from Diverse Geographical Areas 56Follow-up Data and Time Frames 56Internal Classification for Jails/Prisons for Offender Management and Responsivity 56Theoretical Foundations of COMPAS 57Social Learning Theory 57The General Theory of Crime 57Strain Theory/Social Marginalization 57Routine Activities Theory 57Social Disorganization/Sub-cultural Theories 58Control/Restraint Theories 58Data Collection and Interviewing 58Official Records/File Review 58Interview Section 58Self-Report Paper and Pencil Section 59Scoring Guidelines 59Administrator Qualifications and Training Modules 59Basic 2-day Training 59Software Training 59Train-the-Trainer Workshops 60Anomaly or Error Detection 60Coding Integrity Checks 60Reliability Research 60Internal Consistency Reliability 60Test-Retest Reliability 61Inter-rater Reliability 61On-Screen Hyperlinks to Give Staff Immediate Access to Question Definitions 62Systematic Identification, Removal, or Revision of Problematic Questions 62Using Machine Learning (ML) Classifiers to Achieve High Reliability of Prison Classification Assignments 62Latent Factor Structure of the Overall COMPAS Assessment Domain 621. Early Starter, High Violence Risks with Very Low Social Capital 632. Extent of Criminal Involvement 633. Social Marginalization 634. Violence 635. Antisocial Personality and Attitudes 636. Socialization and Social Learning in Antisocial Environments 637. Transience, Unstable Residence, Drugs, and Poor Social Adjustment 638. Antisocial High-Risk Lifestyle 63Criterion Validity: Predictive and Concurrent Validity 64Predictive Validity of COMPAS Has Been Replicated in Multiple Jurisdictions, Multiple Agencies, and by Different Research Teams 64Predictive Validity Across Gender and Racial Groups (Criterion Related Validity) 64Content Validity—Coverage of Relevant Factors 66Meta-analytic Selection 67Theory-based Selection 67Predictive Validity of Specific COMPAS Subscales 67Calibration 68Discrimination and Dispersion 68Construct Validity—Factorial, Concurrent, and Criterion Validity 69Factorial Validity of COMPAS Subscales 69Construct and Convergent Validity 69Substance Abuse 70COMPAS Subscale Correlates with Official Criminal Behaviors and Age of Onset 70Criminal Attitudes and Criminal Personality Scales 70Implementation Research 70User Satisfaction 70Impact Evaluation and Cost/Benefit Analyses of COMPAS 71Current Issues and Future Directions 711. Data Analytic Advances 712. Advances in Automated Data Collection Procedures 723. Diverse Pathways to Crime and New Internal Classifications for Prisons 724. Advances in Decision Analytic Methods for Setting Cut Points in Risk Assessment Scales 72References 724 The Federal Post-Conviction Risk Assessment Instrument: A Tool for Predicting Recidivism for Offenders on Federal Supervision 77Thomas H. Cohen, Christopher T. Lowenkamp, and Charles RobinsonIntroduction 77Overview of Risk Assessment Tools 79History of Actuarial Instruments in the Federal Probation System 79Development and Implementation of the Federal Post Conviction Risk Assessment Instrument 81Administration and Scoring of the PCRA 82Officer Section of the PCRA 82Handling Scoring Omissions and Missing Data 83Offender Section of the PCRA 84Highlighting Addressable Criminogenic Needs 84The Role of Supervision Overrides in the PCRA’s Risk Assessment Mechanism 85The PCRA Assessment and Reassessment Policy 85Overview of Research Using the PCRA 86Revalidation of the PCRA’s Predictive Validity 86Investigating the PCRA’s Dynamic Characteristics 88Studies Investigating the PCRA for Race/Gender Bias 92Evaluating the Predictive Value of the PICTS Criminal Thinking Styles 92The PCRA and Special Offender Populations 93Studies Using the PCRA to Investigate Supervision Overrides, Implementation of the Low-Risk Policy, the Presence of Responsivity Issues, and Convergence of Long-Term Recidivism Rates Across Risk Levels 94PCRA 2.0 and the Future of the PCRA 96Conclusion 97References 975 The Inventory of Offender Risk, Needs, and Strengths (IORNS) 101Holly A. MillerRationale and Development of the IORNS 102Standardization 106Reliability 106Validity 107Construct Validity 107Internal Structure 109Predictive Validity 109Tracking Change through Treatment 110Case Illustration 111Utility and Future Directions of the IORNS 114References 1146 The Level of Service (LS) Instruments 117J. Stephen Wormith and James BontaThe Origins and Evolution of the Level of Service (LS) Instruments 118The Theoretical Underpinnings of the LS 119Administration 121Reliability 124Validity 126Current Issues, Concerns, and Future Directions 132Summary and Conclusions 135References 1397 The Ohio Risk Assessment System 147Edward J. Latessa, Brian Lovins, and Jennifer LuxOverview 147Theoretical Framework 150The Creation and Validation of the ORAS 150Administering the ORAS 151Training Requirements and Qualifications 152The Validity of the Ohio Risk Assessment System 153Ohio Validation 153ORAS Pretrial Assessment Tool 153ORAS Community Supervision Tool 153Indiana Validation 154Texas Validation 155ORAS Prison Intake Tool 156Ohio Risk Assessment System-Reentry and Supplemental Reentry Tools 157The Reliability of the Ohio Risk Assessment System 158Implementation Research 159Use of the ORAS 160Benefits of ORAS 160Current Issues and Future Directions 161References 1628 Self-Appraisal Questionnaire (SAQ): A Tool for Assessing Violent and Non-Violent Recidivism 165Wagdy LozaDevelopment of the SAQ 165Description of the SAQ Scales 166Uses of the SAQ 166Who Can Administer the SAQ? 166Administering the SAQ 167Scoring the SAQ 167Interpreting the Score of the SAQ 168Example of Reporting the Results of the SAQ 168Training 168Limitations of the SAQ 168Studies Demonstrating the Psychometric Properties of the SAQ 169Reliability of the SAQ 169Subscale/Total Correlations 169Construct Validity 170Concurrent Validity 170Predictive Validity 171The Validity of the SAQ for Use with Female Offenders 173SAQ as a Classification Measure 173Comparison Studies with Other Instruments 173The SAQ and Deception 173Cultural Diversity Comparisons 174Predictive Validity of the SAQ for Use with Mentally Ill Offenders 175The Validity of the SAQ for Use with Young Offenders 175Postgraduate Research Projects 175Research Agenda 175Conclusion 175Case Study: High Risk 176Reason for Referral: Intake Assessment 176References 1779 Service Planning Instrument (SPIn) 181Natalie J. Jones and David RobinsonOverview 181History and Development 182Theoretical Foundations 184Content and Measurement 185Measurement of Strengths 186Administration and Training 187Psychometric Properties 187Research Samples 187Reliability Research 188Predictive Validity Research 189Discussion 194References 19610 The Static Risk Offender Needs Guide – Revised (STRONG-R) 199Zachary Hamilton, Xiaohan Mei, and Douglas RouthOverview 199Recidivism Prediction Spotlight 200Item Types 202Instrument Purpose(s) 203Target Population and Setting 204Time Frame of Prediction 204Theoretical Framework 204History of Development 205The Static Risk Assessment 205The Offender Needs Assessment 206The STRONG-R 206Subscale Construction 206Methodological Advances 207Customization and the STRONG-R 207Criminogenic Need Domains 208Administration and Training 209Training at Implementation 209Ongoing Training for Quality Assurance 211Reliability and Validity Research 212Latent Structure 212Content Validity 214Internal Consistency 215Inter-rater Reliability 215Convergent/Divergent Validity 218Concurrent Validity 219Predictive Validity 220Implementation of the STRONG-R 221Current Issues and Future Directions 222Software Applications 222Responsivity, Case Planning, and Management 223Methods of Customization 224References 225Part III Risk/Needs Assessment Abroad 22911 Offender Group Reconviction Scale 231Philip HowardOverview 231History of Development, and Key Differences between the Versions 231The Copas Rate, and Unusual Changes in OGRS Score 232The Introduction of Offence-free Time and Violence Prediction in OGRS4 233Summary of Risk Factors Scored in Each Version 234Operationalizing Recidivism and the Follow-up Period 234Theoretical Framework 236Administration 236Reliability and Validity Research 237Implementation in Correctional Practice 238Current Issues and Future Directions 239References 24012 Forensic Operationalized Therapy/Risk Evaluation System (FOTRES) 243Leonel C. Gonçalves, Astrid Rossegger, and Jérôme EndrassOverview 243History of Development 243Description of FOTRES and its Subscales 243Review of FOTRES Versions 246Theoretical Framework 247Scope of Application 247Risk Characteristics 247Target Offense 247Offense Mechanism Hypothesis 247Administration 248Validity and Reliability Research 249Implementation Research 250Case Studies 251Current Issues and Future Directions 251References 25213 The RisCanvi: A New Tool for Assessing Risk for Violence in Prison and Recidivism 255Antonio Andrés-Pueyo, Karin Arbach-Lucioni, and Santiago RedondoIntroduction 255A Brief Overview of the Catalan Correctional System 255The RisCanvi: A New Tool for Risk Assessment of Prison Recidivism 256The RisCanvi: Structure and Some Psychometric Properties 258Using the RisCanvi in Practice 264Conclusions 266Acknowledgments 267References 267Part IV Conclusion 26914 Risk Assessment: Where Do We Go From Here? 271Faye S. TaxmanWhat Are Major Issues Affecting the Development and Evaluation of Risk Assessment Tools? 272Do Risk Assessment Tools Accurately Identify the Probability of Recidivism (Predictive Validity)? 273Do the Risk Assessment Instruments Differentiate Between Those in the Justice System and Those That Are not in the Justice System? 275Are the Methods to Score Items and Categorize Risk Appropriate? 275Are Instruments Neutral on Race, Gender, and Other Key Demographic Issues? 277Do the Domains on the Risk Assessment Instruments Have Construct and Content Validity? 278How to Implement Risk Assessment Instruments so They Are Useful in Practice Including Case Planning, Resource Allocation, Treatment Referral or Placement, and Clinical Progress? 280Conclusion 281References 282Appendix 285Index 305