r. karl hanson public safety canada
DESCRIPTION
Researching Sex Offenders: A Workshop on Conceptualizing and Implementing Sex Offender Research Projects. R. Karl Hanson Public Safety Canada Presentation at the 13 th Annual Conference of the NYATSA, Saratoga Springs, NY, May 14, 2008. Big Questions. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/1.jpg)
Researching Sex Offenders: A Workshop on Conceptualizing and Implementing Sex
Offender Research Projects
R. Karl HansonPublic Safety Canada
Presentation at the 13th Annual Conference of the NYATSA, Saratoga Springs, NY, May 14, 2008
![Page 2: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/2.jpg)
Big Questions
• Impact of criminal justice and social policy interventions (e.g., community notification, residency restrictions, civil commitment, risk assessment)
• Assessment of reduced risk in high risk offenders
• Active components in sexual offender treatment
• Social policy for prevention on onset
![Page 3: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/3.jpg)
Evaluation of
Sex Offender Treatment Programs
![Page 4: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/4.jpg)
Collaborative Outcome Data Committee
• Formed in 1997
Goals:– Define standards for research on treatment
outcome for sexual offenders (develop consensus)
– Organize existing sexual offender outcome studies
– Promote high quality evaluations
![Page 5: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/5.jpg)
CODC contributors
• Anthony Beech• Darren Bisshop• Guy Bourgon• Dawn Fisher• R. Karl Hanson• Andrew Harris• Calvin Langton• Roxanne Lieb
• Janice Marques• Michael Miner• William Murphy• Michael Seto• Vernon Quinsey• David Thornton• Pamela Yates
![Page 6: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/6.jpg)
CODC Study Quality Guidelines
• Structured rating scale
• Definition of study quality – “…judgement of minimal bias can be made with high
confidence.”
• 20 items (plus 1 additional item rated for cross-institutional designs)– Items fall under 7 categories– Items assess either confidence OR bias (including direction of
bias)
![Page 7: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/7.jpg)
I) Administrative control of independent variables
• Defining treatment (confidence)
• Defining comparison group (confidence)
• Miscellaneous incidental factors (bias)
![Page 8: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/8.jpg)
II) Experimenter expectancies
• Experimenter involvement (bias)
• Blinding in data management (bias)
![Page 9: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/9.jpg)
III) Sample size
• Sample size of treatment (confidence)
• Sample size of comparison (confidence)
• Sample size of institutions (confidence) (for cross-institutional designs only)
![Page 10: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/10.jpg)
IV) Attrition
• Subject selection (bias)
• Program attrition (bias)
• Intent-to-treat (bias)
• Attrition in follow-up (bias)
![Page 11: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/11.jpg)
V) Equivalency of groups
• A priori equivalency of groups (bias)
• Adequacy of search of differences (confidence)
• Findings on group differences (bias)
![Page 12: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/12.jpg)
VI) Outcome variables
• Length of follow-up (confidence)
• Validity/reliability of recidivism information (confidence)
• Equivalency of follow-up (bias)
![Page 13: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/13.jpg)
VII) Correct comparison conducted
• Data dredging (confidence)
• Effectiveness of statistical controls (confidence)
• Computation of least bias comparison (bias)
![Page 14: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/14.jpg)
Global Rating• All items considered in making overall
judgment of bias and confidence – Same three-point scales as individual
items
• Bias and confidence are considered separately
![Page 15: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/15.jpg)
Global Rating Categories• Strong
– High confidence AND negligible bias
• Good– High confidence and some bias, OR– Some confidence and negligible bias
• Weak– Some confidence and some bias
• Reject– Little confidence, OR– Considerable bias
![Page 16: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/16.jpg)
Reliability Study 1: Students
• 2 senior undergrad students
• Approximately one week training (8 practice studies)
• 10 real studies rated independently
![Page 17: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/17.jpg)
Results: Global Ratings
• Overall 9/10 (ICC = 0.95)
• Global confidence 10/10 (ICC = 1.00)
• Global bias 9/10 (ICC = 0.69)• Direction of bias 7/10
![Page 18: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/18.jpg)
Reliability Study 2: Experts
• 12 Experts in sex offender research evaluation• No training on guidelines• 10 hypothetical studies ranging in quality • Rated 1-6 studies each (3 ratings per study)
![Page 19: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/19.jpg)
Expert Raters
• Guy Bourgon• Andrew Harris• Grant Harris• Niklas Langstrom• Roxanne Lieb• Ruth Mann
• Robert McGrath• William Murphy• Vernon Quinsey• Marnie Rice• David Thornton• Pamela Yates
![Page 20: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/20.jpg)
Expert Reliability
– Some agreement on individual items
– No agreement on global ratings
![Page 21: R. Karl Hanson Public Safety Canada](https://reader035.vdocuments.us/reader035/viewer/2022062805/56814cd7550346895db9dd0c/html5/thumbnails/21.jpg)
www.publicsafety.gc.ca
• Collaborative Data Outcome Committee. (2007). Sex offender treatment outcome research: Guidelines for Evaluation (CODC Guidelines). Part 1: Introduction and overview. Corrections User Report No 2007-02. Ottawa: Public Safety Canada.
• Collaborative Data Outcome Committee. (2007). The Collaborative Outcome Data Committee’s Guidelines for the evaluation of sexual offender treatment outcome research. Part 2: CODC Guidelines. Corrections User Report No 2007-03. Ottawa: Public Safety Canada.