Karl Hanson, Ph.D., Douglas Epperson, Ph.D., Dennis Doren, Ph.D., Thomas
Grisso, Ph.D. David Thornton, Ph.D., Robert Prentky, Ph.D., Robert
Hare, Ph.D. and Vernon Quinsey, Ph.D.
13 hours of video with documentation, $495.00
Review By David S. Prescott, LICSW
A passing glance at the research into predicting sex offender recidivism
shows that clinicians have fared poorly at assessing risk. This insulting
revelation has spurred controversy and self-examination. The only solace
is that other types of unguided prediction, such as academic performance,
have not done any better.
Actuarial scales that give (on average) more effective predictions
than clinicians have offered dubious comfort to evaluators accustomed
to comprehensive evaluations. Skeptics ask if less really does equal
more. Whatever the case, there is a strong implication that evaluators
of risk have an obligation to be familiar with the recent research
around these scales.
These scales have an allure that can diminish to confusion, disappointment,
and frustration. Many users hope that they will be able to predict
absolute risk for a specific offender. Questions arise as to whether
the scales apply to groups such as sadists, adolescents, or the developmentally
delayed. As simple as these scales appear at the outset, knowledge
of the research behind them is vital to using them in any meaningful
Lloyd Sinclair's video program is compiled from his three-day symposium
featuring the authors of the most promising scales, including Karl
Hanson, Douglas Epperson, David Thornton, Vern Quinsey and Robert Prentky,
as well as Robert Hare, Ph.D., author of the Hare Psychopathy Checklist – Revised
(PCL-R). The program covers virtually every element of sex offender
risk assessment, along with presentations on related legal and ethical
issues, among other areas. Each author presents their scales, their
underlying research, and (except for Hare) scores typical offenders
portrayed by actors. Hare presents an overview on psychopathy and discusses
how psychopathic offenders differ from other sex offenders.
Most of the actuarial scales discussed in this program are described
in ATSA’s recent pamphlet Risk Assessment, and need no further
introduction. They include the RRASOR, Static-99, MnSOST-R, SONAR,
VRAG, and SORAG. In addition, Prentky presents the Juvenile Sex Offender
Assessment Protocol (J-SOAP) and Thornton presents his Structured Risk
Assessment (SRA). The SRA is an approach for measuring dynamic (changeable)
risk factors. It also contains a brief measure of treatment progress
and supplements Static-99. It is noteworthy that the SRA is currently
available only in this program and in specialized training with Thornton.
The Sinclair program leads the viewer from Hanson and Bussiere’s
1996 meta-analysis through the development of the static factor scales
and into dynamic risk assessment with the SONAR and SRA. Key elements
such as ROC curves are addressed along the way. Aspects such as testifying
in court and writing reports are also covered. Tom Grisso speaks to
ethical issues around these instruments and risk assessment in general.
Dennis Doren ties all the threads together in a thoughtful comparison
of the instruments and suggestions for their use. He makes the point
that many of the instruments can be used in combination to assess various
dimensions of offending, such as pedophilia and antisociality. The
program concludes with a panel discussion of the authors (moderated
by Anna Salter) addressing audience questions ranging from female offenders
to the future of risk prediction.
A unique element of this program is that nearly all the presenters
were present for the entire conference. Interesting discussions travel
across presentations, including the role of dynamic factors in understanding
long-term risk. These discussions also highlight areas yet to be addressed
in understanding the prediction of recidivism, such as whether scale
items are best understood as theory or actuarial framework.
The program places a deliberate focus on the potential use of scales
for civil commitment evaluations. Although this ensures validation
and methodological rigor in the scales presented, it also means that
other approaches are not included. Empirically grounded (but not directly
validated) schemes such as the Sexual Violence Recidivism – 20
are not cited.
Sinclair and his colleagues have compacted, but never reduced, vast
information into a program accessible to beginning and advanced skills
alike. It is comprehensive, versatile, and contains valuable information
for treatment planning and decisions beyond formalized risk assessment.
It also provides a first-class window into the fascinating thought
behind the research into re-offense risk prediction.