Sexual Abuse: A Journal of Research and Treatment, Vol. 17, No. 4, October 2005 (
Predictors of Treatment Attrition as Indicators for
Program Improvement not Offender Shortcomings:
A Study of Sex Offender Treatment Attrition
Michelle J. Beyko
and Stephen C. P. Wong
This study classiﬁed potential attrition predictors under the domains of risk, need
and responsivity (D. Andrews & J. Bonta, 2003). Non-sexual criminogenic needs
(e.g. aggression, rule violating behaviors) and responsivity factors (e.g. lack of
motivation and denial) were the two main clusters of predictors that correctly
classiﬁed 95.3% of program completers and non-completers using discriminant
function analysis in a sample of high-risk male sexual offenders treated in an
accredited inpatient sex offender treatment program. Rapists were more aggressive
than other types of sex offenders and were more likely to drop out of treatment.
Some studies of predictors of treatment attrition have used offender problem
behaviors or psychopathologies to predict attrition and then use the information
to exclude offenders from treatment. Others have argued, and we concur, that
results of attrition research should not be used to develop an “attrition proﬁle” to
exclude offenders from treatment. Predictors of attrition should be seen as markers
for program improvement, rather than shortcomings of the offender. Suggestions
for program improvements to reduce the rate of attrition, based on results of
research, are presented.
KEY WORDS: treatment attrition; treatment dropout; risk; need; responsivity; sex offenders.
Developing strategies to reduce the incidence of sexual offenses has been an
ongoing challenge. One approach is to provide effective treatment to convicted sex
offenders to reduce their risk of sexual recidivism. A number of treatment interven-
tions, especially those that are cognitive-behaviorally based, have demonstrated
Regional Psychiatric Centre, Saskatoon, Saskatchewan, Canada.
Director of Research, Regional Psychiatric Centre, Saskatoon; Adjunct Professor, University of
To whom correspondence should be addressed; Regional Psychiatric Centre, P.O. Box 9243,
Saskatoon, Saskatchewan, Canada S7K 3X5; e-mail: email@example.com.
2005 Springer Science+Business Media, Inc.