FORUM on Corrections Research

Academic Contributions

Sex Offender Research

Predicting sex offender program attrition: The role of denial, motivation and treatment readiness¹

Mark Latendresse²
Department of Psychology, Carleton University

This study examined denial, minimization, motivation and treatment readiness in relation to sex offender treatment outcome in a heterogeneous sample of 449 sex offenders.

Treatment non-completion or drop-out is common among sex offenders. It is therefore important to determine which offenders would benefit from pre-treatment sessions aimed at reducing or removing those factors that prevent the offender from fully benefiting from treatment.

Key findings from this study demonstrate that denial, minimization and possibly treatment readiness are predictors of treatment non-completion among sex offenders. These responsivity factors seemingly affect whether or not an offender completes treatment, even after taking risk into account. It may therefore be worthwhile to assess denial, minimization and treatment readiness before assigning sex offenders to a specific treatment program.

Background

As a result of public concern in Canada, there has been widespread implementation of sex offender treatment programs aimed at reducing sexual reoffending (Polizzi, MacKenzie & Hickman, 1999). Regardless of ongoing debate concerning the efficacy of sex offender treatment, recent research demonstrates that sex offenders treated with current cognitive behavioural approaches will benefit from treatment, and that this treatment will result in modestly reduced recidivism rates for this population (Abracen & Looman, 2004).

. . . this study examined the relationship between treatment readiness, motivation, minimization and denial status ... and determined if these variables predict sex offender treatment outcome.

In fact, the contentious question of treatment efficacy with sex offenders has overshadowed and prevented the advancement of more detailed research on factors affecting treatment outcome, such as treatment readiness, motivation and denial. The responsivity principle is an important concept directly related to increasing behavioural change, and warrants greater systematic investigation. Treatment readiness, motivation, denial and minimization are specific responsivity factors that have started to be investigated in the correctional literature as potential mediators or predictors of treatment outcome.

As mentioned, treatment non-completion or dropout is common among sex offenders. Attrition rates are highly variable across programs, offender types, and institutions, rendering it difficult to yield a precise estimate. Attrition rates in sex offender treatment programs range from 20% to 58% in institutional settings (Geer, Becker, Gray & Krauss, 2001) and from 17% to 47% in community-based programs (Craissati & Beech, 2001). Research suggests that attending to responsivity factors during the planning and delivery phases of treatment increases the probability that an offender will complete the intervention, which in turn will reduce the likelihood of reoffending (Dowden & Serin, 2001).

Present study

The purpose of this study was to further investigate the relationship between treatment readiness, motivation, denial, minimization and treatment outcome (i.e., attrition status) in a heterogeneous sample of incarcerated sex offenders. Specifically, this study examined the relationship between treatment readiness, motivation, minimization and denial status (specific responsivity variables) and determined if these variables predict sex offender treatment outcome.

Methodology

Participants in the present study consisted of adult male sex offenders under the responsibility of Correctional Service Canada (CSC) who were assessed at Millhaven Institution. The sex offender sample included rapists, extra-familial child molesters, intra-familial child molesters, and mixed sex offenders (adult and child victims).

This study divided sex offender treatment participants into three groups: 1) treatment completers, 2) treatment drop-outs, i.e., those who did not complete the program due to either offender-initiated drop-out or agency expulsion from treatment, and 3) treatment non-completers, i.e., those who did not complete the program for either personal or administrative reasons (e.g., transferred, program cancelled, offender released).

Participants ranged in age from 19 to 76 at time of admission to treatment, with a mean age of 43.59 years [standard deviation (SD) = 12.42]. The ethnic composition of the total sample was as follows: Caucasian - 79.2%, Black - 9.2%, Aboriginal - 7.1%, and other - 4.5%. In terms of marital status, most offenders (42.2%) were single, 17.6% were married, 14.5% were in a common-law relationship, 12.5% were separated, 11.4% were divorced, and 1.8% were other. The average sentence length for the sample was 4.21 years (SD = 3.38).

The denial and minimization checklist (DMCL) (Barbaree, 1991), treatment readiness scale (Serin, Kennedy & Mailloux, 2002) and motivation index³ were used to measure the four responsivity variables. The Static-99 was used to control for offender risk level (Hanson & Thornton, 1999).

Results

This study involved 448 male sex offenders admitted to Millhaven Institution for assessment between December 1999 and September 2005. The attrition rate for these offenders was 11.2% during their current sentence; that is, of the 448 offenders, 11.2% (50) dropped out or did not complete sex offender treatment they had begun. Specifically, 398 completed the sex offender program, 17 of the participants were classified as drop-outs, and 33 failed to complete the treatment program for administrative or personal reasons.

The descriptive statistics and correlations for the psychometric tests used in this study are presented in Tables 1 and 2, respectively.

Overall, there were no significant differences between the treatment completers, drop-outs and non-completers in terms of their average minimization total scores [F (2, 445) = 0.05, p < .95]. Similarly, there was no significant difference between groups on their average motivation index scores [F (2, 445) = 1.55, p < .22]. Both minimization and motivation were similar across each of the three treatment outcome groups. Please note that the denial component of the DMCL is measured as yes/no and could not be compared across groups using ANOVA.

Conversely, there were statistically significant differences across treatment outcome groups on the Static-99 [F (2, 445) = 4.11, p < .02] and the treatment readiness scale [F (2, 445) = 3.61, p < .03]. The drop- out group scored significantly higher on the Static-99 (indicating higher risk) than both the completer and non-completer groups. On the treatment readiness scale (TRS), average total scores were highest for the completer group, followed by the non-completer group, and lowest for the drop-out group. Note that on the TRS, higher scores indicate being more ready for treatment.

Table 1

Descriptive Statistics for Psychometric Measures by Treatment Outcome Group

Variable	Total N = 448		Completers n = 398		Dropouts n = 17		Non-Completers n = 33
	Ma	SDb	M	SD	M	SD	M	SD
Minimization total	2.87	2.49	2.86	2.40	3.06	3.04	2.88	2.89
Motivation level	2.05	0.43	2.06	0.43	1.88	0.33	2.00	0.50
Static-99*	3.21	2.22	3.16	2.21	4.71	1.57	3.06	2.38
TRS total*	29.86	14.30	30.49	14.14	23.35	13.56	25.67	15.46

. . . higher levels of risk were associated with lower levels of motivation.

Table 2 presents the correlations for the total sample between the DMCL total minimization score, motivation index, Static-99, and the TRS total scores. Higher levels of treatment readiness were associated with higher levels of motivation (as measured by the motivation index), denial, higher levels of minimization, and treatment outcome (i.e., program completion). The Static-99 was negatively associated with motivation level; in other words, higher levels of risk were associated with lower levels of motivation. Denial was negatively associated with minimization and treatment readiness; in other words, denial was associated with lower levels of minimization and lower levels of treatment readiness.

Table 2 

Correlation Coefficients for the Treatment Readiness Scale and Other Responsivity Factors

	1	2	3	4	5	6
1. aTRS total	1
2. aStatic-99	-.09	1
3. aDenial	-.37**	-.01	1
4. aMinimization total	.20**	-.03	-.52**	1
5. aMotivation level	.44**	-.24**	-.24	.09	1
6. aProgram outcome	-.12**	.07	.14**	.01	-.07	1

. . . denial was associated with lower levels of minimization and lower levels of treatment readiness.

A sequential logistic regression was conducted with treatment attrition status (i.e., treatment completed or not completed) as the outcome variable. Logistic regression permits the prediction of a discrete outcome, such as treatment completion/non-completion, from a set of predictor variables that may be continuous, discrete, dichotomous, or a combination of these. In this study, treatment attrition was coded as a discrete outcome. Due to the small number of drop-outs (17), they were combined with the non-completers so that we had two treatment outcomes for the regression analysis: treatment completed and treatment not completed. As predictor variables - that is, variables used to try to predict treatment outcome - age, denial, minimization, motivation, risk (static-99) and treatment readiness were used in the model.

Static-99 scores and offender age at the beginning of treatment were entered first to ensure that the other predictors in the equation were predicting treatment attrition independently of the risk related variables. In the second block, denial (yes, no), minimization total scores, motivation index scores, and treatment readiness total scores were entered together into the equation. The model correctly classified treatment attrition status for 88.8% of the 448 participants.

Table 3 shows the results of the logistic regression. According to the Wald statistic, denial and minimization were significant predictors of treatment attrition status, while there was a trend towards significance for the TRS (p = .06). Offender age, the motivation index and the Static-99 score did not significantly predict treatment attrition status.

Table 3

Results of Logistic Regression Analysis Predicting  Treatment Attrition Status

Variable	β	SE - β	Wald	Odds Ratio	CI Odds
Step 1
Age	-0.01	0.01	0.63	0.99	0.97 - 1.02
Static-99	0.08	0.07	1.20	1.08	0.94 - 1.24
Step 2
Denial	-1.20	0.38	9.92**	0.30	0.14 - 0.64
Minimization	0.18	0.07	6.41**	1.19	1.04 - 1.37
Motivation	0.14	0.43	0.11	1.15	0.49 - 2.68
TRS	-0.02	0.01	3.38a	0.98	0.95 - 1.01

Put simply, offenders who admitted the sexual assault that led to their criminal conviction were significantly more likely to complete treatment than those who were classified as deniers. Similarly, lower levels of minimization were associated with treatment completion. The motivation index was not significantly related to treatment outcome, and, surprisingly, completers and non-completers demonstrated comparable scores in this area.

Discussion and conclusions

This study identified denial, minimization and treatment readiness as specific responsivity factors worthy of assessment prior to assigning sex offenders to a specific treatment program. These responsivity factors seemingly affect whether or not an offender completes treatment, even after taking risk into account.

Interestingly, the overall attrition rate for this group of sex offenders was 11.2%, a very low rate compared to other sex offender treatment programs where the attrition rates ranged from 20% to 58% in institutional settings.

The current programming model implemented by CSC may contribute to the lower rate of attrition among sex offenders. CSC sex offender programs take a cognitive-behavioural approach with an emphasis on reducing the risk of sexual recidivism by means of self-management and implementing external controls. Furthermore, the program intensity is matched to each offender's risk and need level, so that higher risk offenders are assigned to longer and more intensive treatment. This means that lower risk offenders are not subjected to long, intense treatment programs which may result in their dropping out because of being forced to continue treatment after having reached acceptable levels of functioning with respect to treatment targets.  

---

References:

Abracen, J., & Looman, J. (2004). Issues in the treatment of sexual offenders: Recent developments and directions for future research. Aggression and Violent Behavior, 9, 229-246.

Barbaree, H.E. (1991). Denial and minimization among sex offenders: Assessment and treatment outcome. Forum on Corrections Research, 3(4), 30-33.

Craissati, J., & Beech, A. (2001). Attrition in a community treatment program for child sexual abusers. Journal of Interpersonal Violence, 16, 205-221.

Dowden, C., & Serin, R. (2001). Anger management programming for offenders: The impact of program performance measures. Research Report R-106. Ottawa, ON: Correctional Service Canada.

Geer, T.M., Becker, J.V., Gray, S.R., & Krauss, D. (2001). Predictors of treatment completion in a correctional sex offender treatment program. International Journal of Offender Therapy and Comparative Criminology, 45, 302-313.

Hanson, R.K., & Thornton, D. (1999). Static-99: Improving actuarial risk assessments for sex offenders. (User Report 99-02). Ottawa, ON: Department of the Solicitor General of Canada.

Polizzi, D., MacKenzie, D., & Hickman, L. (1999). What works in adult sex offender treatment? A review of prison-based and non-prison-based treatment programs. International Journal of Offender Therapy and Comparative Criminology, 43, 357-374.

Serin, R., Kennedy, S., & Mailloux, D. (2002). Protocol for the treatment readiness, responsivity, and gain scale: Short version. Ottawa, ON: Correctional Service of Canada.

---

What’s new?

Research reports

R-172 (2006) A program for Aboriginal violent offenders: A preliminary evaluation of “In Search of Your Warrior”
R-176 (2006) The heterogeneity of treatment non-completers
R-177 (2006) Estimating risk of drop-out and expulsion from correctional programs
R-178 (2006) An examination of the effectiveness of the Violence Prevention Program
R-179 (2006) Empirical study of Correctional Officers during their first year at an institution

To obtain copies of specific reports, please contact us at (613) 995-3975 or by e-mail at research@csc-scc.gc.ca. You can also access our publications on the Internet via the Correctional Service Canada website at http://www.csc-scc.gc.ca.