FORUM on Corrections Research

Predicting treatment response in correctional settings

Researchers have recently established that certain correctional treatment programs effectively reduce recidivism. This has led to the identification of the characteristics of programs that "work" and programs that "don't."⁽²⁾ However, little attention has been paid to how individual offender traits might interact with program characteristics and affect treatment outcome.

The Attitudes Toward Correctional Treatment Scale directly addresses this issue.⁽³⁾ There were two main reasons for the development of the scale. A device was needed to reliably identify offenders who are motivated for treatment, as well as offender attitudes and traits that might inhibit treatment and should, therefore, be addressed beforehand.

Further, there had been no real means of specifically assessing offender motivation. In fact, few of the established tools in this area were relevant to correctional settings.

This article, therefore, provides a brief description of the scale, as well as an assessment of its effectiveness. Perhaps most important, the article analyzes the potential impact of this scale on both offender assessment and treatment. Offender attitudes and treatment Although there have been several comprehensive reviews of the general factors that influence psychotherapy outcomes,⁽⁴⁾ few studies have attempted to predict offender response to treatment. Clearly, correctional treatment settings differ from other treatment environments and offenders differ from other treatment clients.

In particular, studies have often suggested that antisocial personality characteristics, psychopathic traits or strong pro-criminal attitudes could be significant obstacles to therapy.⁽⁵⁾

Some studies have recommended certain scales as potential predictors of offender treatment response, but the research in this area has produced conflicting results and few of the studies have had direct relevance to correctional settings.⁽⁶⁾ The Attitudes Toward Correctional Treatment Scale The current version of the Attitudes Toward Correctional Treatment Scale consists of 33 items that offenders score on a five-point scale, ranging from strongly disagree, to uncertain, to strongly agree.

This produces a total score, as well as scores in five subscale categories (the higher the subscale score, the greater the motivation or the more positive the attitude):

• motivation and perceived need for treatment;
• perceptions of treatment and the institution;
• perceptions of staff;
• optimism/pessimism regarding treatment outcome; and
• comfort/discomfort with self-disclosure in groups.

Data relating to the scale has now been compiled for 1,433 men assessed at the Rideau Correctional and Treatment Centre during the past three years. The internal consistency statistics for the subscales were satisfactory (range .70 to .87), as are preliminary test-retest coefficients (range .58 to .72).

It should be noted that during the time between the test and retest (10-14 days), the offenders had several contacts with both correctional and clinical staff, and participated in a pretreatment communication skills group. Consequently, some of the variability between test and retest scores may reflect a desirable sensitivity to short-term changes resulting from the offenders' intervening therapeutic experiences.

Figure 1

A comparison of the sample's basic demographic and offence data with those of other recent studies suggests that the present sample is not atypical for provincially incarcerated inmates, except for a somewhat higher than average prevalence of substance abuse.⁽⁷⁾

The sample's average offender was about 30 years old, had a grade 10 education and had been incarcerated three or four times previously (primarily for property or alcohol/drug offences). Assessment outcome To examine the relationship between the Attitudes Toward Correctional Treatment Scale scores and assessment outcome, 1,327 offenders with confirmed disposition data were divided into three outcome groups: no treatment recommended (55 offenders), treatment recommended but declined (256 offenders), and treatment recommended and completed (1,016 offenders).

All of the offenders completed the Attitudes Toward Correctional Treatment Scale (and several other instruments) during the standard pretreatment assessment process.

The results indicated that higher scale scores were associated with a better assessment outcome. Both the total scale score and two subscale scores (motivation and optimism) showed significant progressive increases across the three groups (see Figure 1). Differences in the remaining subscales were not statistically significant. Treatment outcome A quasi-random procedure was used to select 476 offenders (24 others were rejected because of incomplete data) from those who entered the centre's assessment unit between 1992 and 1994, and went on to participate in the Rideau Addictions Program and/or the Anger Management Program.

Both programs use a basic cognitive-behavioural skills-oriented approach, although the anger program is somewhat smaller and more intensive than the addictions program, with more individual attention.

Treatment outcome was measured through the final ratings of overall program participation and progress for each offender, as rated by program leaders on an eight-point scale, ranging from 1 (unsatisfactory), to 4 (good), to 8 (excellent).

Both the total scale score and all of the subscale scores correlated positively (if modestly) with treatment outcome ratings for both the anger and additions treatment groups.⁽⁸⁾ The highest correlations were with the overall score and the motivation subscale (see Table 1).

Table 1

Corrections Between Attitudes Towards Correctional Treatment Scale Scores and Treatment Outcome
Subscale	Rideau Addictions Program outcome	Anger Management Program outcome
Motivation	0.26	0.26
Treatment perceptions	0.12	0.18
Staff perceptions	0.11	0.19
Optimism	0.16	0.28
Comfort with self-disclosure	0.15	0.18
Total	0.24	0.31
Note: All correlations are significant at p<0.05 or better

The 476 offenders were then divided into three groups based on their overall and subscale scores - the lowest 25% (the low group), the highest 25% (the high group) and the middle 50% (the medium group). The groups differed significantly in age and education, so these variables were entered as covariates, where appropriate, in statistical analyses.

Significant differences in both Rideau Addictions Program and Anger Management Program ratings were found among the groups in their overall scores and in their motivation scores. Significant differences were also found in the addictions program ratings in the treatment perception and optimism scores.

Further, the high group had significantly better anger program outcome ratings than the low and medium groups in relation to their motivation and optimism subscale scores and total scores (see Figure 2).

Figure 2

As for the addictions program, the high group had significantly better outcome ratings than the low group for all scores except the comfort with self-disclosure subscale score, while the medium group did not differ significantly from the other two groups (see Figure 3).

Figure 3

The MMPI-2 Negative Treatment Indicators scale MMPI-2 Negative Treatment Indicators scale (a new "content" scale on the revised Minnesota Multiphasic Personality Inventory) results were inversely correlated with all Attitudes Toward Correctional Treatment Scale subscales.

High scores on this scale indicate personality traits or attitudes toward treatment that suggest resistance to change,⁽⁹⁾ so this provides some evidence of concurrent validity for the Attitudes Toward Correctional Treatment Scale.

However, the correlations between the MMPI-2 Negative Treatment Indicators scale results and the treatment perceptions, staff perceptions and comfort with self-disclosure subscales were substantially higher than those for motivation and optimism.

This suggests, among other things, that the MMPI-2 Negative Treatment Indicators scale should not be interpreted as a measure of motivation for treatment per se, but as a reflection of general negative attitudes toward treatment and mental health professionals.⁽¹⁰⁾

More than anything, this illustrates that "treatment motivation" and "amenability to treatment" are multidimensional concepts, encompassing a variety of attitudes, beliefs, perceptions and misperceptions about the nature of treatment and the therapists involved. What does it all mean? These results suggest that the Attitudes Toward Correctional Treatment Scale is a valid and reliable predictor of offender treatment outcome. There were some differences between the two treatment groups sampled, but this is not surprising given their format and content differences.

In short, the scale can seemingly serve as an objective tool for evaluating offender suitability for treatment which, until now, has been largely based on clinical judgment (informed guesswork).

With the chronic shortage of correctional treatment resources, we must have a reliable means of identifying who will benefit most from treatment. This scale should help prioritize offenders and minimize dropout rates.

Perhaps more important, the scale may help to maximize the benefits of treatment for specific offenders, through early identification of attitudes and beliefs likely to impede treatment progress, allowing therapists to address these attitudes and beliefs in pretreatment counselling.

---

(1)Rideau Correctional and Treatment Centre, RR3, Merrickville, Ontario K0G 1N0.
(2)D.A. Andrews and I. Bonta, The Psychology of Criminal Conduct (Cincinnati: Anderson Publishing, 1994). See also D. A. Andrews, I. Bonta and R. D. Hoge, "Classification for Effective Rehabilitation: Rediscovering Psychology," Criminal Justice and Behavior, 17 (1990): 19-52. And see D. A. Andrews, I. Zinger, R. D. Hoge, I. Bonta, P. Gendreau and F. T. Cullen, "Does Correctional Treatment Work? A Psychologically Informed Meta-analysis," Criminology, 28 (1990): 369-404. And see P. Gendreau and R. R. Ross, "Revivification of Rehabilitation: Evidence from the 1980s," Justice Quarterly, 4 (1987): 349-407.
(3)D. I. Baxter, M. Burchill and M. Tweedale, The Attitudes Toward Correctional Treatment Scale (Merrickville: Ministry of the Solicitor General and Correctional Services, 1992).
(4)D. A. Andrews and J. S. Wormith, Criminal Sentiments and Criminal Behaviour (Ottawa: Solicitor General of Canada, 1984).
(5)A. I. Rabin, "The Antisocial Personality: Psychopathy and Sociopathy," Psychology of Crime and Criminal Justice, H. Toch, ed. (Prospect Heights: Waveland Press, 1992): 322-346. See also R. B. Sloan, F. R. Staples, A. H. Cristol, N.J. Yorkston and K. Whipple, "Patients' Characteristics and Outcome in Psychotherapy and Behaviour Therapy," Journal of Consulting and Clinical Psychology, 44 (1976): 330-339. And see G.E. Woody, A. T. McLellan, L. Luborsky and C. P. O'Brien, "Sociopathy and Psychotherapy Outcome," Archives of General Psychiatry, 42 (1985): 1081-1086.
(6)J N. Butcher, The MMPI-2 in Psychological Treatment (New York: Oxford University Press, 1990). See also I. N. Butcher, I.R. Graham, C. L. Williams, and Y. S. Ben-Porath, Development and Use of the MMPI-2 Content Scales (Minneapolis: University of Minnesota Press, 1990). And see J. R. Graham, MMPI-2: Assessing Personality and Psychopathology (New York: Oxford University Press, 1990).
(7)I.E. McKenzie, Psychometric Correlates of Behaviourial Impulsivity in Adult Male Criminal Offenders, Unpublished M.A. thesis, Carleton University, Department of Psychology, 1992. See also M. Tweedale, Interpersonal Problem-solving and Deterrence: Effects on Prison Adjustment and Recidivism, Ph.D. thesis, University of Ottawa, School of Psychology, 1990.
(8)Correlations between many of the MMPI-2 scales and the outcome ratings were also significant, but the magnitude of the correlations was lower than those obtained with the Attitudes Toward Correctional Treatment Scale.
(9)Butcher, The MMPI-2 in Psychological Treatment.
(10)Butcher, The MMPI-2 in Psychological Treatment. See also I. R. Graham, MMPI-2: Assessing Personality and Psychopathology.