Recommendations for treatment are made on the premise that assessment test data is accurate. This assumes that the client was truthful when tested. For responsible assessment, it must be established beyond a reasonable doubt, that the client (patient, offender) was honest and sincere. A second requirement of accurate assessment involves problem identification within the context of the client’s life situation. Tests designed for specific clients that contain multiple scales can accurately identify client problems. Without accurate test results, we can’t recommend appropriate intervention or treatment.

Truthfulness Scale

One important approach to resolving the dilemma of client truthfulness involves a Truthfulness Scale, which is a series of questions that determine the respondent’s openness, cooperativeness and truthfulness. Historically, the Minnesota Multiphasic Personality Inventory (MMPI) introduced Truthfulness Scales (L, F and K-Scales) almost six decades ago. Yet, for some paradoxical reason, Truthfulness Scales did not catch on. In 1990 Truthfulness Scales were given renewed support by Andrews, Bonta and Hogue’s study that demonstrated identified problem severity must match treatment intensity for optimal outcomes. Subsequent studies reinforced the need for accurate assessment to facilitate effective treatment. Yet, in Jones' (1996) psychometric review, “client truthfulness” research was conspicuously absent. Despite these setbacks, some researchers and psychometricians have continued to develop and research Truthfulness Scales. The Risk & Needs Assessment, Inc. (Risk & Needs) website www.riskandneeds.com discusses much of this Truthfulness Scale research, as all of their tests contain Truthfulness Scales.

Tests with Multiple Scales

Another assessment driven treatment issue involves problem identification within the context of the client’s life situation. The days of one-dimensional or one-factor tests are over. Contemporary tests contain multiple scales to identify clients’ focal problems while concurrently measuring contributing (substance abuse, attitudes and behaviors) factors. Examples of these multiple scale tests are many and include the Driver Risk Inventory-II, Prison Inmate Inventory, Domestic Violence Inventory and Sexual Adjustment Inventory. For sex offender assessment, scales measure sex problems such as sexual adjustment, child molest tendencies, rape potential, etc. But there may be co-occurring problems such as substance abuse, violence potential, antisocial thinking and emotional instability. Tests that are multidimensional give an accurate understanding of all of these offender problems. This enables evaluators to make recommendations for appropriate interventions. Accurate assessment, in this sense, means that clients are tested with scales that are pertinent to understanding them. What scales are in each test is important because these scales have profound implications for recommending treatment. One test is not applicable to all offenders, and just alcohol and drug information is not enough.  Consequently, all Risk & Needs tests have multiple scales.

Many of these test scales have been called “criminogenic needs” (Andrews, et al., 1990) and have been demonstrated to be significant factors in the prediction of recidivism (Gendreau, et. al., 1996). Accuracy of recidivism predictions is another way to study test accuracy. Davignon (2002) demonstrated that test scales measuring substance abuse, aggressive driver behavior and stress coping skills were accurate and reliable in a DUI/DWI offender population.

The Andrews, et al. (1990) study is noteworthy because it demonstrates that optimal treatment is contingent upon accurate problem severity identification and corresponding intensive treatment level placement. For example, offenders with mild problems benefited most from placement in low intensive treatment programs, and clients with severe problems require intensive treatment. These statements sound obvious, but they are now empirically based. And, of equal importance, placing offenders with mild problems in highly intensive treatment programs, or placing offenders with severe problems in mildly intensive treatment programs, increased recidivism and the offenders’ risk to society.

Summary

In summary, there are two important requirements that assessment tests must satisfy. Both requirements involve accuracy. First and foremost, offender truthfulness must be established. Offender truthfulness while completing a test is necessary; otherwise, there can be no confidence in test scores. Tests must include a Truthfulness Scale to measure offenders’ openness, denial, problem minimization and attempts to “fake good.” When Truthfulness Scale scores are very high (90^th percentile and above) the offender was not truthful and invalidated their test results. The second important requirement of testing is accurate identification and measurement of offender problems. To accurately identify offenders’ problems, tests must contain scales relevant to the assessment’s purpose. For specific client or offender groups, e.g., inmates, sex offenders, counseling patients, substance abuse programs, etc., you want specific information. This requires tests developed for specific assessment purposes. Test scales can then be standardized on specific offender groups that the test was designed to evaluate. With these two requirements, “offender truthfulness” and “relevant test scales,” met, the next step is to set forth recommendations for intervention and treatment. These recommendations must be scale-score related, in the sense that intervention addresses the severity and extent of offender problems.  For a review of tests designed for specific offender and client groups you should review the Tests Functionally Grouped webpage.

References

Andrews, D.A., Bonta, J, & Hoge, R.D. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52.

Davignon, D. (2002). DWI/DUI arrests, BAC at the time of arrest, and offender assessment test results for alcohol problems. Impaired Driving Update, 6, 5-44.

Gendreau, P., Little, T., & Goggin, C. (1996). A meta-analysis of the predictors of adult offender recidivism: What works! Criminology, 34, 575-607.

Jones, P.R. (1996). Risk prediction in criminal justice. In A.T. Harland (Ed.), Choosing correctional options that work: Defining the demand and evaluating the supply. Thousand Oaks, CA: Sage.

* * * * *