In clinical practice, CT generally handles this issue by collaboratively developing goals at the outset of treatment and then monitoring progress towards these goals over the course of treatment. For example, if a client starts therapy with a worthwhile but ambiguous goal like "Control my temper" they and I would try to develop a clearer (and more behaviorally defined) idea of what our goal is. I might do this by asking questions like "How would we know if we've accomplished this? What changes would you like to see in how you handle anger?" After a bit of discussion and clarification we might end up with goals such as "Decrease the frequency and intensity of angry outbursts," "Deal with anger and conflicts effectively without swearing or physical fights," "Calm down more quickly after becoming angry," etc. Then, if I collect some baseline data at the beginning of therapy, it is fairly easy to tell whether we are making progress towards these goals as therapy proceeds.

In research, the approach to assessing outcome needs to be a bit different because we need to deal with a large pool of subjects rather than dealing with clients individually, because we need numerical rather than qualitative measures of change for our statistical analysis, and because we need to guard against a variety of possible methodological problems. We still need to think in terms of "What are we trying to accomplish and how can we measure it?" but we need to find reliable, valid, and practical measures.

Too many CBT researchers rely on paper-and-pencil self-report measures because these are cheap and easy to use. However, it often is not safe to assume that subjects will be willing and able to accurately report the data we need, especially since the subjects in your study (i.e. with antisocial personality disorder) do not have a great reputation for being honest and strightforward.

I only remember one controlled outcome study of CT with antisocial personality disorder (Woody, et al, 1985) and I believe they used 22 outcome measures ranging from BDI scores to work attendance. I'd recommend looking at this study as an example of researchers doing a decent job of measuring outcome in the population you will be studying. By the way, they found that CT was effective with antisocial subjects who were also depressed at the outset of treatment but was ineffective with antisocial subjects who were not depressed at the outset of treatment, so you may want to include pre-treatment BDI scores in your study.

The other source I'd recommend taking a look at is Turkat's book on CBT with personality disorders (Turkat, 1990). He describes an empirically-based approach to personality disorders using a series of single-case designs and includes a brief section on antisocial personality disorder.

• Turkat, I. D. (1990). The Personality Disorders: A Psychological Approach to Clinical Management. New York: Pergamon.




• Woody, G. E., McLellan, A. T., Luborsky, L., & O'Brien, C. P., (1985). Sociopathy and psychotherapy outcome. Archives of General Psychiatry, 42, 1081-1086.