Still more on Ozanoff et al., 2005

As I've said before, one of the major advantages of being a grad student is access to the faculty. I finally got to speak to someone familiar with the MMPI about Ozanoff et al., and several interesting things came from the discussion.

First off, the L scale (one of the validity scales, intended to detect attempts to represent yourself as "better" than you really are) probably isn't valid for assessing autistics. I commented on that a bit in my previous post, but I wasn't fully aware of the sort of questions that were involved in the L scale.

Simply put, autistics would legitimately answer many of them in a different manner than neurotypicals.

What I found out about the restructured clinical (RC) scales, however, calls into question a number of things about my first set of observations. At least some of the effects I noted are probably explainable by the differences in scores on the RCd, or demoralization, scale, which measures common factors that influence all of the original scales.

For the record, the autistic group scored higher on that one.

Of course, even within the RC scales, there were effects. Aside from the aforementioned effect on the RCd scale, the autistic group scored higher on the RC2 scale, which measures low positive emotions. This is probably explainable by the fact that nearly a quarter of the selected sample had a diagnosis of major depression... and the change in the statistics when these persons are factored out supports this conclusion.

Finally, autistic individuals scored lower on the RC9 (hypomanic activation) scale, which basically measures hyperactivity. I'd comment on how interesting this was if the effect was anywhere near substantial.

You see, even with everything I've mentioned so far? None -- none -- of the autistic means were in the clinically significant range. While there was one scale (8/8-K, officially known as the schizophrenia scale, which measures odd thinking and social alienation) in which 40% of the sample was in the clinical range, this only really means that 40% of the sample thought "strangely" enough and were socially alienated enough for the result to attract a doctor's (or a psychologist's) attention.