William M Grove
I also perform assessment research related to the (MMPI and MMPI-2 and certain other widely-used instruments,. Another interest is statistical (actuarial, mechanical) versus clinical (judgment-based) prediction methods, and the development of sound statistical procedures for accurately predicting important behaviors, e.g., criminal offender recidivism.
These interests concern latent categorical versus dimensional psychopathology constructs. For example, investigators debate whether DSM-IV disorders represent distinct disease entities (latent categories). One reason this debate matters is that if a disorder is categorical, it may have a single, categorical cause strongly influencing development of the disorder. This kind of study could greatly facilitate etiological research.
Taxometric methods, originally developed by Paul E. Meehl include procedures for analyzing multivariate data to determine whether there is a taxon and its complement class. These procedures provide risky tests of categorical theories. I develop new procedures to supplement existing ones. I also study the statistical properties of such procedures, by mathematical and simulation studies.
Finally, I study the relationship between dimensional and categorical psychological measures (e.g., test scores versus DSM-IV diagnoses) and latent categories, in several ways. First, I am interested in ways of classifying individuals into the taxon versus the complement class. Second, I am interested in relating inferred taxon membership (i.e., diagnosis), measures that allow one to infer taxon membership, and other variables that one may wish to predict (e.g., prognosis, future behavior).
I also have an interest in forensic psychology. I am interested in forensic assessment and prediction (e.g., sex offender recidivism), and in improving the scientific quality of expert witness testimony in courtroom proceedings.