Raymond A. Knight - US grants

There is growing evidence that some form of information input dysfunction predates the onset of schizophrenia and is a consistent trait of schizophrenics in both the acute and remitted stages of the disorder. Yet, despite the increasing consensus that some input deficit is implicated in schizophrenia, progress on the specification of this deficit has been sluggish. Although the study of schizophrenia has more than its share of nuisance variables that hinder research (7,14,16), the slow progress of process-specification seems to stem mostly from the task-oriented strategy that developed because of the absence of adequate models of normal cognition to guide research. Recent advances in cognitive psychology have produced converging models of early visual processing that have provided experimental psychopathologists with numerous paradigms to study clearly specified stages of processing from multiple perspectives. Thus, a powerful process-oriented research strategy that has several methodological advantages (56) is now possible. Using well-established theoretical models for early visual information processing, one can assess schizophrenics' input dysfunction with several paradigms that quantify these theoretical constructs in specific terms and predict a priori particular patterns of performance for both adequate and inadequate processing at each stage. By measuring the same processes from different perspectives over a series of studies, one can provide construct validation for hypothesized deficits, and circumvent confounding psychometric artifacts (14,15) that make the discrimination of specific cognitive deficits problematic. Converging lines of evidence from several investigators (57,62,63,95,106,114) suggest that poor prognosis schizophrenics have a perceptual organization dysfunction in their initial, wholistic processing stage. This theory not only integrates results of several studies of schizophrenics that had seemed contradictory, it also provides an explanation of other cognitive deficiencies in these subjects (56). It is the major intent of the research we are proposing to test this theory rigorously and to provide evidence that will allow greater specification of the nature of this deficit. Moreover, since the data suggest that this deficit is characteristic of only a specific subgroup of schizophrenics, poor prognosis patients, the proposed research will attempt to isolate some important discriminators of this subgroup of schizophrenics.

During the past four years, our research team has been engaged in a project aimed at organizing a complex multivariate database on sexual offenders into homogenous and reliable scales and dimensions for the purpose of generating useful taxonomies of this important, but inadequately studied, offender sample. Initial reliability and concurrent validity results of both our clinical and empirical strategies of clustering life history and offense data have been very promising. These results, however, have been based on a rather select sample--offenders currently committed to the Massachusetts Treatment Center as "sexually dangerous persons" --and have relied solely on concurrent data sources. The practical usefulness of our taxonomic solutions can only be evaluated in the context of generalization and follow-up data. The proposed research, therefore, aims at: (1) assessing the generalization and coverage of our typologies when applied to (a) the large sample of offenders observed at the Massachusetts Treatment Center and found to be "not sexually dangerous" and (b) a smaller sample of sexual offenders seen at the Oak Ridge Division of the Mental Health Centre in Penetanguishene, Ontario, Canada; (2) examining the cross-situational and prognostic utility of our dimensions and types by a follow-up of released sexual offenders (both "sexually dangerous" and "not sexually dangerous") through state and federal records; (3) using this follow-up data also as a means of unconfounding or collapsing subtypes as a function of their predictive validity; and (4) beginning to asses aspects of person-situation interactions and institutional programs as they affect post-release outcome and recidivism. The significance of the proposed research lies primarily in the benefits a valid, predictive, widely applicable taxonomy would have both for subsequent research into the etiology, diagnosis, treatment, and management of sexual aggression and for the elucidation this admittedly preliminary data would provided for current decision-making about such offenders.

The frequent clinical observation that rapists constitute a markedly heterogeneous group has led numerous clinicians and researchers to propose taxonomic systems aimed at identifying more homogeneous subgroups (7). Because all decisions about such offenders inevitably involve the interpretation of a particular case as a member of a larger group, the creation of a reliable and valid classification system could have important implications for treatment, management. and disposition. In spite of the potential importance of taxonomic considerations and their prominence in clinical theorizing, there have been a paucity of empirical studies of the various proposed systems. During the last decade we have been studying the critical taxonomic problems of rapists. We revised and operationalized a rationally derived system that incorporated the major types of rapists described in the clinical literature. In a series of studies (21- 24) we assessed the reliability and validity of this system, determining its strengths and weaknesses and identifying problem areas that required revision. We also cluster analyzed samples of rapists, and tested the validity of various types that evidenced cross-sample stability (e.g.,25). Finally, we integrated these two sets of studies and created anew system. This new system solves all of the problems we identified in our analyses of our original rationally derived topology, incorporates the new stable types that emerged in our cluster analytic studies, and provides anew empirically-based structure that has substantial heuristic potential. The major aim of the current proposal is to test the reliability and concurrent and predictive validity of this new system. Because of the extensive computerized clinical and criminal file data base that we have established on a large sample of rapists and the post release follow-up data we have on a subsample of these offenders, a major validation study can be conducted efficiently in a relatively short time.

DESCRIPTION (Adapted from applicant's abstract): Rape is a serious physical assault that has devastating long-term physical and emotional effects. The extensiveness and seriousness of this problem demands a concerted, effective social response. Substantial evidence indicates that sexual aggression is determined by a multiplicity of variables and is committed by a heterogeneous group of offenders. The present research proposal is designed to address three critical issues about the assessment of sexual aggression. First, the investigators propose to continue validation research on an easily administered and processed inventory, the Multidimensional Assessment of Sex and Aggression (MASA), for which both computerized and paper-and-pencil administration forms exist. This inventory has been designed to assess the critical domains that discriminate among sex offenders, and it has already demonstrated some evidence of reliability and validity. The investigators propose to improve the MASA by expanding its assessment of both the developmental antecedents of aggression and of domains critical to the evaluation of child molesters. Second, the investigators propose to use the MASA to revise and simplify our rapist typology (the Massachusetts Treatment Center Typology for Rapists [MTC-R3], which is currently the only empirically validated, reliable typology for rapists. Third, the investigators propose to continue the development of a version of the MASA for juvenile sex offenders and to use this instrument to develop a valid typology for this population.