Psychometric properties of the MacArthur Competence Assessment Tool–Criminal Adjudication.

Abstract

University of VirginiaThis article describes the development of a new clinical instrument for use in assessments of adultcriminal defendants' competence to proceed to adjudication, the MacArthur Competence AssessmentTool-Criminal Adjudication (MacCAT-CA). The MacCAT-CA was derived from a more comprehen-sive research instrument (MacArthur Structured Assessment of Competencies of Criminal Defendants;Hoge, Bonnie, Poythress, Monahan, & Eisenberg, 1997) on the basis of considerations efface validityfor use in legal contexts, psychometric analyses, and advice from mental health experts who reviewedan earlier prototype. This article presents the results from an National Institute of Mental Health-sponsored validation study that investigated the psychometric properties of the MacCAT-CA.At least 25,000 criminal defendants are referred annually forevaluation of their competence to participate in legal proceed-ings (Steadman & Hartstone, 1983). Although a number ofmeasures have been specifically designed to assess defendants'capacities in this area (e.g., Competence Screening Test [CST],Lipsitt, Lelos, & McGarry, 1971; Georgia Court CompetencyTest [GCCT], Wildman et al., 1980; CADCOMP, Barnard etal., 1991; Competency Assessment Instrument [CAI], Labora-tory of Community Psychiatry, 1974; Interdisciplinary FitnessInterview [IFI], Golding, Roesch, & Schreiber, 1984; Fitnessto Stand Trial Interview, Menzies, Webster, Roesch, Jensen, R RobertA. Nicholson, Department of Psychology, University of Tulsa; JohnMonahan and Richard J. Bonnie, School of Law, University of Virginia;Steven K. Hoge, School of Law and School of Medicine, University ofVirginia; Marlene Eisenberg, Institute of Law, Psychiatry, and PublicPolicy, University of Virginia.John F. Edens is now at the Department of Psychology, Sam HoustonState University.This research was supported by the MacArthur Foundation ResearchNetwork on Mental Health and the Law, and by Grant RO1MH54517-O1A1 from the National Institute of Mental Health. We are grateful toTom Feucht-Haviar for his assistance on a pilot test of this measure.Norman G. Poythress, John Monahan, Richard J. Bonnie, and StevenK. Hoge are authors of the MacArthur Competence Assessment Tool-Criminal Adjudication (MacCAT-CA) published by Psychological As-sessment Resources. Randy K. Otto, Norman G. Poythress, Robert A.Nicholson, John F. Edens, John Monahan, Richard J. Bonnie, and StevenK. Hoge are authors of the MacCAT-CA manual, also published byPsychological Assessment Resources.Correspondence concerning this article should be addressed to RandyK. Otto, Department of Mental Health Law and Policy, Florida MentalHealth Institute, University of South Florida, 13301 Bruce B. DownsBoulevard, Tampa, Florida 33612-3899. Electronic mail may be sent toorto@hal.fmhi.usf.edu.Eaves, 1984), each exhibits various limitations (Grisso, 1991;Melton, Petrila, Poythress, & Slobogin, 1997; Nicholson, 1992).Briefly, some measures were designed to serve primarily asscreening instruments (e.g., CST and GCCT) and appear mainlyto be collections of items having various legal content but littleor no underlying conceptual structure. Others (e.g., CAI and IFI)lack standardized administration and criterion-based scoring,relying instead on the judgment of the clinician to determine,on a case-by-case basis, which questions to pose to the defendantand how to rate the responses received. None is designed toprovide quantitative indexes of discrete competence-related abil-ities, nor have interpretative norms based on large, nationalsamples been developed for any of these measures.Given the pivotal role of a defendant's competence, one of themajor research initiatives of the MacArthur Foundation ResearchNetwork on Mental Health and the Law was the developmentof a standardized research instrument for evaluating criminaldefendants' psycholegal abilities related to competence to pro-ceed to adjudication. The MacArthur Structured Assessmentof Competencies of Criminal Defendants (MacSAC-CD; Hoge,Bonnie, Poythress, Monahan, & Eisenberg, 1997) was based onBonnie's (1992, 1993) comprehensive theory of legal compe-tence, and it contains measures of discrete competence-relatedabilities. Items in the MacSAC-CD were modeled after thosecreated by Grisso, Appelbaum, Mulvey, and Fletcher (1995) forthe assessment of competence to consent to treatment (see, moregenerally, Grisso & Appelbaum, 1998). Each measure in theMacSAC-CD involves standardize d administration an crite-rion-based scoring, thus reducing much of the discretionary(and potentially idiosyncratic) scoring that plagues existingmeasures.Hoge et al. (1997) recently described the MacSAC-CDand presented the primary findings from a field study that in-volved both incompetent and competent defendants as researchparticipants. The results of this field study indicated that the