作者: James W. Langenbucher , Erich Labouvie , Christopher S. Martin , Pilar M. Sanjuan , Lawrence Bavly
DOI: 10.1037/0021-843X.113.1.72
关键词:
摘要: Item response theory (IRT) is supplanting classical test as the basis for measures development. This study demonstrated utility of IRT evaluating DSM-IV diagnostic criteria. Data on alcohol, cannabis, and cocaine symptoms from 372 adult clinical participants interviewed with Composite International Diagnostic Interview--Expanded Substance Abuse Module (CIDI-SAM) were analyzed Mplus (B. Muthen & L. Muthen, 1998) MULTILOG (D. Thissen, 1991) software. Tolerance legal problems criteria dropped because poor fit a unidimensional model. curves, information testing variously constrained models suggested that in CIDI-SAM discriminate between only impaired less cases may not be useful to scale case severity. can used construct validity diagnoses identify performance.