Using Diagnostic Classification Models to Diagnose Children's Emotional and Behavioral Problems in Preschool

• In Event: Important Considerations for Survey Data

Abstract

Survey practitioners are interested in obtaining detailed information on each respondent and survey question. The Pediatric Symptom Checklist-17 (PSC-17) is a high-quality universal screening tool to identify children with emotional and behavioral risks in the school environment. In this study, we analyzed a large sample (N= 1147) of teachers’ ratings on preschool children using PSC-17 by a new group of psychometric models, named Diagnostic Classification Models (DCMs). Different from most psychometric models that require a cutoff score to make diagnostic decisions, DCMs provide classifications directly. We investigated multiple DCMs and selected the optimal model using information-based indices. We then conducted the item analysis and person classification for this model. This study is a didactic application of DCMs to survey instruments.

Authors

• Yu Bao, James Madison University

• Jin Liu, University of South Carolina - Columbia

• Christine DiStefano, University of South Carolina