Computer-adaptive test to measure community reintegration of veterans Linda Resnik, PT, PhD, OCS; Feng Tian; Pengsheng Ni, MD; Alan Jette, PhD, PT
Aim • Develop and evaluate briefer version of Community Reintegration of Injured Service Members (CRIS) measure. • Used item response theory (IRT) and computer-adaptive test (CAT) methodologies. • Relevance • Early identification and prevention of community reintegration problems and evaluation of clinical interventions to promote healthy social role functioning require accurate assessment and monitoring of community reintegration.
Methods • Constructed large item banks for each CRIS scale. • Tested on convenience sample of 517 Veterans. • Used exploratory and confirmatory factor analyses to identify dimensionality within each domain. • Used IRT to calibrate items. • Compared accuracy and precision of CATs of different lengths with full-item bank. • Examined data for differential item functioning (DIF).
Results • Confirmatory factor analyses supported scales’ unidimensionality. • Acceptable item fit statistics found for final models. • High accuracy (0.88+) found for 10-, 15-, 20-, and variable-item CATS for all scales. • CAT precision increased with number of items and decreased at upper ranges of each scale. • Three items exhibited moderate DIF by sex.
Conclusions • CRIS-CAT demonstrated promising measurement properties and is recommended for use in community reintegration assessment. • Routine community reintegration assessment would: • Enhance patient assessment and targeting of referrals to mental health, social services, and Veterans Health Administration benefit programs. • Inform primary care and other interventions that address underlying factors related to poor community reintegration.