Evaluation of newly developed item banks for child-reported outcomes of participation following spinal cord injury

Study design: Multi-center, prospective calibration study. Objectives: To examine the psychometric properties of item banks designed for a computer adaptive test (CAT) of participation. Setting: Three pediatric specialty care hospitals in North America. Methods: Newly developed item banks containing questions about participation were administered to 381 children, 8-21 years. Unidimensionality was evaluated using categorical confirmatory factor analysis (CFA). CFA model fit was assessed by the comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA) and residual correlations. Pearson's correlations were calculated between simulated CATs and the full-item banks. Results: CFI = 0.905, TLI = 0.951 and RMSEA = 0.089 for the self-participation item bank; CFI - 0.924, TLI = 0.963 and RMSEA = 0.083 friend-participation item bank. The 15-(r = 0.96 self; r = 0.95 friend), 10-(r = 0.93 self; r = 0.92 friend) and 5-(r = 0.85 self; r = 0.85 friend) item simulated CATs and full-item banks were highly correlated. Conclusion: The newly developed self and friend-participation item banks met the required assumptions for CAT. Simulated CATs yielded scores comparable to full-item banks. The results of the study support the use of the item banks for a CAT of participation in children following spinal cord injury. Sponsorship: The study was funded by the Shriners Hospitals for Children Research Advisory Board (RAB #9146). Spinal Cord (2012) 50, 915-919; doi:10.1038/sc.2012.80; published online 21 August 2012