Objective: Evaluate the measurement properties of the Neurobehavioral Cogni
tive Status Examination (NCSE) using rating scale analysis. Design: Calibra
tion of item responses collected as part of a study examining characteristi
cs of case-management programs and treatment outcomes. Setting: Three outpa
tient rehabilitation facilities. Participants: Convenience sample of 186 co
mmunity-dwelling adults with TBI. Results: Several rating scale analyses we
re performed to construct a unidimensional measure. Deletion of easy and mi
sfitting items created a better targeted test (generated more spread among
individuals) without increased error. Conclusions: If used with a community
-based sample, three strata can be differentiated despite a skewed distribu
tion. Recommended applications are with samples with considerably more cogn
itive impairment.