Rasch analysis of the Executive Interview (The EXIT-25) and introduction of an abridged version (The Quick EXIT).

Abstract

To evaluate the psychometric properties of the Executive Interview (EXIT-25) and to propose modifications that will improve those properties. Rasch analysis of existing datasets contributed by 3 prior projects, all of which examined criterion-related validity of the EXIT-25. A large, urban, academic free-standing rehabilitation facility. The sample of 147 was comprised of 109 adults diagnosed with stroke evaluated during inpatient rehabilitation and 38 adults with traumatic brain injury evaluated during inpatient (n=11) or outpatient rehabilitation (n=27). Not applicable. The EXIT-25, Repeatable Battery for the Assessment of Neuropsychological Status, and Trails A and B. Eleven of the 25 items correlated weakly with the total measure and misfit the rating scale model. Deleting these 11 items improved the internal consistency of the remaining 14 items and enhanced the measure's criterion-related validity. The EXIT-25 can be reduced from 25 to 14 items without reducing internal consistency. Convergent validity of the abbreviated measure is supported by moderate-size correlations with standard measures of cognitive deficits.