Verbal fluency performance after traumatic brain injury: a cluster analysis.

Abstract

This study examines the sensitivity of two different analyses for comparing verbal fluency test performance of participants with mild cognitive impairment resulting from traumatic brain injury (TBI) and healthy controls. The study uses a cluster analysis based on Troyer, Moscovitch, and Winocur (1997) of switching and clustering for the first time in a TBI-only sample. The sensitivity of number of correct words produced, a standard measure of verbal fluency performance, is also examined. The effects of demographic variables of education and estimated verbal IQ are reported. Participants included 31 adults with TBI and 26 adult controls. Significant group differences were found in estimated verbal IQ. Number of correct words produced was no more sensitive for detecting group differences in verbal fluency performance than total responses (correct and incorrect) in letter and category fluency conditions, but may be more sensitive in category switching. No group differences were found in any cluster analysis measure when estimated verbal IQ and total responses were controlled. Results highlighted the importance of controlling for demographic variables including estimated verbal IQ when interpreting verbal fluency data and that complex, multifaceted analyses of verbal fluency performance may not add to the ability of verbal fluency to detect mild cognitive impairments following TBI. Clinical implications are discussed.