Abstract
Introduction
The semantic fluency task is described as a measure of semantic function and utilised in schizophrenia (SZ) research to demonstrate semantic deficits. Two meta-analyses support the efficacy of this task in measuring semantic function in SZ; however, a more recent meta-analysis suggests that executive dysfunction is the predominant determinant of semantic fluency performance in this group. By (1) matching the semantic and executive fluency tasks on discriminant validity, and (2) including an additional fluency task containing both semantic and executive elements (animals by size), this study aimed to determine whether semantic fluency is in fact an appropriate measure of semantic function in SZ.
Methods
Forty-two SZ participants and 40 healthy controls performed 3 fluency tasks: executive (F, A, S), semantic (body parts), and semantic/executive (animals by size). Performance on these tasks was analysed in two ways, (1) based on discriminant validity and (2) by matching output between the semantic and semantic/executive fluency tasks.
Results
When the semantic and executive fluency tasks were matched, executive fluency performance was either (1) mildly impaired or (2) not impaired in SZ. Both semantic and semantic/executive performance was impaired in SZ regardless of calculation. Group differences on the semantic/executive task remained when executive function was controlled for, but disappeared when semantic fluency effects were controlled for.
Conclusions
The findings support earlier meta-analyses in finding that the semantic fluency task is a robust measure of semantic memory function in SZ.
Notes
1. The ANCOVAs were rerun with only patients on atypical medications (i.e., removing five participants) for both methods and there were no changes in results.