Clustering, Switching & Verbal Fluency Assessment: A Critical Evaluation of Introspection and Interview-based Scoring Procedures

Abstract

Semantic Verbal Fluency (SVF) is a neuropsychological assessment that requires respondents to rapidly generate words in specified semantic categories (e.g., different kinds of animals). It provides information on the integrity of examinees’ executive function and semantic memory by looking at clustering of words generated and switching between clusters. According to recent literature, extant methods for scoring clustering and switching based on examiner-identified semantic associations among words should be abandoned in favor of scores derived from asking participants post hoc how they think they generated their words the way they did (Body & Muskett, 2013). Repeated measures ANOVA indicated that leading questions, a demand characteristic inherent in interviewing a participant after administering SVF, lead to more clusters, clustered words, cluster switches (CS) and decreased hard switches (HS) on interview-based scoring procedures across low demand and high demand queries – evidence supporting a threat to internal validity inherent in Body & Muskett’s (2013) claim and methodology. Cueing three semantic categories (e.g., pets, sea creatures, farm animals) before administering SVF was also found to increase mean number of clusters, clustered words, CS and decrease HS compared to un-cued groups – evidence consistent with literature (Abwender, Preston, & Steffenella, 2003; Hurks, 2012) Additionally, HS on SVF traditional scoring measures correlated with total RFFT designs, while HS on SVF interview scoring measures correlated with total RFFT perseverations, suggesting the two scoring procedures are measuring different variables. No correlations were observed between CVLT-II clustering and SVF clustering using neither traditional scoring methods nor interview-based methods. We bring caution to the validity of interview-based methods employed by Body & Muskett (2013) as our preliminary study empirically supports a major threat to internal validity in their design.