Pandas and Penguins, Monkeys and Caterpillars: Problems of Cluster Analysis in Semantic Verbal Fluency

Abstract

Semantic verbal fluency, the most common example of which involves naming as many animals as possible in a specified time period, is one of the most widely used cognitive assessments both in research and work with clinical populations. One increasingly common way in which verbal fluency tasks are analysed is via the post-hoc categorisation of participants' responses into clusters by the experimenter. This qualitative process is seen as an important adjunct to simply counting the number of responses and is held to be a sensitive indicator of pathological cognitive processes and a window into semantic organisation. In this article, we examine the theoretical assumptions underlying cluster analyses and present data from description by 10 participants of their own performance on a semantic verbal fluency task. These qualitative descriptions cast doubt on (a) the assumption that semantic relationships can be reliably identified via external analysis and (b) the related assumption that clustering (results) can be taken as representative of specific features of semantic organisation and cognitive processes.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 8 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.