Now you know it, now you don’t: Asking the right question about category knowledge

Nolan ConawayBinghamton University
Kenneth KurtzBinghamton University

Abstract

A key goal of category learning research is to describe how categories are represented. Essential to this research are measures that provide investigators insight into exactly what learners have gained from their training experience. In this paper, we review and explore three commonly used measures: A) ease of acquisition, B) generalization, and C) single feature classification. We report results of a category learning experiment in which these measures are compared side-by-side. We find that generalization and single feature classification data are the more informative measures; we also find a novel inconsistency between them. Specifically, many learners who generalize based on only a single dimension demonstrate robust knowledge of both dimensions during the single feature classification test. We discuss implications for methodology in the field, as well as for selective attention and theories of human category learning.

Files

Now you know it, now you don’t: Asking the right question about category knowledge (448 KB)



Back to Table of Contents