Switch it up: Learning Categories via Feature Switching


This research introduces the switch task, a novel learning mode that fits with calls for a broader explanatory account of human category learning (Kurtz, 2015; Markman & Ross, 2003; Murphy, 2002). This paper presents the switch task to further explore the contingencies between learning goals, learning modes, outcomes, and category representations. Given that the ability to switch items between categories nicely encapsulates category knowledge, how does this relate to more familiar tasks like inferring features and classifying exemplars? To address this question we present an empirical investigation of this new task, side-by-side with the well-established alternative of classification learning. The results show that the category knowledge acquired through switch learning shares similarities with inference learning and provides insight into the processes at work. The implications of this research, particularly the distinctions between this learning mode and well-known alternatives, are discussed.

Back to Table of Contents