Beyond Dichotomies in Generalization Research: A Reply to Lee and Schlegelmilch (2025)

Abstract

Lee and Schlegelmilch (2025) reanalyzed data from Zaman, Yu, and Verheyen (2023), arguing that the role of perception in generalization is overemphasized and that higher level cognitive processes (in the form of a similarity-based rule) provide a better account. In this reply, we make the argument that their reanalysis contains flaws and inconsistencies and present additional evidence for consideration. While acknowledging the role of higher level cognition in generalization, we maintain that the available evidence supports stimulus perception and memory as important predictors. We view this exchange as a reminder that stimulus representation and higher level cognition exist on a spectrum, with their relative importance varying across contexts. Consequently, more research into the interactions between these processes is necessary. We argue that rather than adhering to existing theoretical frameworks, researchers should continue to challenge our understanding of human generalization by incorporating new evidence and insights into existing models and comparing them across diverse contexts. This approach will foster a more comprehensive and nuanced understanding of human generalization processes.