Inductive generalization (i.e., making generalizations from instances) is ubiquitous in human cognition. In the developmental literature, researchers have proposed two theoretical accounts of this important process: a na€ ıve theory account and a similarity-based account. However, neither of these theoretical accounts explains marked developmental changes in inductive generalization with familiar categories that have been documented in prior research. In this article, I describe briefly a revised version of the similarity-based account of inductive generalization that can explain individual variability as well as developmental change in inductive generalization with familiar categories. I also highlight several unresolved issues in the study of development of inductive generalization.
KEYWORDS-inductive generalization; category-based induction; conceptual development; semantic developmentInduction involves making generalizations from instances. It is a powerful and effective tool for generating knowledge that people use countless times daily. For example, we can predict that a never-before-seen dog barks and a novel ball bounces. Inductive generalizations are not always veridical-some breeds of dogs do not bark (e.g., Basenji)-but induction is a useful, if not foolproof, mode of thinking.