Proceedings of the Annual Meeting of the Cognitive Science Society

Recent work suggests that classification training and observational learning may differ with regard to the benefitsof different types of item presentation. In particular, there is evidence that between-category contrast is most helpful fortraditional classification learning of feature-based categories, while a supervised observational mode promotes learning ofrelational categories via within-category comparison. The purpose of this study is to begin to tease apart the role of the learningmode versus the type of category in producing this pattern of results by replicating an earlier study that used classificationtraining and feature-based categories, and adding in observational learning conditions. If under these conditions, contrast isbeneficial for both learning modes, it will suggest that the type of category being learned is the key to the previously observeddifference. If, on the other hand, that same difference is observed, it will suggest that the cause is the learning mode itself.