Proceedings of the Annual Meeting of the Cognitive Science Society

Using a cued-recall paradigm, Experiment 1 had two conditions retrieve base situations in response to surface-dissimilar analogs: one in which participants were prompted to re-represent the base situations in general terms, and one in which they reproduced them verbatim. Although the prompted generalizations tended to be either too tied to the original situation or too broad so as to encompass disanalogous cases, they increased retrieval of the base analogs compared to the unprompted condition. Besides replicating the retrieval advantage of prompting one-shot generalizations, Experiment 2 addressed whether a self-administered training could aid participants in producing very general reformulations of base situations of the kind that, in Experiment 1, were associated with greater retrieval. Even though trained participants produced a greater proportion of highly abstract generalizations, the training yielded no retrieval advantage over and above the prompted abstraction condition. We discuss the implications of the present findings for transfer-oriented instruction.