| Abstract: | Tested predictions of a structure-mapping model for word problems in 4 experiments involving 132 undergraduates. In Exps I and II, Ss rated the potential usefulness of solutions for pairs of problems—mixture problems in Exp I and work problems in Exp II. The problems were either equivalent (same story, same procedure), similar (same story, different procedure), isomorphic (different story, same procedure), or unrelated (different story, different procedure). Ss in Exp III used an example solution for a work problem and a mixture problem to generate equations to related test problems that differed in their mappings from the example. In Exp IV, Ss matched concepts in the test problems to corresponding concepts in the examples to provide a direct measure of their ability to construct mappings across different problems. In Exps III and IV, Ss performed significantly better on isomorphic problems than on similar problems, and significantly better on work isomorphs than on mixture isomorphs. Results suggest that a structure-mapping model that emphasizes the transparency and structure of the mapping can be used to predict the usefulness of a solution. (19 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
