Schema-Based Theories of Problem Solving
Abstract
The objective of this research is to construct a schema-based model of problem solving to represent construction of equations for solving algebra word problems. The research summarized in this report is concerned with the selection, use, and description of instructional examples. Experiment 1 shows that mathematical experience was beneficial for improving the selection of good analogies when the analogies are isomorphic to the test problems, but was not beneficial when the analogies are more inclusive than the test problems. In Experiment 2 students were able to effectively combine information from two analogous problems but did significantly worse when combining information form one example and a set of procedures. The last three experiments required that students categorize motion problems according to whether the two distances in a problem should be equated, added, or subtracted. Categorization significantly improved as the number of training examples representing a category increased from one to four (Experiment 3). Categorization was also significantly better when students received both specific and general descriptions of the examples than when they received only a single description (Experiment 4). However, as shown in Experiment 5, students were unable to form their own general descriptions by comparing similar examples. (RRH)
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1989
- Accession Number
- ADA216717
Entities
People
- Stephen K. Reed
Organizations
- San Diego State University