Bringing Problem-Solving Transfer to the Science of Instruction

• In Event: The Form and Function of Transfer in Diverse Contexts

Abstract

Problem-solving transfer refers to being able to use what was previously learned to solve new problems. Problem-solving transfer is a fundamental educational goal. The goal of this presentation is to review examples of instructional studies that assess learning outcomes by using measures of retention and problem-solving transfer, in order to better understand the role of measures of problem-solving transfer in educational research.

Consider three examples. First, some students learn the statistical concept of binomial probability by reading a formula-based lesson that starts with the formula and emphasizes how to apply it (deductive group), and other students learn by reading a concept-based lesson that begins with familiar concepts and leads up the formula (inductive group). On a retention test involving solving problems that were very similar to those in the lesson, the deductive group performed better, whereas on a transfer test involving recognizing that a problem was unanswerable (e.g., “N = 2, r = 3, p = ½, Find binomial probability”) or answering a question about a problem situation (e.g., “Can r be greater than N?”), the inductive group performed better (Mayer & Greeno, 1972).

Second, some students read a non-signaled lesson on how red tides develop (non-signaled group), and other students read a signaled version of lesson that contained an outline, headings, and pointer words (signaled group). On a retention test involving recall of the factual details in the lesson, the non-signaled group performed best, whereas on a transfer test involving solving problems about red tides (e.g., generating ideas for how to prevent red tides), the signaled group performed better (Loman & Mayer, 1983).

Third, some students read a text passage that explained how a car’s braking system works (no-illustrations group), and other students read the same text passage along with corresponding illustrations showing the steps in the process (illustrations group). On a retention test involving recall of the factual material, the no-illustrations group performed best, whereas on a transfer test involving solving problems about brakes (e.g., generating reasons for why a braking system might fail), the illustrations group performed best (Mayer, 1989). Additional examples will be presented in the presentation.

When researchers assess learning outcomes mainly by focusing on measures of retention, the results may show that instructional features intended to promote learning are not effective. When researchers expand the assessment of learning outcomes to include measures of problem-solving transfer, the results may show important differences in learning caused by different instructional methods. Overall, to the extent that a goal of education is to enable students to use what they have learned in new situations, learning outcomes should be assessed using valid and reliable measures of problem-solving transfer.

Author

• Richard E. Mayer, University of California - Santa Barbara