Training Learners to Self-Explain: Designing Instructions and Examples to Improve Problem Solving

Abstract

In this experiment, we integrated two learning methods – subgoal learning and constructive learning – to explore their interactions and effects on solving computer programming problems. We taught learners to solve problems using worked example and practice problem pairs with one of three kinds of instructional design that either did not highlight the subgoals, described the subgoals, or prompted participants to describe the subgoals for themselves. In addition, we varied the distance of transfer between the worked example and practice problem pairs. We found that instructions that highlighted subgoals improved performance on later problem solving tasks. The groups that performed best were those that received subgoal descriptions with farther transfer between examples and practice problems and those that described subgoals for themselves with nearer transfer.