Abstract
We used a 2 × 2 design to compare the performance of low-aptitude remedial participants learning to solve mathematics word problems. Participants worked either alone or as a member of a collaborative dyad, and received up to twelve hours of instruction and practice using either the Word Problem-Solving Environment (WPSE), an exploratory system over which users exercise considerable control, or Solver, a very structured and sometimes directive tutor. Individuals who worked with the WPSE showed the greatest average improvement between pretest and posttest scores. Dyad members showed relatively moderate improvements after working with either the directive tutor or the problem-solving environment. There were no significant changes between pretest and posttest scores for individuals who worked with the directive tutor. We discuss possible reasons for this pattern of results and present an analysis of data concerning the number of operations used in solving problems, the number of errors made, and the number of help requests, to support our discussion.
Get full access to this article
View all access options for this article.
