Educational Software—It isn't Always

Abstract

It is commonly assumed that well-designed instructional software for problem-solving will be effective in promoting learning. Unfortunately, many computer-aided instructional systems have fallen short of this goal. This article describes a laboratory study to investigate this issue of effectiveness in the case of a problem-solving tool that was designed to assist students in solving non-linear optimization problems. It is shown that manual procedures resulted in greater learning than with the use of the computer-based tool. Implications of these experimental results are discussed, and it is concluded that empirical testing of instructional software is required before adoption for classroom use.