Analysing Learning Strategies Through Microcomputer-Based Problem Solving Tasks

Abstract

Introduction Many different kinds of problem-solving tests have been studied by psychologists over the years and attempts have been made to classify them in terms of given and goal states (Beitman, 1965) and by typology (Greeno, 1978). Many of the problems which have been studied are those in which all the information required to solve the problem has been made explicit to the subject from the outset. Typical of such problems are Raven’s matrices, Towers of Hanoi, cryptarithmetic, and ‘river crossing’ problems. A much smaller group of problems is exemplified by ‘naughts and crosses’ as studied by Lawler (1985). In this group, the goal is to win against an adversary who is capable of changing the problem conditions in a non-predictable way. For such problems, the use of a microcomputer to present the problem could only be justified if the quantity or quality of information on the sequence of steps taken in the solution process (and recorded by the computer) would be more comprehensive or more reliable than could be inferred from other methods of recording the subjects’ activities.