COMPARABLE MEASUREMENT SCALES IN TASK-COMPARISON EXPERIMENTS

The primary goal of this comment on the article by Schacter, Cooper, and Delaney (1990) is to illustrate two kinds of advantages of comparable measurement scales for task-comparison experiments. First, a comparable metric allows quantitative comparisons of the degree of accurate performance across tasks, which yields more information than the qualitative conclusions from tests of significance on noncomparable measurement scales. Second, such a scale can highlight situations in which fundamentally different kinds of comparisons are being attempted. A few remarks about the inferences drawn from the research are also offered.