Comparing Test Scores Using Information From Criterion-Related Validity Studies

摘要

There is frequently a need to compare a client's test scores from different instruments. If the scores come from instruments that use the same scale, it is tempting to compare the scores directly. Unfortunately, this method can lead clinicians to believe that there is a large difference between scores when the difference is minimal. As an alternative, we outline a method for score comparison that uses information from criterion-related validity studies. Using three examples, we show why this method is more psychometrically sound, produces more accurate comparison scores, and requires little extra work for clinicians than the direct comparison approach. To make the score comparison process easy for clinicians to use, we include an appendix that demonstrates how to implement this method in Microsoft Excel and the free R program.