ABSTRACT
In Chapter 2, we discussed the conceptual foundations and means of estimating the reliability of scores on assessment instruments. Each form of reliability calculation provides an estimate of the degree to which scores on an instrument are repeatable, reproducible, stable, or internally consistent. As we pointed out in Chapter 2, none of these terms speaks to the question of what the scores actually mean, a distinction often missed in selecting and interpreting the results from clinical assessment instruments. Most basically, the concept of validity refers to the question of what an instrument mea sures: Validity refers to the degree to which variation in scores on an instrument reflects variation in the psychological entity or process of interest. Therefore, it is best to think of reliability as a necessary precursor to validity (we also noted exceptions to the usual relation between reliability and validity, such as when we expect behavior to change over time).
