Although many of you may not be interested in the psychometric details of the ORS and SRS, it does bear importantly on whether there are seen as credible. Jeff Reese and I (Duncan & Reese, 2013) recently exchanged views with Halstead, Youn, and Armijo (2013), debating when a measure is too brief and when it is too long. Here is our paper. First regarding when a measure is too brief: There is no doubt that 45 items, 30 items, or even 19 items is psychometrically better than 4 items, and that the increased reliability and validity of longer measures likely result in better detection, prediction, and ultimate measurement of outcome. But how much better is the really the question. Are these differences clinically meaningful and do they offset the low compliance rates and resulting data integrity issues from missing data? These are the questions that require empirical investigation to determine how brief is too brief, although from my experience, the verdict has already been rendered. But when is a measure too long? The answer is simple: When clinicians won’t use it.
Posted on: Wed, 14 Aug 2013 17:25:09 +0000