Below is a response to my prior post regarding Wechsler-like scaled score issues. The response was on the NASP listserv and the authors gave me permission to reproduce it "as is" below. I'm pleased that they concur with the recommendations at the end of the paper post.
Larry Weiss
Mark Daniel
Pearson
Technorati Tags: psychology, school psychology, forensic psychology, clinical psychology, neuropsychology, Wechsler batteries, WJ III, scaled scores, standard scores, psychometrics
Kevin McGrew's argument can be turned around to show that using subtest score metrics with larger SDs also may lead to misinterpretation if a change of 1 raw score point leads to a change of many standard score points. So, the issue is not as simple as which subtest metric is better (e.g, the Wechsler / Kaufman metric or the WJ metric). The issue is better framed in terms of making the right choice of metric based on how it fits with the underlying RS distribution. Appropriate fit between the RS and SS distributions is necessary to avoid
misinterpretation due to SS metrics that are either too large or small.
We agree with his suggested guidelines at the end of the full paper.
Larry Weiss
Mark Daniel
Pearson
Technorati Tags: psychology, school psychology, forensic psychology, clinical psychology, neuropsychology, Wechsler batteries, WJ III, scaled scores, standard scores, psychometrics
