|
Since 2002, 4 million visitors plus:
| |||||||
|
|
WIAT II Decile Scores to Standard Scores John Willis My students and I have been struggling with this question. The description of decile scores in the Process Assessment of the Learner (PAL) Manual (Virginia Wise Berninger, The Psychological Corporation, 2001, p. 109) is a little confusing to me, since it seems to cover 101% of the distribution (0 through 100).As far as I can determine, deciles seem to be based on percentiles, rather than percentile ranks, but the Manual (p. 109) says, "For each selected decile score, the percentage represents those students who obtained scores at or below the decile score," which sounds more like percentile ranks. The Manual (p. 109) also notes that, "technically, the deciles are the points separating the tenths of the distribution, not the tenths themselves," but calls the tenths "decile scores." Howard Lyman (Test Scores and What They Mean, 5th ed., Boston: Allyn & Bacon, pp. 104-105), usually a paragon of clarity upon whom I rely heavily, blames decile scores on the Institute for Personality and Ability Testing (IPAT) and says the first decile cutting point is at the tenth percentile, the second at the twentieth percentile, etc. [He also notes that R. B. Cattell did it differently. Decile 1 was the 5th to 15th percentile, which added a decile 0 (below PR 5) and put decile 10 at above PR 95, but we don't want to go there!] Lyman says, "In order to prevent confusion between decile and decile score (and to point out their similarity to percentile ranks), I prefer to use the term "decile ranks.'" So . . .Here's my guess, adapted from Table 5.4 (p. 109) in the PAL Manual:
This is what is known technically as a SWAG (Scientific Wild-Ass Guess). I'd appreciate corrections, as my students are becoming justifiably skeptical of my explanations. I cannot even prove my assertion that "decile" should be pronounced with a long i (as opposed to "dessill"). |