Why are the point differences between predicted achievement and actual achievement that are needed for statistical significance so much smaller with the WIAT–II as compared to the original WIAT?

Frequently Asked Question:

Why are the point differences between predicted achievement and actual achievement that are needed for statistical significance so much smaller with the WIAT–II as compared to the original WIAT?

Answer:

Some of the difference in tabled values is the result of improved reliability of some of the WIAT–II subtests (the better the reliability of the subtest, the smaller the difference required to reach statistical significance). However, the majority of the difference is attributable to a change in one aspect of the statistical formula used to calculate statistically significant differences between predicted and actual achievement scores.

The change in formula is discussed in the WIAT–II Updated Examiner’s Manual on page 157, where it is stated: “Users of the WIAT will note the similarity of formulas presented here but also that the calculations of extreme discrepancies (The Psychological Corporation, 1992, p. 188) are no longer included, in favor of the more widely used standard error of the residual.”

“Calculation” or recognition of a “severe” discrepancy must still be based on statistical, as well as clinical significance. Base rate information, as included in the WIAT–II Manual and Scoring Assistants, help to document the rarity of discrepancies. When a discrepancy is reviewed for eligibility for special education services, all local and state special education regulations should also be considered.