Substitution, proration, or a retest? The optimal strategy when standard administration of the WPPSI-IV is infeasible

Abstract

Administration following a standard procedure is undoubtedly the ultimate goal of assessment. Practitioners face the challenge of unexpected events interrupting a formal evaluation, substitution, proration, or a retest, which may be used as alternative strategies to perform estimations. Such nonstandardized estimation strategies have introduced additional measurement errors because the manual reports only 1 set of norms, which are derived following the standard testing procedure. Thus, mismatching with norms is the root cause of additional errors introduced when applying these alternative strategies. This study examined the additional measurement errors introduced in using normative and clinical samples, as well as a retest sample of the 3 estimation approaches in the Wechsler Primary Preschool Scale for Children-Fourth Edition (WPPSI-IV). Results revealed that substituting, prorating, or retesting 1 of 6 core subtests for the Full Scale IQ increased the SEM by 0.61 to 1.92 score points (20-64%), generating wider confidence intervals (CIs) by ±1.2 to ±3.8 IQ points and misclassifications as high as 22%. Furthermore, substituting 1 missing subtest for the Verbal Comprehension Index or Processing Speed Index resulted in an additional 4.27 to 5.24 score point increase in SEM (79.7-132.5%), creating wider CIs by ±8.37 to ±10.27 IQ points and misclassifications as high as 35-44%. We remind practitioners to be cautious about the various increases in error rate resulting from such applications and to avoid introducing unnecessary additional errors when possible. Under circumstances when alternative procedures are considered a must, current work provides information for making evidence-based decision. (PsycINFO Database Record