Interpreting Neuropsychological Test Discrepancies: An Error-Adjusted Approach

Introduction:Differences between neuropsychological test scores are often interpreted using simple thresholds based on standard deviation units. Although widely used, these approaches do not explicitly account for measurement error and may lead clinicians to infer the presence of deficits where none exist. Method: This paper describes an error-adjusted approach in which observed differences are evaluated relative to the expected measurement error of the tests involved. Drawing on classical test theory and methods already used within existing test batteries, the standard error of the difference is derived from the standard errors of measurement of the two tests. The size of the observed discrepancy is then expressed relative to the size of the shared error. Results: The resulting error-adjusted difference (EAD) provides an index of the size of a discrepancy relative to measurement error. A worked example illustrates how interpretation of the same difference varies as a function of test reliability. Conclusions: The method is intended as a practical complement to existing approaches, particularly in contexts where normative or regression-based comparisons are not readily available.