Date of Conferral







Sandra Street


Cognitive tests are typically scored and interpreted using an appropriate normative reference group, often similar age individuals with similar levels of education. Psychometric testing theory presupposes that demographic correction is always beneficial, supporting the ubiquitous use of age and education correction in clinical practice. In the context of dementia, however, there is some evidence suggesting that demographic correction (specifically age correction) may reduce the sensitivity of cognitive tests to age-related cognitive decline. It was hypothesized that age correction would reduce the utility of cognitive tests for detecting cognitive change in individuals with mild cognitive impairment and mild dementia due to Alzheimer’s disease. This hypothesis was investigated using the National Alzheimer’s Coordinating Center (NACC) database. NACC data are contributed by the NIA-funded Alzheimer’s Disease Centers (ADCs). A series of hierarchical multiple linear regressions predicted the CDR® Dementia Staging Instrument Sum of Boxes Score (CDR-SB) from domain specific composite scores derived using different types of demographic correction (i.e., no correction, age correction, education correction, and both age and education correction). When looking at memory scores alone, raw scores captured more variation in the CDR-SB. However, when using a typical neuropsychological (NP) battery approach, correcting for education only produced a superior model. Findings may be used by clinicians for positive social change by recognizing that a diagnosis between normal cognitive aging and dementia is never determined by a single cut-off score in clinical practice, correcting for education is an essential component when processing standardized test scores.