Abstract
Neuropsychological tests are essential tools for evaluating dementia and related neurocognitive disorders, with their clinical utility determined mainly by their validity. This paper critically reviews the diverse evidence supporting the validity of neuropsychological tests in dementia assessment. Criterion validity is discussed in relation to the tests' ability to predict clinical diagnoses and underlying brain pathology, with a focus on their sensitivity to functional impairments and progressive neuropathological changes. Construct validity is explored through the lens of cognitive processes underlying test performance, using evidence from correlation structures and experimental paradigms. Furthermore, the paper examines the impact of emerging digital technologies on the evaluation of neuropsychological test validity, highlighting contrasts with traditional validation methods. The review identifies discrepancies between different types of validity evidence, emphasizing the need to contextualize validity within specific clinical and research applications. By addressing the conceptual limitations and trade-offs between validation approaches, this study proposes a comprehensive framework for interpreting validity evidence. Ultimately, it offers theoretical and practical implications for enhancing the robustness of neuropsychological tests in clinical practice and research.