Overview of Modern Measurement Theory and Examples of Its Use to Measure Execution Function in Children

Researchers and clinicians in neuropsychology are interested in measuring constructs that cannot be directly observed, such as cognitive ability, executive function, or depression. Modern measurement theory (MMT) provides a framework for understanding and measuring these constructs (also called latent variables) through the use of test batteries and other assessments that provide observable responses that are theoretically related to the underlying construct of interest. MMT subsumes a number of statistical methods, including factor analysis and item response theory (IRT), that are focused on using observable responses to model these unobservable constructs. These methods are able to provide information on how test batteries or individual items relate to each other and the latent variable across varying levels of the construct of interest (e.g., mild ADHD and severe ADHD). Finally, we review several applications of IRT to the measurement of executive function in children to highlight the unique information IRT provides when evaluating the relationship of observed responses to unobservable constructs. In summary, this paper seeks to provide an accessible overview of latent variable models within MMT (with a focus on IRT) that could improve measurement in neuropsychology, if more widely used, and offer readers the foundations of a knowledge base to become more critical consumers of measurement-related research in pediatric neuropsychology.