Reliability and Validity in Psychometric Assessments

This document is a technical report that discusses the concepts of reliability and validity in psychometric assessments, particularly in the context of hiring decisions. It outlines the distinction between reliability coefficients and validity coefficients, emphasizing that validity cannot be represented by a single numerical figure. The report details various methods to measure reliability, including temporal consistency, alternate-forms reliability, internal consistency, and generalizability theory. It explains that validity is a composite measure that requires substantial evidence to support the claims made by assessments. The document also describes five dimensions of validity, including test content validity and internal structure validity, and discusses the implications of these concepts on hiring practices. It concludes that psychometric assessments should comprise no more than 33% of the hiring decision, highlighting the importance of balancing these assessments with other candidate evaluations.