Validscale: A command to validate measurement scales

Abstract

Subjective measurement scales are used to measure nonobservable respondent characteristics in several fields such as clinical research, educational sciences, or psychology. To be useful, the scores resulting from the questionnaire must be validated; that is, they must provide the psychometric properties validity, reliability, and sensitivity. In this article, we present the validscale command, which carries out the required statistical analyses to validate a subjective measurement scale. We have also developed a dialog box, and validscale will soon be implemented online with Numerics by Stata.