In this article, we describe how a latent variable modeling approach to the specification of measurement error unifies and benefits traditional methods of examining reliability in psychology and medicine. The models presented include classical reliability and generalizability theory to account for measurement error, latent class analysis to assess sensitivity and specificity, and item response theory to improve questionnaire development. We also illustrate how working with latent variables, in addition to addressing measurement error, may help deal with some instances of missing data. Throughout the article, analyses and results from examples and published articles are presented to illustrate the advantage of working with latent variables.
- latent variables
- measurement error
ASJC Scopus subject areas
- Psychiatry and Mental health
- Applied Psychology
- Arts and Humanities (miscellaneous)
- Developmental and Educational Psychology