Latent state-trait analysis

Latent state-trait (LST) models allow researchers to separate person-specific sources of variability from situation-specific variance and random measurement error variance. In this chapter, I review the fundamental concepts of LST theory (Steyer et al., 2015) and discuss two prototypical LST models that can be used to determine consistency, occasion-specificity, reliability, and indicator-specificity in empirical data. I present an application to spatial ability test data from two measurement occasions (N = 519). In the discussion, I highlight extensions of LST approaches to other longitudinal research questions, including those related to the convergent validity of multiple raters, the question of variability versus trait change, and the study of specific fixed situations.