Latent variables are variables that are not directly observed but are theoretically or empirically derived from observed variables. To use these latent variables we use methods such as the following:

  • Structural equation models
  • Multilevel models
  • Post-response models
  • Ordinal logistic regression
Latent variables are derived from variables that have been observed.

The research line Modelling latent variables has four sub-areas:

  • Item response theory
  • Structural equation modeling
  • Multilevel modeling
  • Logistic prediction model