Latent Covariates in Generalized Linear Models: A Rasch Model Approach

摘要: Study of multivariate data in situations where a variable interest is unobservable (latent) and only measured indirectly widely applied. Item response models are powerful tools for measurement have been extended to incorporate latent structure. The (log-linear) Rasch model simple item tests fit parameter estimation can take place without assumptions about the distribution variable. Inclusion as predictor standard regression such logistic or Poisson discussed, study relation between psychosocial work environment absence from used illustrate motivate results.