The 7th Day of the LCSR Summer School: Unconditional Latent Curve Models

On Tuesday, Prof. Kenneth Bollen discussed unconditional latent curve models. Unconditional models do not include any explanatory variables that predict a trajectory. The model includes both fixed and random effects for the intercept and the slope. Fixed effects reflect mean values of a dependent variable in the starting time point and mean values of the slope estimated on the basis of the whole sample. In contrast to fixed components, random effects reflect individual deviations from the mean intercepts and slopes. This logic is similar to that of multilevel models. However, latent curve models are estimated from a structural equation perspective.

Moreover, Prof. Kenneth Bollen explained model identification problems, estimation methods, goodness-of-fit criteria, and competitive advantages of the method as compared to parameter estimation using regression models for each sample unit.