New Statistical Techniques for Evaluating Longitudinal Models

Publisher Website
Google Scholar
Abstract
In this paper the statistical problems of longitudinal data are discussed and some new work on model building is reported. The models and methods are applicable to any part of the life cycle. As illustrative, a model is presented for treating qualitative data and for validating a stage theory of cognitive development in the child. Another illustration is a model for treating quantitative data and for disentangling antecedent-consequent relationships as in the disengagment hypothesis that psychological withdrawal occurs before social withdrawal in aging persons. The general qualitative model is conceptually related to Lazarsfeld’s latent structure analysis; the general quantitative model is regarded as a covariance structure model. The models are built to carry out confirmatory analysis of developmental theories in ways not previously available.

This publication has 0 references indexed in Scilit: