Title: Methodological issues in categorical data analysis (Friday 13:30-14:30).
Summary: Against the background of the Pearson-Yule discussion about the nature of categorical variables, several ways of viewing categorical variables will be discussed. Among other things, it will be argued that rating scales (Likert items) might perhaps better be treated as strictly categorical, using latent class outcomes as illustration. Finally, SEMs for categorical variables will be dealt with, paying extra attention to the ‘causal’ interpretation (and comparability) of loglinear and logit (logistic) coefficients.
Jacques A. Hagenaars is emeritus-professor in Methods and Techniques of Social Science Research, Department of Methodology, Faculty of Social Sciences, Tilburg University. He also has a professorship in IMPALLA-Advanced Master in Social Policy Analysis, Leuven University (KULeuven). His main research interests are in the area of categorical data analysis, longitudinal studies and latent variable models. He has published several books and a number of articles on these topics