Research project: Models for longitudinal categorical data
Longitudinal data arise when repeated observations are made on the same unit, or subject, of observation.
Longitudinal data arise when repeated observations are made on the same unit, or subject, of observation.
Models and inference procedures need to be sensitive to the possibility of within-subject dependence between observations. Research at S3RI has focussed on modelling binary and ordinal categorical longitudinal data using a flexible family of association/marginal models, the parameters of which have a particularly straightforward interpretation.
Jokinen, J., McDonald, J. W. and Smith, P. W. F. 2006 Meaningful regression and association models for clustered ordinal data. Sociological Methodology, 36, 173-199.