Analysing longitudinal data in the presence of missing responses with application to SLID data.

摘要

In longitudinal studies, outcomes that are repeatedly measured over time may be correlated and some may be missing. In this practicum, we empirically examine the performance of a recently proposed generalized quasi-likelihood (GQL) approach for the analysis of longitudinal data that includes observation that are missing completely at random (MCAR) or missing at random (MAR). This GQL approach is also illustrated by reanalyzing the Survey of Labour and Income Dynamics (SLID) data from Statistics Canada.