{"id":106,"date":"2016-09-04T21:57:26","date_gmt":"2016-09-04T21:57:26","guid":{"rendered":"http:\/\/jsr.isrt.ac.bd\/?post_type=article&p=106"},"modified":"2016-09-06T02:17:34","modified_gmt":"2016-09-06T02:17:34","slug":"marginal-models-for-binary-longitudinal-data-with-dropouts","status":"publish","type":"article","link":"http:\/\/jsr.isrt.ac.bd\/article\/marginal-models-for-binary-longitudinal-data-with-dropouts\/","title":{"rendered":"Marginal models for binary longitudinal data with dropouts"},"content":{"rendered":"
In this paper, we propose and explore a set of weighted generalized estimating equations for\u00a0fitting regression models to longitudinal binary responses when there are dropouts. Under\u00a0a given missing data mechanism, the proposed method provides unbiased estimators of\u00a0the regression parameters and the association parameters. Simulations were carried out to\u00a0study the robustness properties of the proposed method under both correctly specified and\u00a0misspecified correlation structures. The method is also illustrated in an analysis of some\u00a0actual incomplete longitudinal data on cigarette smoking trends, which were used to study\u00a0coronary artery development in young adults.<\/p>\n