A Tutorial on GEE with Applications to Diabetes and Hypertension Data from a Complex Survey
Correlated data frequently arise from cross-sectional studies with complex cluster design because individuals from the same cluster or region share some common characteristics. Analyzing correlated data using standard statistical methods, which are applicable for independent data, may produce misleading inference. This article reviews the GEE and its software implementations and provides some guidelines for using it in practice. To illustrate GEE, data from the 2011 Bangladesh Demographic and Health Survey, a two-stage complex cluster survey have been used to identify the risk factors for diabetes and hypertension. The results suggest that age, current working status, education, socioeconomic status, and body mass index are significantly associated with hypertension and diabetes. Further, we found significant positive correlation between the responses from the same cluster, justifying the use of GEE.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Submission of any work for publication in this journal would imply that the authors acknowledge that the work is their own and that they have taken all necessary permissions for all the materials used in their work.
Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-NonCommercial License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
Authors permit us for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.