Regression models for analyzing clustered binary and continuous outcomes under the assumption of exchangeability
E. Olusegun George, Dale A. Bowman, Qi An · 2013
The advent of sophisticated tools of measurement has given rise to new modes of data collection. As a result, data often come with complex dependence structures. These complex structures typically require non-standard statistical approaches that usually entail computationally intensive methodologies. Conventional tools generally rely on the assumption that the data, or some suitable transformations of them, follow a normal distribution. This assumption no longer directly applies in these contexts. Over the past 20 years, there have been remarkable developments in statistical methodology for the analysis of such data. The development of statistical software and packages has unfortunately not kept pace with these methodological advances, but practitioners nonetheless now have a host of increasingly sophisticated tools available to them for handling the complex data. This has made possible their adoption and application in solving important substantive problems across a number of disciplines, particularly in engineering and finance, and in medicine and health.