Analysis of Clustered Binary Data in Intervention Studies under Completely Randomized Design.

Comparative trials in which interventions are randomized to a group of individuals (clusters) are gaining importance due to its specific advantages. The data gathered in group or cluster tend to exhibit intra correlation coefficient which the analysis must be addressed in order to draw valid inference. The usual statistical methods are not appropriate for this design, since they do not take into account of dependency among subjects within the clusters. Recently, several methods are being evolved to compare event rate for clustered binary data. In this paper, an attempt is made to compare these methods using real life data set. It is demonstrated that the adjusted chi-square method and pooled adjusted chi-square method are appropriate in analysing the clustered binary data.