No abstract

In this paper, we have attempted to study the behaviour of the family wise error rate (FWER) for Bonferroni's procedure and false discovery rate (FDR) of the Benjamini-Hodgeberg procedure for simultaneous testing problem with equicorrelated normal observations. By simulation study, we have shown that F.W.E.R. is a concave function for small no. of hypotheses and asymptotically becomes a convex function of the correlation. The plots of F.W.E.R. and F.D.R. confirms that if non-negative correlation is present, then these procedures control the type-I error rate at a much smaller rate than the desired level of significance. This confirms the conservative nature of these popular methods when correlation is present and provides a scope for improvement in power by appropriate adjustment for correlation.Recently multiple hypothesis testing under dependence has gained importance due to its increased relevance in modern scientific investigations. Although many efforts have been made to generalize the existing methods under dependence (Yekutieli and Benjamini (1999), Benjamini et al. (2001), Sarkar et al. (2002),Sarkar (2008), Efron (2007),Efron (2012) etc.), very few literature is available which explicates the effect of dependence on the existing methods.Correlation is one of the most important measure of dependence in the study of normal random variables. Although it is not an exhaustive measure of dependence, it might be a good starting point in order to study how dependence among observations affect the existing multiple testing algorithms. Efron (2010) in his study of empirical Bayes methods, has shown that, the correlation penalty depends on the root mean square (RMS) of correlations. An excellent review of the whole literature can be found in Efron (2012). However, we wanted to focus on the effect of correlation in a different manner. The traditional family-wise error rate (F.W.E.R.) and the false discovery rate (F.D.R.) are the 1

No abstract