Title: Testing normality of a large number of populations
Authors: Maria Dolores Jimenez-Gamero - Universidad de Sevilla (Spain) [presenting]
Abstract: Suppose we have independent samples coming from $k$ populations. We consider the problem of (simultaneously) testing that the populations are normal. The means and the variances may be different. The number of populations is allowed to increase to infinite. It may be even larger than the sample sizes. In our developments, it is assumed that the data is univariate in order to keep the notation simple. Nevertheless, the results also apply (with minor modifications) to multivariate data.