This paper deals with treatment of missing data in some multidimensional testing problems. Cases in which missing data are not missing at random are studied. In this context, to obtain valid parametric inferences, the process that causes missing data must be specified. Through a nonparametric approach some problems of missing data not missing at random are discussed and characterized. The proposed solution is based on a nonparametric combination of dependent permutation tests, not requiring any specification of non-response model.
CITATION STYLE
Giraldo, A., Pallini, A., & Pesarin, F. (1994). A Nonparametric Testing Procedure for Missing Data. In Compstat (pp. 503–508). Physica-Verlag HD. https://doi.org/10.1007/978-3-642-52463-9_62
Mendeley helps you to discover research relevant for your work.