Abstract

In recent years, many neutrosophic statistical methods have been proposed as generalizations of classical statistical methods and claimed to be superior when there is indeterminacy in the data values, the sample size, and/or the design parameters of the statistical method. We discuss several issues regarding these methods here. We argue that sample sizes and design parameter values selected by the practitioner are always known. Furthermore, the applicability and usefulness of neutrosophic statistical methods is questioned. In several examples we show that the results of neutrosophic analysis were either incorrect or not useful. We recommend a simpler and much more informative simulation-based method be used instead for the statistical analysis of interval data.