ABSTRACT
Classical tests of statistical hypotheses are often based on the assumption that the populations sampled are normal. In addition, these tests are frequently confined to statements about a finite number of unknown parameters on which the specification of the probability distribution function (pdf) of the random variable under consideration depends. Efforts to eliminate the necessity of restrictive assumptions about the population sampled have resulted in statistical methods called nonparametric, directing attention to the fact that these methods are not limited to inferences about population parameters. These methods are also referred to as distribution free, emphasizing their applicability in cases in which little is known about the functional form of the pdf of the random variable observed. Non-parametric tests of statistical hypotheses are tests whose validity generally requires only the assumption of continuity of the cumulative distribution function (cdf) of the random variable involved.
