Assessment of Dispersion in Categorical Data

Four separate approaches to the problem of assessing variation in categorical data are developed, each of which results in an identical index of dispersion (D). An unbiased sample estimator of D is pre sented, which bears an analogy with the common unbiased sample estimator of variance for quantitative data. An additional index, re lated to a measure of entropy for categorical data, is mentioned and some applications are discussed.