Generalized Method for Rank Determination in Rank-Ordered Statistics

This paper proposes a generalized and structured method for use for rank determination in rankorder statistics.The sampled populations may be measurements on as low as the ordinal scale and need not be continuous or numeric.The proposed method would readily enable the researcher assign ranks to sample observations without the need to first arrange the observations in some form as is often the case with the traditional approach in the ranking of observations.The method also provides expressions that are intrinsically and structurally formulated to enable one easily break ties and assign appropriate ranks to any tied observations if need be in a straight forward manner.The proposed method is illustrated with some sample data and shown to be often easier to use in practice than the traditional method and of more generalized and wider applicability than some other existing formulations which are often limited in their use.