Dear list participant,
My questions (two) are about Ridit Analysis. (See below for some brief ridit information). I want to use ridit analyses to analyse differences between groups on an ordinal scale.
Question 1.: when I use ridit scores, is it appropriate to us Kruskall-Walis / Mann Whitney to assess statistical significance, or are there other, more appropriate, options for use?
Question 2.: is there a statistical application to compute ridit analyses and related P-values availible on line?
Thank you for your time and trouble.
Wolter Paans
HanzeUniversity, Groningen
Netherlands
[log in to unmask]
(Brief Ridit information)
Ridit means Relative to an Identified Distribution (Bross 1958). The use of ridits implies the selection of a reference distribution, since phenomena measured on an ordinal level are always relative in nature. The riditscore for a given category stands for the proportion of all subjects from the reference group falling in all categories of lower rank plus half the proportion falling in the given category. The ridit of the reference group itself is, by definition, 0.500. Therefore, if two subjects are randomly selected from the same population, then the second subject will have a higher score than the first subject half the time, and will have a lower score half the time (Sermeus & Delesie, 1996). The average ridit, usually calculated for the reference group, has a probability interpretation; it is an estimate of the chance that an individual in a given category has a lower score than an individual in the reference group.
|