Some Rank-Order Tests for Trend in a Set of Correlated Means,
Abstract
A statistic, J, has been proposed as a test for the differences between a set of n correlated means when m subjects have been subjected to n treatments. The null hypothesis is that all pormutations of the n scores for each subject are equally likely. The alternative hypothesis is that there is a trend which can be specified by the experimenter in the form of a theoretical rank-order. The J statistic is the sum of the S(d-squared) values that can be computed between the hypothetical rank-order and the observed rank-order for each of the m subjects. It is therefore equivalent to an average Spearman rank-order correlation. There is no parametric test that can be substituted for the non-parametric trend test.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1957
- Accession Number
- ADA074102
Entities
People
- Ardie Lubin
Organizations
- Walter Reed Army Institute of Research