Sign Tests
E. Jacquelin Dietz
Abstract
E. Jacquelin Dietz
Abstract
Abstract The sign test is a simple distribution‐free test that can be applied easily in a variety of situations. Observations, which may be difference scores, are replaced by their positive or negative signs and analyzed using a binomial test. The specific applications discussed here are a location test for the median, the Cox–Stuart test for tend, and the McNemar test for correlated proportions.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Abstract The sign test is a simple distribution‐free test that can be applied easily in a variety of situations. Observations, which may be difference scores, are replaced by their positive or negative signs and analyzed using a binomial test. The specific applications discussed here are a location test for the median, the Cox–Stuart test for tend, and the McNemar test for correlated proportions.
Key concepts: McNemar's test, Sign test, Test (biology), Statistics, Sign (mathematics), Mathematics, Negative binomial distribution, Wilcoxon signed-rank test