ERIC Number: ED463299
Record Type: Non-Journal
Publication Date: 2002-Apr-11
Reference Count: N/A
An Empirical Study of Eight Nonparametric Tests in Hierarchical Regression.
Harwell, Michael; Serlin, Ronald C.
When normality does not hold, nonparametric tests represent an important data-analytic alternative to parametric tests. However, the use of nonparametric tests in educational research has been limited by the absence of easily performed tests for complex experimental designs and analyses, such as factorial designs and multiple regression analyses, and limited information about the properties of these tests for realistic data conditions. Efforts to remedy this deficiency have begun with the introduction of general linear model-based nonparametric tests. The results of a computer simulation of the properties of several of these tests in hierarchical regression analyses indicated that, on balance, the top-performing test for the nonnormal distributions studied was the McKean-Hettmansperger F-test (J. McKean and T. Hettmansperger, 1976) using a confidence interval estimate of the scale parameter tau. The Serlin-Harwell aligned-rank chi-square test (R. Serlin and M. Harwell, 2001) performed almost as well, which, combined with the fact that it is easier to compute, makes it an attractive competitor to the Mc-Kean-Hettmansperger test. (Contains 5 tables and 59 references.) (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Authoring Institution: N/A