Northern Prairie Wildlife Research Center
Statistical Sirens: The Allure of Nonparametrics
Douglas H. Johnson*
Abstract: Although nonparametric statistical methods have a role
to play in the analysis of data, often their virtues are overstated and their
deficiencies overlooked. A recent Special Feature in Ecology advocated
nonparametric methods because of an erroneously stated advantage that they require
no assumptions regarding the distribution underlying the observations. The present
paper points out some often-ignored features of nonparametric tests comparing
two means, and advocates parameter estimation as a preferred alternative to
hypothesis testing in many situations.
Key words: Hypothesis testing; nonparametric methods; normal
distribution; parameter estimation; Wilcoxon-Mann-Whitney test vs. t
test.
This resource is based on the following source (Northern Prairie Publication
0934):
Johnson, Douglas H. 1995. Statistical sirens: the allure of
nonparametrics. Ecology 76(6):1998-2000.
This resource should be cited as:
Johnson, Douglas H. 1995. Statistical sirens: the allure of
nonparametrics. Ecology 76(6):1998-2000. Jamestown, ND:
Northern Prairie Wildlife Research Center Online.
http://www.npwrc.usgs.gov/resource/methods/nonparam/index.htm
(Version 05OCT2000).
Table of Contents
* Northern Prairie Science Center, National Biological Service, Jamestown,
North Dakota 58401 USA
Downloading Instructions
-- Instructions on downloading and extracting files from this site.
nonparam.zip
( 22K) -- Statistical Sirens: The Allure of Nonparametrics
Installation: Extract all files and open
index.htm in a web browser.