- Free Articles
- Extreme Value Theory International Encyclopedia of Statistical Science
- Reactive Scheduling of Batch Processes Encyclopedia of Optimization
- Variational methods in shape analysis Handbook of Mathematical Methods in Imaging
- History of Geomathematics: Navigation on Sea Handbook of Geomathematics
- Geomagnetic Field: Satellite Data Handbook of Geomathematics
**More Free Articles**

This is the free portion of the full article.
The full article
is available to licensed users only.

How do I get access?

# Parametric Versus Nonparametric Tests

A common distinction made with reference to statistical tests/procedures is the classification of a procedure as *parametric* versus *nonparametric*. This distinction is generally predicated on the number and severity of assumptions regarding the population that underlies a specific test. Although some sources use the term *assumption free* (as well as *distribution free*) in reference to nonparametric tests, the latter label is misleading, in that nonparametric tests are not typically assumption free. Whereas *parametric *statistical tests make certain assumptions with respect to the characteristics and/or parameters of the underlying population distribution upon which a test is based, *nonparametrictests* make fewer or less rigorous assumptions. Thus, as Marascuilo and McSweeney (1977) suggest, nonparametric tests should be viewed as *assumption freer* tests. Perhaps the most common assumption associated with parametric tests that does not apply to nonparametric