Nonparametric tests (distribution free tests) are
statistical tests, which can be used with data and with skewed distributions at
the nominal or ordinal level of measurement (Weaver, 2002). The reason they are
also called distribution free tests is because the assumptions underlying their
use are weaker and fewer than the assumptions associated with parametric tests
(Weaver, 2002). Nonparametric tests barely require any assumptions concerning
the shapes of the underlying population distributions (Weaver, 2002). Which is
why nonparametric test are used instead of parametric tests if the assumptions
of the parametric test have been grossly violated (Weaver, 2002). Nonparametric
test are also referred to as rank-order tests because rank or ordinal data
usually requires nonparametric analysis. Researchers use nonparametric tests
when certain assumptions cannot be made about the population and when the type
of data is ordinal in nature and not at least interval. A psychological
research situation where the effects of deliberately induced stress on
participants could be ranked according to how fast each participant recovers to
a normal stress free state (first, second third, and so forth) and ranked on
how accurately they perform test right after stress is induced (first, second
third, and so forth).

## Monday, March 18, 2013

Subscribe to:
Post Comments (Atom)

## No comments:

## Post a Comment

Note: Only a member of this blog may post a comment.