Nonparametric Hypothesis Testing Paper

Nonparametric Hypothesis Testing

Joey Bunnell, Darlene Aceves, Jagbir Dhanda, Russell Fivecoate, Lawrence Fee

RES/342

January 23, 2012

Jack McNicholas

Nonparametric Hypothesis Testing

Nonparametric statistical procedures rely on no or few assumptions about the shape or parameters of the population distribution from which the sample was drawn (Walsh, 1962). Within this paper team C will be conducting several different nonparametric tests in order to conclude what variables affect the amount that a person makes at a particular business. In order to conduct this research, team C will develop a hypothesis for this research and then perform a five-step hypothesis test. Once the test has reached a conclusion, team C will then interpret the results and discuss what the results show. These results will then be compared to the findings of the tests conducted in week 3 by team C.

As a team we decided to use the Kruskal-Wallis non-parametric test. “It is used for comparing more than two samples that are independent, or not related. The parametric equivalence of the Kruskal-Wallis test is the one-way analysis of variance (ANOVA).” (McDonald, J.H) There is not much calculating involved, and a quick answer to our question could be granted with this non-parametric test. Unlike many standard parametric tests there are not many assumptions and conditions that need to be met for a non-parametric test to follow through. “This method provides an air of objectivity when there is no reliable, which is universally recognized, underlying scale for the original data and there is some concern that the results of standard parametric techniques would be criticized for their dependence on an artificial metric.” We did not want inaccurate data for the p-value. The Kruskal-Wallis test will allow for the use of one nominal variable and one measurement variable. We can compare to see if two non-related samples are identical with this test. This is also done while assuming that they are...