Statistical Hypothesis Testing with SAS and R

Chapter 12

Tests on other Distributions

In this chapter we present goodness-of-fit tests for distributions other than Gaussian. Tests in Section 12.1 are based on the empirical distribution function (EDF). Section 12.1.1 deals with Pearson's -test, which is an omnibus test for goodness-of-fit but not so powerful if alternative tests on specific distributions are available.

12.1 Tests based on the EDF

12.1.1 Kolmogorov–Smirnov test

Description:	Tests if a sample is sampled from a specific distribution function .
Assumptions:	Data are at least measured on an ordinal scale. The random variables are independently distributed with observations and a continuous distribution function .
Hypotheses:	(A) vs
for at least one
(B) vs
with for at least one
(C) vs
with