In this chapter we present goodness-of-fit tests for the Gaussian distribution. In Section 11.1 tests based on the empirical distribution function (EDF) are treated. A good resource for this kind of test is Stephens (1986). We start with the Kolmogorov–Smirnov test. It evaluates the greatest vertical distance between the EDF and the theoretical cumulative distribution function (CDF). If both, or one parameter are estimated from the sample the distribution of the test statistic changes and the test is called the Lilliefors test on normality.
Section 11.1.1 deals with tests not based on the EDF such as the Jarque–Bera test which compares observed and expected moments of the normal distribution.
|Description:||Tests if a sample is sampled from a normal distribution with parameter and .|