Normality Test

A normality test is used to determine whether sample data has been drawn from a normally distributed population (within some tolerance). Normality is a prerequisite of several statistical tests, such as the student t-test and the 1-way or 2-way ANOVA, which require a normally distributed sample population. If the assumption of normality is not valid, the results of the tests will be unreliable.

Six different normality tests are available in Origin

Normality Test	Summary
Shapiro-Wilk	Common normality test, but does not work well with duplicated data or large sample sizes.
Kolmogorov-Smirnov	For testing Gaussian distribution with specific mean and variance.
Lilliefor	Kolmogorov-Smirnov test with corrected P. Best for symmetric distributions with small sample sizes.
Anderson-Darling	Can give better results for some datasets than Kolmogorov-Smirnov.
D'Agostino-K Squared	Based on transformations of sample kurtosis and skewness. Especially effective for ?non-normal? values.
Chen-Shapiro	Extends Shapiro-Wilk test without loss of power. Supports limited sample size (10<=200).

The Normality Test Dialog Box

Examples

Algorithm

To determine a normality test for a column or range of values:

Select Statistics: Descriptive Statistics: Normality Test. This opens the NormalityTest dialog box. You can specify any of the six test methods in the Quantities to Compute branch. Box charts and histograms may also be produced.
Upon clicking OK, a report table sheet is created showing the basic summary statistics, degrees of freedom, test statistic, associated p-value, and the test conclusion.