6.0. Overview

• One Sample Tests
• Two Sample Tests
• Tests with Paired Data
• Multisample Tests
• Tests with Matrix Data
• Tests with Binary Data
• 2 x 2 Tables
• R x C Tables

A wide range of parametric and nonparametric tests, correlation coefficients and statistical tables are covered in this chapter. Most test statistics are displayed with their probability values and, where relevant, with confidence intervals. Output also includes summary statistics about the columns or groups of data used in the test. Contingency Table, Cross-Tabulation and Break-Down analyses can be performed with an unlimited number of factors.

In this chapter of the User’s Guide formulas used in tests will be given where possible. The following notation:

n1, …, nk, M1, …, Mk, s1, …, sk 

will stand for the number of observations, means and standard deviations of samples 1, …, k respectively. In one sample tests n stands for the sample size and in tests with more than one variable n will stand for the sum of all valid cases in all samples. For instance, in a two sample test n = n1 + n2.

All tests can handle missing observations. The tests that do not require paired or matrix data will ignore only the individual missing values. Tests on paired data will omit a pair if either or both of the observations in a pair are missing. Tests that require data in matrix format will omit any rows containing one or more missing observations.

The data format for tests covered in this chapter will be one of the following types.