Chi-Square

Abstract

The Chi-square test is perhaps the most frequently used (or overused) nonparamteric statistical test. The Chi-square test, named for the Greek letter χ (i.e., Chi or the Greek letter for x), is typically used to test for differences in proportions between two or more groups. The Chi-square test is also called a goodness of fit test. That is to say, the Chi-square test is used to see if grouped data actually fit into declared groups, or if the data instead do not fit into the group. For this lesson, Chi-square will be demonstrated using data in two formats: (1) Chi-square using R will first be demonstrated where the data are presented as an external file imported into R, with data organized at the level of individual subjects, (i.e., each row represents the data for an individual subject) and (2) Chi-square using R will also be demonstrated where data are not at the level of individual subjects but data are instead presented in summary format, as a collapsed contingency table.