ABSTRACT
The chi-square (2) test presented in this chapter is concerned entirely with categorical variables; those producing nominal data, that is, frequencies by categories or levels. l Chi-square is first used to analyse a simple division of one variable into two levels of
frequencies. l The concept of expected frequencies under the null hypothesis is introduced. l Cross-tabs tables are then introduced and chi-square used to analyse for association
between two categorical variables with two levels each (a 2 × 2 analysis). l The generalised form of chi-square testing r × c tables (those with any number of rows and
columns) is then covered. l Chi-square can also be used as a goodness of fit test to check whether a distribution of
frequencies in categories is a close fit to a theoretical distribution (e.g., whether a college’s pattern of degree classifications matches the average pattern for the country).
