One of the most valuable and helpful statistics is a non-parametric procedure called Chi Square analysis. It is also called the test of “goodness of fit”. Its symbol is “x squared” (x²). This is a commonly used statistical procedure by many graduate students and faculty.
Because the Chi Square relies on frequency data, its value lays in the statistic’s ability to answer questions about data that are nominal. Variables in many settings are measured very often by their categories - and not exact intervals. Chi Square allows you to answer important questions with variables measured with nominal or ordinal scales. Unlike the t-test and ANOVA procedures, the Chi Square statistic is not as powerful to reject the null. It does not use the mean or standard deviation for computation; it does not rely on an interval or ratio scaling.
An example of where a Chi Square statistic would be helful is to see if there are differences between male and female college students on choice of major field of study- Engineering or English. The null hypothesis would be: There is no difference between male and female college students on their choice between taking quantitative versus qualitative elective courses. Both variables GENDER and COURSE are categorical. Perfect for a Chi Square.
Return from chi square analysis to statistical tests.