What are chisquare tests? Statistics Read more from Chapter Applied Mathematics

The chi-square test is a way to determine the odds for or against a given deviation from the expected statistical distribution. This somewhat complex statistical test computes the probability that there is no major difference between the expected frequency of an event with the observed frequency of that event—and especially to determine if the set of responses is significantly different from an expected set of responses only because of chance.

There are even various ways to perform this type of test, such as the most common type, called the Pearson’s chi-square test. Another is called the likelihood ratio chi-square test (or G test), in which a hypothesis is tested of no association of columns and rows in nominal-level tabular data. Yet another is the chi-square goodness-of-fit test, which is just a different use of the Pearson chi-square; it is used to test if an observed distribution conforms to any other distribution. Taking this test one step further (making it more powerful) is the Kolmogorov-Smirnov goodness-of-fit test, which is preferred for interval data.