Abstract

Chi square analysis, with probability distribution that goes by the same name, is used to test hypotheses referred to frequency distribution. This test contrasts observed frequencies with expected frequencies according to the nil hypothesis. The article describes the use of chi square analysis to test the association between two variables using a hypothetical situation and simulated data. Next, it describes its use in assessing how well a theoretical distribution can result when it attempts to represent the real distribution of data within a given sample. This is called obtaining the goodness of fit. Estimating the goodness of fit helps to see to what extent observed data are adjusted to a theoretical or expected distribution. For this, a second hypothetical situation is used with simulated data.