Statistics > Multinomial > Chi-Square Goodness-of-Fit Statistics > Multinomial > Chi-Square Goodness-of-Fit The Chi-Square Goodness-of-Fit test determines whether an observed frequency distribution agrees with some claimed distribution. It applies to data of different categories resulting from a multinomial experiment. The test statistic for Goodness-of-Fit tests in multinomial experiments is When using this utility, you must provide the observed frequencies (frequency counts or categorical data) and the expected frequencies or probabilities. You must also provide the confidence level of the test. Dialog Inputs Observed Frequencies If the frequency counts of the different categories are in one column of the Datasheet, select the Frequencies in Column: radio button and select the column in the drop-down menu. If the category names corresponding to the frequency counts are also in one column of the Datasheet, you can optionally select the column containing the category names in the drop-down menu. If the categorical data are in one column (instead of frequency counts), selet the Categorical Data in Column: radio button and select the column in the drop-down menu. Expected Frequencies If the different categories of data are expected to have the same frequency, select the Equal Frequencies radio button. If unequal frequencies are expected, select the Unequal Frequencies radio button. If the expected frequencies are provided, select the column containing the expected frequencies in the Frequencies in Column drop-down menu. If the expected probabilities are given, select the column containing the probabilities in the Probabilities in Column drop-down menu. If a past sample of categorical data is available, select the Categorical Data radio button. Select the column containing the categorical data values in the Past Sample Data in Column drop-down menu. Significance Level Enter the significance level in the given text box. The significance level must be between 0 and 1.