How to conduct a Chi-square Goodness-of-Fit Test Step 1: The chi-square goodness-of-fit test typically starts with a single column of observed frequency data, the distribution of which is to be tested. Excel does not have a built-in function to perform the goodness-of-fit test, but it is easy to compute the required quantities. Compute a column of expected counts based on the null hypothesis proportions. (Typically the total observed count times the null hypothesis proportion for each cell.) Compute the contribution of each row to the chi-square test statistic, namely (observed frequency-expected count)2/(expected count) for that row. Compute the chi-square tests statistic as the sum of all the chi-square contributions. Use the Excel function chisq. dist. rt(x, df) to compute the p-value.