bayesplain.frequentist.chi_square_independence

bayesplain.frequentist.chi_square_independence(table, correction=False)[source]

Pearson chi-square test of independence for an r x c table.

Parameters:
table : array_like

Table of counts.

correction : bool, default False

Apply Yates’ continuity correction. Off by default so that the 2x2 statistic matches the square of the two-proportion z.

Returns:

The conventional result. estimate carries the sample Cramer’s V, as a point estimate with no uncertainty attached – which is precisely the gap the Bayesian version fills.

Return type:

FrequentistTwin