Studies in the history of probability and statistics XLIII Karl Pearson and quasi-independence

Abstract

SUMMARY In 1913 Karl Pearson introduced and correctly fitted the model of quasi-independence to an incomplete contingency table, in the context of a study in his research on eugenics. In assessing the significance of the chi-square statistic, he recognized the need to allow for restrictions on the cell entries, but he did not see the need to compensate for the estimation of cell probabilities in this case, nor did he generalize the ideas to tables with other structures.