Large values for the number of groups with the Pearson chi-squared goodness-of-fit tes
Pearson 卡方拟合优度检验中的大量分组
Dahiya (1971) offered evidence against using a large number of equally likely groups with the Pearson chi-squared goodness-of-fit test, and indicated that when the number of groups exceeds the sample size the assumed level of the test is inflated. This paper provides data which question the former statement and disprove the latter.