Academic reference on the "minimum of 5 expected counts per cell" rule of thumb for Chi-Square test?

Question

I remember very clearly an academic paper that investigated thoroughly the "minimum of 5 expected counts per cell" rule of thumb when conducting a chi-square test on a contingency table. It explained where this rule of thumb comes from, its history, and explored its reliability through simulations. The paper concluded with some precise recommendations.

To my embarrassment, I made the mistake of not saving the reference at the time I saw it, and now I'm looking for it again, to no avail.

Does anyone know what reference it would be? Of course, I'm also interested in other good references relatively to this subject.

Thanks,

Perhaps see: stats.stackexchange.com/questions/30765/for-chi-square-on-any-2-by-x-contingency-table-should-no-more-than-20-of-the-c — Sal Mangiafico, Feb 03 '23 at 14:22

score 4 · Accepted Answer · answered Feb 03 '23 at 13:55

Check the paper "The Tale of Cochran's Rule: My Contingency Table has so Many Expected Values Smaller than 5, What Am I to Do?" by P. M. Kroonenberg & Albert Verbeek.

Full citation: P. M. Kroonenberg & Albert Verbeek (2018) The Tale of Cochran's Rule: My Contingency Table has so Many Expected Values Smaller than 5, What Am I to Do?, The American Statistician, 72:2, 175-183, DOI: 10.1080/00031305.2017.1286260

Academic reference on the "minimum of 5 expected counts per cell" rule of thumb for Chi-Square test?

1 Answers1

Related