A 2x2 contingency table comparing an exposure with a rare disease outcome contains several cells with expected frequencies below 5. Which test should the researcher apply instead of the ordinary chi-square test?
- A Yates corrected chi-square test
- B Fisher exact test ✓
- C Student t-test
- D Wilcoxon rank sum test
Correct answer: B. Fisher exact test
Explanation
The chi-square approximation becomes unreliable when any expected cell frequency falls below 5 in a 2x2 table. The Fisher exact test calculates the exact probability of the observed table under the null hypothesis and is the standard substitute for small expected counts. Yates correction is applied when expected counts are above 5 but relatively small; it does not rescue tables with cells below 5. Wilcoxon rank sum compares ordinal distributions, not categorical counts.
Reference: Park's Textbook of Preventive and Social Medicine, 27th ed.
High-yield for: NEET PGINI-CETNExTFMGEUSMLEPLABMRCP
Written and medically reviewed by the StethoPrep medical team.