Hello Ernest Tewfik SPSS software (v27) uses Fisher-Freeman-Halton Exact Test in its outputs.

When Expected count is less than 5 Fisher exact test included in the output of SPSS for all versions.

Thanks a lot!

M. Shihab,

thanks for your answer.

I also had the understanding that one should use the Fisher test when the expected count is less than 5 for at least one of the cells in a 2x2 table.

However, the following article (ref1) says that the "N-1 chi-square" (= linear-by-linear association in SPSS, ref2) is good enough when all expected counts are >1!

Thus, I was wondering if I could use "1" in my article.

Thanks.

1. Campbell I. Chi-squared and Fisher-Irwin tests of two-by-two tables with small sample recommendations. Stat Med. 30 août 2007;26(19):3661‑75.

2. Busing FMTA, Weaver B, Dubois S. 2 × 2 Tables: a note on Campbell’s recommendation. Statistics in Medicine. 2016;35(8):1354‑8.