Home > Statistics Every Writer Should Know > The Stats Board > Discusssion
fisher exact test
Message posted by john miner on October 22, 1999 at 12:00 AM (ET)
what are the indications for the use of this test?Tell me a little about it.
READERS RESPOND:
(In chronological order. Most recent at the bottom.)
Re: fisher exact test
Message posted by on October 22, 1999 at 12:00 AM (ET)
Re: fisher exact test
Message posted by Bill on October 22, 1999 at 12:00 AM (ET)
The chi-square test should not be used when the expected frequencies in a 2x2 table are small. "Small" has been defined as less that 5 for any cell, less than 2, or less than 5 for half (2) of the cells. Once the definition of small is meet then the Fisher Exact test is the alternative test for this 2x2 table.
Re: fisher exact test
Message posted by Jack Tomsky on October 24, 1999 at 12:00 AM (ET)
Suppose that you have samples from two binomial distributions, x1~Bin(p1, N1) and x2~Bin(p2, N2). You want to test that the two probability parameters, p1 and p2, are equal. Fisher's exact test uses the fact that the conditional distribution of x2 given x1+x2 = t is a hypergeometric distribution.
Your $5 contribution helps cover part the $500 annual cost of keeping this site online.