Definition


The Fisher’s test is used to compare two proportions from two independent groups. It helps to conclude if there seems to be an association between two factors or two diseases or a factor and a disease. It is used when the Chi² test cannot be performed because of too small samples.


In practice


If there is less than 5 patients for at least one case of a contingency table, as : 


Tobacco / COPDCOPD positiveCOPD negative
Smoker502
No smoker550


The Chi² test will be replaced by the Fisher's test.


How to use it on EasyMedStat ?


  1. Go to Statistics > Test variables
  2. Select a first variable which could be of types Yes-no or List
  3. Select a second variable which could be of types Yes-no or List
  4. Click on "Find a statistical association between ..."


If your samples are large enough, a chi-squared test will be performed automatically.


See also