Skip to content
Prev 174498 / 398503 Next

Fisher test problem

Viju Moses 
As you say, I guess that's not possible. Meanwhile I've got R's results verified in Stata

. tabi 17 6 \ 2 1, chi2 exact

           |          col
       row |         1          2 |     Total
-----------+----------------------+----------
         1 |        17          6 |        23 
         2 |         2          1 |         3 
-----------+----------------------+----------
     Total |        19          7 |        26 

          Pearson chi2(1) =   0.0708   Pr = 0.790
           Fisher's exact =                 1.000
   1-sided Fisher's exact =                 0.627

The OpenEpi.com output was: (note the 2 tailed P value (fisher test))

Chi Square and Exact Measures of Association
	Test	Value	p-value(1-tail)	p-value(2-tail)
	Uncorrected chi square	0.07083	0.3951	0.7901
	Yates corrected chi square	0.1813	0.3351	0.6702
	Mantel-Haenszel chi square	0.06811	0.3971	0.7941
	Fisher exact		0.6273	1.25
	Mid-P exact		0.3971	0.7942

Thanks for pointing it out.. Had been under the notion that OpenEpi was very... strong (being associated with EpiInfo and WHO).

Viju
Tal Galili wrote:

Thread (6 messages)

Viju Moses Fisher test problem Mar 21 Tal Galili Fisher test problem Mar 21 David Winsemius Fisher test problem Mar 21 C.H. Fisher test problem Mar 21 Daniel Nordlund Fisher test problem Mar 21 Viju Moses Fisher test problem Mar 21