Hello all, I have the following. Two sets of p x 3 matrices where p is say relatively large. My data looks like : mat1 2 3 4 2 3 4 1 2 3 mat2 2 3 4 12 12 4 10 12 3 when p = 3. I form a 2 x 3 contingency table using the i^th row from each matrix. Its Fisher's exact test I am doing for the 2 x 3 table. P(X|Ho is true) = sum (P(X| Ho true, Two column totals fixed) * P(Two column totals fixed) ) The first probability is fisher's exact test for the column totals after the equal sign while the second probability is to be computed from the data. Any ideas how to do this in R? -- Thanks, Jim. [[alternative HTML version deleted]]