This calculator performs multiple analyses of a 2 X 2 experiment. To use it, fill in the boxes with counts for each dependent and independent variable value. For convenience, you can also change the names of each treatment and outcome.
Output
Error estimates are computed assuming a normal distribution of counts, and may therefore be inaccurate for very small counts.
The χ2 test uses a normal approximation to the count distribution; it may therefore be inaccurate for very small counts.
The Fisher exact test is accurate even for small counts; it is performed only for N ≤ 200.
Operations on 2 X 2 and N X M contingency tables are part of the functionality of the Meta.Numerics.Statistics library.
|