Irwin-Fisher Approximation

Note: This is a large-sample approximation to the Irwin-Fisher Exact Test. When n1 and n2 are large, so that n1p1 > 5, n1q1 > 5, n2p2 > 5, and n2q2 > 5, large-sample normal-curve theory can be used to determine the confidence interval for delta = p1 - p2. If these conditions are not met, an alert will be noticed. Recent studies suggest that this rule of thumb can be relaxed. However, this conservative postion will be alerted. If an alert is noticed, the Irwin-Fisher Exact Test can be obtained via probability tables (see Marascuilo,L. A., and McSweeney, M. (1977). Nonparametric and Distribution-Free Methods for the Social Sciences. Brooks and Cole Publishing: Monterey, Calif.).

Property	Sample 1	Sample 2	Total
+	t1	t2	t
-	n1-t1	n2-t2	n - t
Total	n1	n2	n

Enter the values based on a 2 x 2 table:

Cell t1 Cell t2

Cell n1-t1 Cell n2-t2

Proprotional value of sample 1 Proportional value of sample 2

Irwin-Fisher Approximation - Z Score, significant at alpha 05 if above or below + or - 1.96

Lower 95% Confidence Interval

Difference in proportional values

Upper 95% Confidence Interval