Predicted sample size for significance level 0.01 and statistical power 0.90 (in parentheses for power 0.80) under assumption of different success frequencies p versus q (refers to example 3)
| q = 10% | q = 20% | q = 30% | q = 40% | q = 50% | |
|---|---|---|---|---|---|
| p = 20% | 382 (303) | ||||
| p = 30% | 122 (98) | 553 (473) | |||
| p = 40% | 64 (52) | 158 (126) | 668 (526) | ||
| p = 50% | 40 (33) | 77 (62) | 179 (143) | 725 (571) | |
| p = 60% | 28 (23) | 46 (37) | 83 (67) | 186 (148) | 732 (574) |









