Table 2 Sample size required to detect a 2, 3, and 4 percentage point increase in condom use (4%, 5%, and 7% increase)
Hypothesized standard deviation of condom use | Hypothesized increase in proportion of condom use between groups | Sample size required in each group (ratio 1:1) | Adjusted for effect of clustering (design effect = 2) | 10% refusal rate (estimated) |
|---|---|---|---|---|
0.15 | 7% | 296 | 592 | 651 |
0.15 | 5% | 526 | 1052 | 1157 |
0.15 | 4% | 1183 | 2366 | 2603 |