Table 6.
Sample size required for a range of effect sizes and different hypotheses to be tested
| |
(i) Main effects of treatments versus placebo |
(ii) Comparison of pairs of main effects versus one another |
(iii) Interactions between pairs of treatments |
|||
|---|---|---|---|---|---|---|
| Effect size (% difference in TI) | Full dataset | Reduced dataset* | Full dataset | Reduced dataset* | Full dataset | Reduced dataset* |
| 5% |
847 |
3388 |
1694 |
3388 |
3388 |
6776 |
| 7.5% |
385 |
1542 |
771 |
1542 |
1542 |
3084 |
| 10% | 222 | 888 | 444 | 888 | 888 | 1776 |
* Avoiding interference from irrelevant treatment.