Table 4.
Required sample size per group for ANCOVA vs. ANOVA assuming α = 0.05
| ANCOVA Power |
ANOVA Power |
|||||
|---|---|---|---|---|---|---|
| Type of TT Test | Effect of Altitude | Effect of Treatment* | 80% | 95% | 80% | 95% |
| 720-kJ cycle | 20% | 5% | 243 | 402 | 258 | 429 |
| 20% | 17 | 27 | 17 | 29 | ||
| 33% | 5% | 198 | 328 | 210 | 349 | |
| 20% | 14 | 22 | 14 | 23 | ||
| 11.2-km run | 20% | 5% | 71 | 120 | 74 | 123 |
| 20% | 6 | 8 | 6 | 9 | ||
| 33% | 5% | 60 | 98 | 60 | 100 | |
| 20% | 5 | 7 | 5 | 8 | ||
| 3.2-km run | 20% | 5% | 154 | 254 | 247 | 409 |
| 20% | 11 | 18 | 17 | 27 | ||
| 33% | 5% | 126 | 208 | 201 | 310 | |
| 20% | 10 | 15 | 14 | 22 | ||
ANCOVA, analysis of covariance; TT, time trial.
Assumes an attenuation from placebo group, e.g., if there is a 20% increase from altitude alone, then a 5% treatment effect would result in a 1% attenuation in the increase from altitude (20% vs. 19% increase in control vs. treatment groups. respectively).