Effective Sample Size and Power Holding k Constant
| Number of Physicians | Number of Patients | Total Number | ρ= 0.017 | Power | |
|---|---|---|---|---|---|
| k | m | (mk) | DE | ESS | tTest* |
| *Power to detect a minimum effect size of 0.5 when one half the clusters is randomly assigned to treatment and the other half to placebo, and mean responses are compared using a 2-sample t test made at the .05 level of significance. | |||||
| m = number of subjects in a cluster; k= number of clusters; mk = total number of subjects in a clustered study; DE = design effect; and ESS = effective sample size. | |||||
| 4 | 10 | 40 | 1.153 | 34 | 29 |
| 4 | 20 | 80 | 1.323 | 60 | 47 |
| 4 | 40 | 160 | 1.663 | 96 | 67 |
| 4 | 80 | 320 | 2.343 | 136 | 82 |