Table 1.
Summary of simulation study data scenarios.
| Description | Similar to motivating example? | |
|---|---|---|
| Common to all simulations | ||
| Number of groups | 3 | Yes |
| Number of time periods |
2. In period 1, group 1 received the intervention. In period 2, groups 1 and 2 received the intervention. |
Yes |
| Number of clusters | 75 | Yes |
| Cluster size | Log‐normal(6.9, 0.74) in each year. Geometric mean = 1027 | Yes |
| Correlation of measurements within clusters | Independent within cluster‐periods | No |
| Mean outcome in year 1 | Odds = 6.61 | Yes |
| Mean change in outcome from year 1 to year 2 | Odds ratio = 0.32 | Yes |
|
Different scenarios |
||
| Period effect | (1) Common period effect, high variability | No |
| (2) Common period effect, low variability | No | |
| (3) Varying period effect, decreasing variability | Yes | |
| (4) Varying period effect, stable variability | No | |
| Intervention effect | (a) Log(OR) = 0.41 common to all clusters | No |
| (b) Log(OR) = 0.41, varying between clusters | No | |
| Intervention effect in group 2 | ||
| Simulation study 1 | Intervention effect in group 2 the same as group 1 log(OR) = 0.41 | No |
| Simulation study 2 | Intervention effect in group 2 is log(OR) = 1.5, and group 1 is log(OR) = 0.41 | No |