Table 1.

Data from a fictional cluster-randomized trial

Cluster ($i$)	Number of participants ($n_i$)	True treatment effect in cluster $i$
1	10	5
2	10	5
3	10	5
4	100	1
5	100	1
6	100	1

$i$ denotes cluster number and $n_i$ denotes the number of participants in cluster $i$.