Table 1. Parameters from data generating models needed to simulate power.
notation | text | outcome data type | ||
continuous | binary | count | ||
K | n.clusters | ✓ | ✓ | ✓ |
J | n.periods | ✓ | ✓ | ✓ |
Njk | clust.size | ✓ | ✓ | ✓ |
, | period.effecta, period.var b | ✓ | ✓ | ✓ |
effect.size | ✓ | ✓ | ✓ | |
btw.clust.var | ✓ | ✓ | ✓ | |
indiv.varc | ✓ | |||
ICC | ICCc | ✓ | ||
Tijk | at.risk.params | ✓ |
The period effects are drawn from a normal distribution centered at period.effect with variance period.var.
If period.var = 0, then period.effect is assumed to be the same for all periods.
Only one of the ICC and needs to be specified in continuous data generating models.