
154 Sample Size Calculations for Clustered and Longitudinal Outcomes
TABLE 5.1
Sample size and power for detecting a main effect δ
(2)
in model (5.3) when
randomizations occur at the second level (two-sided significance level α =
0.05)
λ
2
∆
(2)
ρ N
1
N
(0)
2
Total N ϕ
(2)
1 0.4 0.1 10 19 380 0.807
20 15 600 0.820
0.2 10 28 560 0.807
20 24 960 0.807
0.5 0.1 10 12 240 0.802
20 10 400 0.836
0.2 10 18 360 0.809
20 16 640 0.823
1.5 0.4 0.1 10 16 400 0.811
20 12 600 0.807
0.2 10 23 575 0.801
20 20 1000 0.807
0.5 0.1 10 10 250 0.802
20 8 400 0.820
0.2 10 15 375 0.809
20 13 650 0.813
Note: λ
2
represents the ratio of the number of level 2 units in the intervention
arm to the corresponding ...