Table 2.

Design effects to determine the required study sample size given a fixed sample size from each cluster (M)

M	ρ	CRT	CRT-BA		SW-CRT
Design effect	Implied r	Design effect	t	Design effect
30	0.001	1.03	0.01	2.03	2	3.03
30	0.01	1.29	0.13	2.24	2	3.22
30	0.05	2.45	0.44	2.74	2	3.58
30	0.1	3.90	0.63	2.93	2	3.63
30	0.25	8.25	0.83	2.75	2	3.23
60	0.001	1.06	0.03	2.06	5	1.92
60	0.01	1.59	0.23	2.44	5	2.20
60	0.05	3.95	0.61	3.06	5	2.61
60	0.1	6.90	0.77	3.18	5	2.65
60	0.25	15.75	0.91	2.86	5	2.33
150	0.001	1.15	0.07	2.14	2	3.13
150	0.01	2.49	0.43	2.83	2	3.72
150	0.05	8.45	0.80	3.42	2	4.05
150	0.1	15.90	0.89	3.41	2	3.94
150	0.25	38.25	0.96	2.94	2	3.34
300	0.001	1.30	0.13	2.26	5	2.07
300	0.01	3.99	0.60	3.17	5	2.70
300	0.05	15.95	0.89	3.59	5	2.93
300	0.1	30.90	0.94	3.50	5	2.83
300	0.25	75.75	0.98	2.97	5	2.39

Abbreviations: ρ, intracluster correlation coefficient; t, number of steps; CRT, cluster randomized trial; CRT-BA, cluster randomized trial with before and after observations; SW-CRT, stepped wedge cluster randomised trial.