Table 3.

Approximate precision of the mean effect for a random effects design with M patients and N paired treatment periods per patient relative to one with M = 46 and N = 1, using several within- and between-patient variances: (A) 100,200; (B) 10,200; (C) 100;20; respectively. Reference is 46 patients and 1 pair of treatment periods.

A) Between-Patient Variance = 100; Within-Patient Variance = 200
Pairs of Treatments	Number of Participants Enrolled
	10	20	25	30	46	60	100
1 pair	0.22	0.43	0.54	0.65	1.00	1.30	2.17
2 pairs	0.36	0.72	0.91	1.09	1.67	2.17	3.62
3 pairs	0.47	0.93	1.16	1.40	2.14	2.80	4.66
4 pairs	0.54	1.09	1.36	1.63	2.50	3.26	5.43
5 pairs	0.60	1.21	1.51	1.81	2.78	3.62	6.04
7 pairs	0.69	1.38	1.73	2.08	3.18	4.15	6.92
10 pairs	0.78	1.55	1.94	2.33	3.57	4.66	7.76
100 pairs	1.05	2.09	2.61	3.14	4.81	6.27	10.45

B) Between-Patient Variance = 10; Within-Patient Variance = 200
Pairs of Treatments	Number of Participants Enrolled
	10	20	25	30	46	60	100
1 pair	0.22	0.43	0.54	0.65	1.00	1.30	2.17
2 pairs	0.42	0.85	1.06	1.27	1.95	2.55	4.24
3 pairs	0.62	1.24	1.55	1.87	2.86	3.73	6.22
4 pairs	0.81	1.62	2.03	2.43	3.73	4.86	8.10
5 pairs	0.99	1.98	2.48	2.97	4.56	5.94	9.90
7 pairs	1.33	2.65	3.32	3.98	6.11	7.96	13.27
10 pairs	1.78	3.57	4.46	5.35	8.20	10.70	17.83
100 pairs	6.37	12.73	15.92	19.10	29.29	38.20	63.66

C) Between-Patient Variance =100; Within-Patient Variance = 20
Pairs of Treatments	Number of Participants Enrolled
	10	20	25	30	46	60	100
1 pair	0.22	0.43	0.54	0.65	1.00	1.30	2.17
2 pairs	0.25	0.51	0.63	0.76	1.17	1.52	2.54
3 pairs	0.27	0.54	0.67	0.81	1.24	1.61	2.69
4 pairs	0.28	0.55	0.69	0.83	1.27	1.66	2.77
5 pairs	0.28	0.56	0.70	0.85	1.30	1.69	2.82
7 pairs	0.29	0.58	0.72	0.86	1.32	1.73	2.88
10 pairs	0.29	0.59	0.73	0.88	1.35	1.76	2.93
100 pairs	0.30	0.61	0.76	0.91	1.39	1.82	3.03