Table 4: N = 120. Designs for testing H0: p p0 vs. H1: p > p0 with = 0: 05 and power at least 80% at p1 = p0 + 0:15. For each setting the Rows 1-3 contain the Types 1-3 two-stage designs. The left and rights parts are for the first type and second type of optimal designs.
| p0 (1-stage) | 2-stage minimizing EN0 | 2-stage minimizing n2 | ||||||
|---|---|---|---|---|---|---|---|---|
| Design | EN0 | n2 | P ET0 | Design | EN0 | n2 | P ET0 | |
0.1 (29; 5) |
(19, 4; 29, 5) (22, 3; 29, 5) (20, 3, 4; 29, 5) |
28.8 24.7 22.6 |
29 29 29 |
0.02 0.62 0.71 |
(19, 4; 29, 5) (22, 3; 29, 5) (20, 3, 4; 29, 5) |
28.8 24.7 22.6 |
29 29 29 |
0.02 0.62 0.71 |
0.2 (39; 11) |
(35, 10; 38, 11) (23, 5; 39, 11) (22, 5, 8; 39, 11) |
37.9 31.1 29.7 |
38 39 39 |
0.04 0.49 0.55 |
(35, 10; 38, 11) (23, 5; 39, 11) (35, 1, 10; 38, 11) |
37.9 31.1 37.9 |
38 39 38 |
0.04 0.49 0.04 |
0.3 (44; 17) |
(24, 11; 44, 17) (26, 9; 42, 16) (26, 9, 13; 42, 16) |
43.6 31.8 31.7 |
44 42 42 |
0.01 0.64 0.64 |
(24, 11; 44, 17) (24, 8; 42, 16) (24, 8, 13; 42, 16) |
43.6 31.8 31.8 |
44 42 42 |
0.01 0.57 0.57 |
|
0.4 (45; 22) |
(29, 16; 45, 22) (31, 13; 45, 22) (30, 13, 17; 45, 22) |
42.7 37.7 36 |
45 45 45 |
0.02 0.52 0.6 |
(29, 16; 45, 22) (31, 13; 45, 22) (30, 13, 17; 45, 22) |
42.7 37.7 36 |
45 45 45 |
0.02 0.52 0.6 |
|
0.5 (46; 27) |
(27, 18; 44, 26) (24, 13; 46, 27) (24, 13, 17; 46, 27) |
43.8 33 32.9 |
44 46 46 |
0.01 0.59 0.6 |
(22, 16; 44, 26) (24, 13; 46, 27) (22, 1, 16; 44, 26) |
43.9 33 43.9 |
44 46 44 |
0.01 0.59 0.61 |
|
0.6 (42; 29) |
(29, 22; 42, 29) (21, 13; 42, 29) (26, 17, 20; 42, 29) |
41.8 32 31.3 |
42 42 42 |
0.01 0.48 0.67 |
(29, 22; 42, 29) (21, 13; 42, 29) (26, 17, 20; 42, 29) |
41.8 32 31.3 |
42 42 42 |
0.01 0.48 0.67 |
|
0.7 (38; 30) |
(23, 20; 38, 30) (24, 19; 37, 29) (22, 17, 19; 38, 30) |
37.9 26.6 26.5 |
38 37 38 |
0.01 0.8 0.72 |
(23, 20; 38, 30) (24, 19; 37, 29) (24, 19, 22; 37, 29) |
37.9 26.6 26.6 |
38 37 37 |
0.01 0.8 0.8 |