Skip to main content
. Author manuscript; available in PMC: 2016 May 4.
Published in final edited form as: Stat Sci. 2015;30(2):199–215. doi: 10.1214/14-STS504

Table 6.

Comparison of different four-arm trial designs of size T = 423. Fa: Fisher's adjusted test; α: family-wise type-I error; 1 − β: power; p*: expected proportion of patients in the trial assigned to the best treatment; ENS: expected number of patient successes; UB: upper bound

Crit.
value
H0 : p0 = pi = 0.3 for i = 1, …, 3
H1 : p0 = pi = 0.3, i = 1, 2, p3 = 0.5
α p* (s.e.) ENS (s.e.) (1 − β) p* (s.e.) ENS (s.e.)
FR 2.128 0.047 0.250 (0.02) 126.86 (9.41) 0.814 0.250 (0.02) 148.03 (9.77)
TS 2.128 0.056 0.251 (0.07) 126.93 (9.47) 0.884 0.529 (0.09) 172.15 (13.0)
UCB 2.128 0.055 0.251 (0.06) 126.97 (9.41) 0.877 0.526 (0.07) 171.70 (11.9)
RBI 2.128 0.049 0.250 (0.03) 126.77 (9.40) 0.846 0.368 (0.04) 158.34 (10.4)
RGI 2.128 0.046 0.250 (0.03) 126.80 (9.36) 0.847 0.358 (0.03) 157.26 (10.3)
CB Fa 0.047 0.269 (0.39) 126.89 (9.61) 0.213 0.677 (0.41) 184.87 (36.8)
GI Fa 0.048 0.248 (0.18) 126.68 (9.40) 0.428 0.831 (0.10) 198.25 (13.7)
CG 2.128 0.034 0.250 (0.02) 127.16 (9.46) 0.925 0.640 (0.08) 182.10 (12.3)
UB 126.90 (0.00) 1 211.50 (0.00)