Skip to main content
. Author manuscript; available in PMC: 2011 May 10.
Published in final edited form as: Stat Biopharm Res. 2011 Feb 1;3(1):97–105. doi: 10.1198/sbr.2010.09044

Table 3.

Type I error rates for the proposed test (Q) by tumor rate (π), shape of the incidence curve (γ1), number of historical control groups (k), and absence/presence of extra variation. All tests were performed at the nominal 5% level, with a sample size of n=50 per group and with 70% of the animals (on average) surviving to the terminal sacrifice.

πa γ1 No extra variationb
Extra variationb
k=1 k=2 k=5 k=10 k=2 k=5 k=10
0.01 1.5 2.3 5.5 4.6 5.0 5.5 4.4 4.5
3.0 2.2 5.5 4.5 5.0 5.5 4.4 5.1
6.0 2.4 5.5 4.3 5.7 5.8 4.1 4.8
0.05 1.5 5.4 5.1 3.9 4.1 5.1 4.4 4.6
3.0 5.3 5.0 3.7 4.3 5.1 4.3 4.7
6.0 5.3 5.0 3.8 4.2 5.0 4.1 4.4
0.15 1.5 4.8 5.1 4.8 4.8 4.9 4.8 4.8
3.0 4.7 5.0 4.7 4.9 5.1 4.7 4.8
6.0 4.4 4.9 4.9 4.8 4.8 4.9 4.9
0.30 1.5 5.3 5.0 5.1 5.0 4.8 5.0 5.0
3.0 5.2 5.0 5.2 4.8 4.9 5.1 5.2
6.0 5.2 5.0 5.1 5.1 5.0 5.0 5.3
a

The mean tumor rate among the historical control groups (πh) equals the mean tumor rate in the current control group (πc) under the null hypothesis (H0: πh = πc = π).

b

The variance of the tumor rates among the historical control groups in the presence of extra variation is 20% larger than in the absence of extra variation.