Skip to main content
. 2019 Dec 10;85(1):101–120. doi: 10.1007/s11336-019-09692-3

Table 2.

Computed sample size, estimated power, and simulated power for the ANOVA, approximate ANCOVA, and exact ANCOVA methods when G=3, σY2=10000, ρ=0.5, σ2=7500, {μ1,μ2,μ3}={400,450,500}, δ2=0.2222, Type I error α=0.05, and nominal power 1-β=0.80

P ANOVA Approximate ANCOVA Exact ANCOVA
NTO Estimated power Simulated power Error NTA Estimated power Simulated power Error NTE Estimated power Simulated power Error
1 63 0.8148 0.9073 - 0.0925 48 0.8136 0.7994 0.0142 48 0.8042 0.8006 0.0036
2 63 0.8148 0.9022 - 0.0874 48 0.8130 0.7993 0.0137 51 0.8217 0.8215 0.0002
3 63 0.8148 0.9006 - 0.0858 48 0.8123 0.7883 0.0240 51 0.8122 0.8062 0.0060
4 63 0.8148 0.8945 - 0.0797 48 0.8115 0.7705 0.0410 51 0.8022 0.8053 - 0.0031
5 63 0.8148 0.8926 - 0.0778 48 0.8108 0.7592 0.0516 54 0.8201 0.8271 - 0.0070
6 63 0.8148 0.8799 - 0.0651 48 0.8099 0.7528 0.0571 54 0.8104 0.8177 - 0.0073
7 63 0.8148 0.8737 - 0.0589 48 0.8091 0.7366 0.0725 54 0.8004 0.7976 0.0028
8 63 0.8148 0.8653 - 0.0505 48 0.8082 0.7196 0.0886 57 0.8185 0.8291 - 0.0106
9 63 0.8148 0.8539 - 0.0391 48 0.8072 0.7087 0.0985 57 0.8088 0.8157 - 0.0069
10 63 0.8148 0.8526 - 0.0378 48 0.8062 0.6957 0.1105 60 0.8263 0.8259 0.0004