Table 4.
Required sample sizes of the best 3T3 guideline-based design (with spread factor chosen based on hill slope) for the standard parameterization of the log-logistic function when the true parameters are not standard, compared with the optimal design with 100 observations in the same situation
True b | True e | Required n equally spaced design | Required n optimal design |
---|---|---|---|
1 | 1 | 131 | 100 (reference) |
0.9 | 1 | 130 | 101 |
1 | 0.9 | 131 | 100 |
0.9 | 0.9 | 129 | 101 |
1.1 | 1.1 | 134 | 100 |
0.5 | 1 | 131 | 110 |
1 | 0.5 | 131 | 109 |
0.5 | 0.5 | 128 | 112 |
1 | 2 | 139 | 109 |
2 | 1 | 195 | 132 |
2 | 2 | 211 | 164 |
Parameters c and d do not affect the design