Skip to main content
. 2017 Dec 6;181(3):843–867. doi: 10.1111/rssa.12341

Figure 6.

Figure 6

Invariant region formed from threshold lines for bias adjustment to the two relative effect estimates from study 1 (Inline graphic, invariant region): the new optimal treatments on the boundary are indicated by k~*; optimal treatment without bias adjustment is k *=3