Appendix 12—figure 1. Simulation study demonstrating how likelihood re-weighting can improve estimation accuracy in case-control studies.

Panels (A) and (B) show histograms of the case probability weights used in the simulations for the scenarios when 50% of cases are true cases and when 100% of cases are true cases, respectively. Panel (C) shows the estimated effect sizes as a function of the proportion of mis-classified cases. Panel (D) shows the standard errors of effect estimates as a proportion of mis-classified cases.