Figure 5From: A Wild Bootstrap approach for the selection of biomarkers in early diagnostic trialsEmpirical type-I error for varying ccr. Empirical type-I error of the MCP, the Logit and the WB-Normal approach for varying case-control ratios.Back to article page