Fig. 3

Nested loop plot of root mean squared error (RMSE) of \( {\hat{\beta}}_1 \) by the expected value of Y, E(Y) ∈ {0.1, 0.25}, the number of predictors K ∈ {2, 5, 10}, noise absent or present (full and dashed lines), the sample size N ∈ {100, 250, 500} and the size of true coefficients β₁ ∈ {1.04, 2.08} for simulated scenarios. Due to poor performance some results lie outside the plot range. OEX, explanation oracle; D, deviance; GCV, generalized cross-validation; CE, classification error; RCV50, repeated 10-fold cross-validated deviance with θ = 0.5; RCV95, repeated 10-fold cross-validated deviance with θ = 0.95; AIC, Akaike’s information criterion; IP, shrinkage based on informative priors; WP, shrinkage based on weakly informative priors; FC, Firth’s correction. See Table S2 and S3 for results on scenarios with N = 1000. Results regarding RMSE of β₂ are contained in Table S4 and S5