Fig. 6

First row: Error rate and 95% confidence intervals (y-axis; 95% confidence intervals are not shown in cases where they are shorter than the symbols) for the test of the positive acceleration factor for different sample sizes (x-axis), methods (colors/shapes), and true acceleration factors (columns) averaged across all simulation conditions. Second row: Power and 95% confidence intervals of the test of the positive acceleration factor (y-axis) for different sample sizes (x-axis), methods (colors/shapes), and true acceleration factors (columns) averaged across conditions with different parametric families. Methods: Bayesian model-averaging (BMA, deep purple circles) and model selection over parametric families with: Bayes factors (BF = deep purple triangles), AIC (light green circles), and BIC (light green squares). Note the different scaling of the y-axis for the error rate (first row) and power (second row)