Skip to main content
Fig. 2 | BMC Medical Research Methodology

Fig. 2

From: Informed Bayesian survival analysis

Fig. 2

Left: Distribution of the inclusion Bayes factors for the presence of the treatment effect in the fixed-n design under the null hypothesis (assuming the absence of the treatment effect; light green) and the alternative hypothesis (assuming the presence of the treatment effect; deep purple). 32.5% Bayes factors under the alternative hypothesis are larger than 1000 and not shown. The vertical dashed lines visualize boundaries for obtaining 10% false-negative evidence and 5% false-positive evidence. Right: Trajectories of the inclusion Bayes factors for the presence of the treatment effect in the sequential design under the null hypothesis (assuming the absence of the treatment effect; light green) and the alternative hypothesis (assuming the presence of the treatment effect; deep purple). Ten example trajectories are visualized in the full colored lines. The bounds are truncated in the range of 1/15 and 15. The horizontal dashed lines visualize boundaries for obtaining 10% false-negative evidence and 5% false-positive evidence

Back to article page