Fig. 1: Simulation of the Bayesian two-sided sequential design. | Nature Human Behaviour

Fig. 1: Simulation of the Bayesian two-sided sequential design.

From: Self-administered mindfulness interventions reduce stress in a large, randomized controlled multi-site study

Fig. 1

After 10,000 iterations, the simulation indicates that under the proposed design, there is a 79% chance (72% under H1 and 7% erroneously under H0) that the test will reach compelling evidence boundaries (BF10 = 10 or 1/10). There is a 21% chance that the test will conclude by reaching the maximum (max.) sample size of 720 per condition, with a 5% probability of providing some evidence in favour of H1 (BF10 > 3).

Back to article page