In other words, the Bayes Factor and thus the credibility of an association depends explicitly on what we assume for the typical magnitude of likely genetic effects. For example, if we assume that the average effect is not an odds ratio of 1.15 as in Figure 1, but an odds ratio of ORav=1.5, then the prior odds of association will be less, because fewer variants—with larger effects than in the ORav =1.15 scenario—would suffice to explain the genetic variability. A larger Bayes Factor would be needed to reach a 3:1 posterior. Moreover, large effects emerging from small studies will be more credible than in the ORav =1.15 scenario, while very small effects emerging with similar p-values from large studies will be less credible.[13]
