Résumé en anglais

Following the critical review of Seaman III et al (2012), we re ect onwhat is presumably the most essential aspect of Bayesian statistics, namely theselection of a prior density. In some cases, Bayesian inference remains fairly stableunder a large range of noninformative prior distributions. However, as discussed bySeaman III et al (2012), there may also be unintended consequences of a choice of anoninformative prior and, these authors consider this problem ignored in Bayesianstudies. As they based their argumentation on four examples, we reassess theseexamples and their Bayesian processing via di erent prior choices. Our conclusionis to lower the degree of worry about the impact of the prior, exhibiting an overallstability of the posterior distributions. We thus consider that the warnings ofSeaman III et al (2012), while commendable, do not jeopardize the use of mostnoninformative priors.