On the potential mismatch between the function of the Bayes factor and researchers’ expectations