The Objectivity of Subjective Bayesian Inference

Publication Year:
Usage 646
Downloads 646
Social Media 3
Tweets 3
Repository URL:
Sprenger, Jan
Most Recent Tweet View All Tweets
preprint description
Subjective Bayesianism is a major school of uncertain reasoning and statistical inference. Yet, it is often criticized for an apparent lack of objectivity. By and large, these criticisms come in three different forms. First, the lack of constraints on prior probabilities, second, the entanglement of statistical evidence and degree of belief, third, the apparent blindness to bias in experimental design. This paper responds to the above criticisms and argues in addition that frequentist statistics is no more objective than Bayesian statistics. In particular, the various arguments are related to different senses of scientific objectivity that philosophers have worked out in recent years.