More evidence in favor of the hypothesis that we are natural frequentists: Even though I try to think like a Bayesian, I am mentally incapable of assigning probabilities without perfect information. The best I can do is offer a range of probabilities that seem reasonable, whereas a real Bayesian should be able to average the probabilities that it would assign given all possible sets of perfect information, weighted by likelihood.
More evidence in favor of the hypothesis that we are natural frequentists: Even though I try to think like a Bayesian, I am mentally incapable of assigning probabilities without perfect information. The best I can do is offer a range of probabilities that seem reasonable, whereas a real Bayesian should be able to average the probabilities that it would assign given all possible sets of perfect information, weighted by likelihood.