Perhaps multiple choice tests in schools make people extra susceptible to privileging hypotheses. As a simplified example, if a student’s probability distribution to the answer of a question on a multiple choice test before seeing the choices is uniformly distributed amongst all integers from 1 to n, simply seeing an arbitrary integer as one of the, say, four options in a multiple choice test justifiably increases its probability of being correct to 0.25, a tremendous increase when n is large. Thus, on multiple choice tests, privileging a possible answer being “suggested” by being one of the answers the student can pick is a useful strategy on multiple choice tests, so we need to remember that privileging the hypothesis is not a viable strategy in other contexts. Maybe multiple choice tests aren’t so good after all...