could someone help explain the difference to me between bayesian and frequentist statistics? from what i understand, it has to do with how you treat the alternative hypothesis. i know in frequentist statistics, we do not reject the null hypothesis unless we beat a critical t score. do bayesian's estimate the probability of the alternative hypothesis (rejecting the null hypothesis) being true before having data? thanks for the help with this, trying to wrap m head around the logic. also, in which sort of situations would bayesian probability be more effective? it seems like in a general academic context, frequentist would be more useful.
A frequentist would describe the flip as a random variable, the result of a single trial sampling from a probability distribution with a fixed but unknown probability of landing heads. This is the more commonly accepted view of probability.
A Bayesian would say that the coin landed hands. It is not a random variable. What is random is the unknown probability of landing heads that generated the result. This is the random (i.e., not fixed) portion of the outcome.
This may seem like a subtle difference and in a way it is, but these different views on probability theory impact the way inferences are drawn from things such as hypothesis testing.
