Bayesian Hypothesis testing
What is Bayesian Hypothesis testing?
In order to standardize a test, the statistician has some basic knowledge: for example, they know that most male human individuals would weigh around a certain amount and thus they can put the average weight of someone in this setting is between 65 and 120 kilograms
What are the benefits of Bayesian hypothesis testing?
You get to do things quickly based on previous knowledge.
One example of a hypothesis test is to calculate the average weight for American people (Male or female, or combined). As more data on the population becomes available, we should expect that height online will be increasingly close to the average.
Does it have any cons?
A common problem with initial Bayesian analysis is the tenuous connection with current knowledge that it ensues.
Thus, for example, that the probability of a true conversion rate is meaningful in Bayesian models and can be calculated by our knowledge of the value based on prior information or data.
This varies from a frequentist approach, in which this number is unknown and fixed.
A hypothesis is given a degree of certainty based on the highest posterior probability.