|
In relation to my last post, yes, you could use Bayesian analysis but what would be the point. Considering its inherent subjectivity, your results would be biased right from the start. Assuming the prior conditions of the statistical set is just as accurate as saying "the milkman did it". Given that the number of assumptions one can make about the priors of any analysis, you might as well use "the milkman did it" because everyone with an opinion on what initially happened could make an analysis and get what they wanted out of it. And, that's with exactly the same set of values!!!.
That's why I believe we have to be careful with using statistics. It can be made to say or justify anything. If it were as unambiguous as algebra or calculus, for example, then there would be no problem. But it's not.
|