最近看了一篇论文,其中要用到贝叶斯的posterior odds tatio检验,有点不太明白,希望哪个高手指教一下,谢谢!!现在把那篇论文相关文献摘录如下:
The problems of interpreting large samples may be avoided by casting our
inferences in terms of a real-world decision maker. In particular, we provide a
Bayesian interpretation of our results via posterior odds ratios. The posterior
odds ratio simply represents the ratio of the probabilities of the null to the
alternative hypothesis given the decision maker’s prior beliefs and the sample
information. With diffuse prior beliefs, the posterior odds ratios correspond to
standard significance levels and would equal about 0.053 at the 0.05 level and
0.0101 at the 0.01 level. In presenting our results we use two weakly informa-
tive priors. Both cases assume that the null hypothesis of no tax premium is
true with probability 0.5. Our prior beliefs about the alternative hypotheses are
represented as a 0.5 probability that (1) the mean ex-day SER is between - 1
and + 1 with uniform probability, and (2) the mean SER is distributed as
normal with a mean of zero and standard deviation of 0.316.‘“.” Besides
reporting these posterior odds ratios, we also report standard significance
levels.