Conclusions

The primary purpose of this book has been to develop an approach to statistical anal-

yses that is based on being explicit about measuring statistical evidence and having

the measure of evidence determine all inferences. One could argue about the ap-

propriateness of the specific measure of evidence chosen, namely, the relative belief

ratio, but if there is a better measure, then by all means it should be used and the

corresponding theory developed. For the author the relative belief ratio is justified

as the measure of evidence by the many nice properties possessed by the associated

theory. The key point is that any theory of inference has to be based on a well-defined

measure of statistical evidence.