I have been thinking for a (short) while about EBM. It would seem that there
are several strands to EBM. Among these, there might be an overarching idea
that we should ask whether what we are doing is any good, and another strand
about what formal methods we use in EBM (largely statistical and Bayesian
ones), as well as other strands.
To take just these two, though, could we seperate these them? Are there
other formalisms that might work instead of a Bayesian approach, for
example. If so, does anybody have any experience of using them, or any ideas
about what the advantages or disadvantages might be?
As usual, I shall compile the answers for the list.