Dear All,
It's nice to see that this is stimulating such a discussion.
I would suggest two minor points:
1. An SR is important because of its systematic nature. The technique by
which you combine the results seem less set in stone.
2. The type of trade off that Jon asks about is implicit in all our
behaviour and is (IMHO) rational. There is always a point when we stop
looking for another piece of evidence. What might make more sense would
be to pause earlier, assess the evidence that we have, and decide
whether to continue. If the evidence we have is unclear, we might choose
to; however, if it is very clear in one direction then we might chose as
ask "what would I have to find to over-turn what I already know, and how
likely is it to exist, given that I haven't yet found it?". That would
then allow us to make some transparent, rational decisions about how to
allocate scare resources (both money and time) - and thus help to answer
Jon's original question about one expensive SR or many cheap ones.
BW,
Matt
P.S. For those who asked for info on our other work, you should have had
it by now - please let me know if not.
|