There have been many studies of prayer; if you do enough studies some will
come up significant
and get published. Nobody would publish a study showing prayer is
ineffective (they will if
there's a significant claim out that it's effective, but not now). The
elaborate scheme of this
study (group a prays for the patients, group b prays for a, group c prays for
a and b) very
strongly suggests at least 2-3 negative studies in the file drawer. The
first experiment would
be group a only; a negative result would probably get repeated; then you
might add b, then
c.
The observed difference is large compared to psi experiments because the
groups are small.
With small groups and binary outcomes huge differences are expected by chance
(remember,
there's multiple studies and the published ones are the best). In psi
experiments you can
repeat ad nauseam and the law of large numbers gets rid of the large
deviations.
This archive was generated by hypermail 2b30 : Sat May 11 2002 - 17:44:11 MDT