What the Ganzfeld Database Needs.

This is what the Ganzfeld needs. This is what parapsychologists still have to do:

(1) Apply stringent, well thought out, fair exclusion and inclusion criteria and apply it to the entire database. This will lead to database (A).
(2) Apply independent heterogeneity (standardness) criteria to database (A), resulting in database (B). Bem's criteria will do fine, as long as it's independently applied.
(3) Slice the database's (A) and (B) into post-joint communique and pre-joint communique databases, resulting in (A)#1, (A)#2, (B)#1, (B)#2.
(4) Calculate the chance-odds for all four databases (p-values).
(5) Calculate the file-drawer requirement (Rosenthal's N) in an attempt to either outrule or suggest as possible publication bias or the file drawer effect as contributing in any significant way to the results.
(6) Create polynomial AND linear regressions for the Effect Sizes for the four databases outlined in (3) to outrule or support the hypothesis that methodological flaw is contributing in any significant way to the chance deviations.
(7) Make conclusions, or be very very close to one.


(1), (2) have been done for sporadic portions of the database.
(3) hasn't been done properly, thus (4) hasn't been done appropriately
(6) has never been done, except for Bierman's flawed attempt (only captured about 1/2 of the database, is thus flawed).
(7) has been attempted, but since (1) (2) (3) (4) (6) haven't been done either at all or to an appropriate extent, any attempt at (7) is possible but at the same time iffy - it'll be speculation of a conclusion, not an actual conclusion.