The member of the Zaring family who studies such matters passes along this editorial, which yes, indeed, "prior to publication, authors will have to remove all vestiges of the NHSTP [null hypothesis significance testing procedure] (p-values, t-values, F-values, statements about ‘‘significant’’ differences or lack thereof, and so on)."  The idea is that this sort of testing is less useful in psychology research than is effects testing, given that it's a largely experimental field that often doesn't have huge sample sizes (though huge sample sizes present their own problems of spurious p-values).  It doesn't seem to have led to a stampede away from significance testing yet, but there's an interesting overview of the controversy here.  And I might as well link to the research on which a couple of my colleagues will be dining out for years, demonstrating the suspicious ease with which it is possible to find statistical significance in lots of cases.  And here's a rant.  My sense is that there is very little attention paid in empirical legal research to this issue; Dave Hoffman mulled this over last month, and said that at CELS, every single paper he saw came with a p-value.

Empirical Legal Studies | Bookmark

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345157d569e201b8d1099ddc970c

Links to weblogs that reference Psychology Journal Forbids Significance Testing: