Subject: The Cult of Statistical Significance

Posted by Richard Hake on 10/14/2010

Message:

Some subscribers to K-12SD might be interested in a recent post "The Cult of Statistical Significance" [Hake (2010)]. The abstract reads:

******************************************
ABSTRACT: Math-Teach's Domenico Rosa has called attention to the review of Ziliak & McCloskey (2008) by Olle Häggström (2010). According to Häggström, Ziliak & McCloskey's major point is that "many researchers are so obsessed with statistical significance that they neglect to ask themselves whether the detected discrepancies are large enough to be of any subject-matter significance."

Consistent with that outlook, in "Lessons from the Physics Education Reform Effort" [Hake (2002)] I cited the position of many psychologists and biologists that the "effect size" is a preferred alternative (or at least addition) to the usually inappropriate t-tests and p values associated with Null Hypothesis Statistical Significance Testing (NHSST).

Nevertheless, many educational researchers (even some physicists) still utilize *only* NHSST to gauge the significance of their research results.
*****************************************

To access the complete 14 kB post please click on .

Follow Ups:

The Cult of Statistical Significance - Bob Gorman 10/14/2010