The Mistake of "Accepting the Null"

Almost all researchers who engage in hypothesis testing have been taught that it is improper to conclude that a null hypothesis is true simply because the hypothesis testing procedure leads to a fail-to-reject decision. Nevertheless, many of these same researchers use language in their research reports suggesting that they have completely forgotten that a fail-to-reject decision does not logically permit one to leave a study believing that the tested H_o is true. In your review of studies that utilize two-way ANOVAs (or, for that matter, any procedure for testing null hypotheses), remain vigilant to erroneous statements as to what a nonsignificant finding means.

(From Chapter 13 in the 6^th edition, p. 310)

Copyright © 2012

Schuyler W. Huck
All rights reserved.