Practical vs. Statistical Significance in Mixed ANOVAs

Turning to the distinction between practical and statistical significance, researchers have a variety of techniques at their disposal for assessing the degree to which statistically significant results are nontrivial. When you come across a study in which one of these techniques has been used in conjunction with a three-way mixed ANOVA, give the researcher(s) credit for going beyond the basics in analyzing the study's data. In Excerpts 17.29 and 17.30 [not shown here], we see two examples where this was done. In the first of the excerpts, the researchers conducted a power analysis to determine whether their study had sufficient protection against Type II errors. In Excerpt 17.30, eta squared was computed, after the study's data had been collected, to measure how much variability was explained by the main and interaction effects that turned out to be statistically significant.

(From Chapter 17, p. 528)

Copyright © 2003 Schuyler W. Huck
All rights reserved.