Practical vs. Statistical Significance

At various points in this chapter, I tried to help you understand that statistical significance does not always signify practical significance. I did this by means of the words I have written and the excerpts I have chosen to include.

I hope you will remember two things. First, a very small p may indicate that nothing big was discovered, only that a big sample can make molehills look like mountains, and, conversely, a large p may indicate that something big was left undetected because the sample size was too small. The second thing to remember is that studies can be planned, using the nine-step version of hypothesis testing, such that neither of those two possible problems is likely to occur.

(From Chapter 13 in the 6^th edition, p. 340)

Copyright © 2012

Schuyler W. Huck
All rights reserved.