Sample Size, Inadequate Power, and Type II Errors

You'll be better able to decipher and critique research reports if you know well the connection between these three concepts [sample size, power, and Type II errors]. I say that because you are bound to come across research reports in which the authors claim that no differences exist or that variables are not related, with these claims based on statistical tests that did not reveal any statistically significant differences or relationships. If the sample sizes in such studies are small, I hope you'll entertain the possibility that the research "findings" are incorrect and represent a Type II error brought about by inadequate statistical power.

(From Chapter 8, pp. 184-185)

Copyright © 2003 Schuyler W. Huck
All rights reserved.