The Importance of the Research Question(s)

Our first warning [in Chapter 21] concerns the quality of the research question(s) associated with the study you find yourself examining. If the study focuses on a trivial topic, no statistical procedure has the ability to "turn a sow's ear into a silk purse." This is as true of nonparametric procedures as it is of the parametric techniques discussed earlier in the book. Accordingly, we once again urge you to refrain from using data-based p-levels as the criterion for assessing the worth of empirical investigations.

(From Chapter 21, pp. 673-674)

Copyright © 2003 Schuyler W. Huck
All rights reserved.