The Importance of the Null Hypothesis

This skill [of being able to discern what a researcher's null hypothesis was] is important to have because the final decision of hypothesis testing always has reference to the point of departure. Researchers never end up by rejecting (or failing to reject) in the abstract; instead, they always will terminate the hypothesis testing procedure by rejecting (or failing to reject) a specific H_o. Accordingly, no decision to reject should be viewed as important unless we consider what specifically has been rejected.

On occasion, the hypothesis testing procedure is used to evaluate a null hypothesis that could have been rejected, or not rejected, from the very beginning, strictly on the basis of common sense. Although it is statistically possible to test such an H_o, no real discovery is made by rejecting something that was known to be false from the outset, or by reaching a fail-to-reject decision when such an outcome was guaranteed from the start.

I cannot exaggerate the importance of the null hypothesis to the potential meaningfulness of results that come from someone using the hypothesis testing procedure. Remember that a reject or fail-to-reject decision, by itself, is not indicative of a useful finding. A reject decision could be easily brought about simply by setting up, in Step 1 [of the hypothesis testing procedure], an outrageous H_o; a fail-to-reject decision could be just as easily brough about by comparing two or more things that are known to be the same. Consequently, you should always be interested in not only the ultimate decision reached at the end of the hypothesis testing procedure but also the target of that decision--H_o.

(From Chapter 7 of the 6^th edition, pp. 155-156)

Copyright © 2012

Schuyler W. Huck
All rights reserved.