Behavior analysts have paid an enormous amount of attention to internal validity―ways to ensure that scientific statements are based upon the data. But the consumption of scientific research is a matter of external validity. The external validity of research does not flow logically from internal validity (despite the arguments of textbooks to the contrary: see Hayes, 1988, for a discussion).【419頁】

I often ask my students a key question the late Aaron Brownstein taught me to view in a different way: Why do we replicate research? Students almost invariably answer that the purpose is to see whether the same result will occur if we did the same thing. This is clearly false. We are not testing the consistency of the universe when we replicate research. If we did exactly the same thing in every detail, the same results would occur. Rather, our purpose is to see whether doing what the author said is doing the same thing. We are testing the functional adequacy of the researcher's verbalizations in guiding our behavior.