What Behavioral Scientists Are Unwilling to Accept
Meehl (1986) published a brilliant paper with the title “What Social Scientists Don’t Understand.” I believe that that a better characterization of the situation is to broaden the reference class to include behavioral scientists in general. Also, I believe that most behavioral scientists do understand these issues at some level, but that they are not willing to accept the implications that an explicit understanding would force.
Concern has been expressed, for the past 40 years or so, that the strategies of research design and statistical analysis used by behavioral scientists are woefully inadequate to support a progressive scientific enterprise. In this article I will summarize the nature of some of these concerns, and will identify some of the impediments that they impose to the development of progressive conceptual frameworks adequate to the task of achieving an understanding of the behavior of complex organisms in their environment. Although there has been little disagreement regarding the soundness of the methodological criticisms that have been made, there is little reason to believe that the methodological and statistical criticisms have had any great impact on the activities of either journal editors or research scientists. A review of these concerns is appropriate at this time because a number of articles have appeared recently that defend the status quo, and which are based either on faulty premises or a questionable view of the problems that impede scientific progress. And, hopefully, yet another critique might enhance the rate at which we arrive at a more satisfying state of affairs.
Keywords: Context of Discovery, Context of Justification, Alternative Hypotheses, Multitrait-Multimethod, Validity, Null Hypothesis Testing