Ex post facto hypothesizing, or the development of hypotheses
“predicting” relationships that have already been observed, is invalid in
science, because it’s impossible to disconfirm such hypotheses. While nothing
prevents us from suggesting reasons that observed relationships may be the
way they are, we should not frame those reasons in the form of “hypotheses.”
More important, one observed relationship and possible reasons for it may
suggest hypotheses about other relationships that have not been examined.
The elaboration model is an excellent logical device for this kind of unfolding
analysis of data.