What Studies Show — and Don’t
The presuppositions of studies merit scrutiny
Hummingbirds are wonderful creatures. Happily, their range is expanding. Studies show this expansion!
Of course, we often read of studies that reach less welcome conclusions. Studies have shown that fish oil does little to lower cholesterol; they’ve shown that aspirin doesn’t reduce heart attacks. New studies cast doubt on the salubrious nature of red wine. Will coffee take the next hit? Who knows? After all, studies show that spending time on social media can result in anti-social behavior.
These bothersome studies lead me to speculate that someday studies will show that…most studies don’t show what they purport to show! Of course, were it to happen, that study would be doubtful as well. Paradox would rule the roost!
Maybe we’ll escape such paradox. But the presuppositions of studies, new and old, merit scrutiny. For a start, we need to ask: (a) who did the study, (b) why, and (c) under what conditions. Such questions prompt others. Who will review the accumulated studies, and for what purpose? Will there be need for another group of reviewers? (Compare Plato’s question in The Republic: Who will watch the watchmen?)
In addition, we need to ask whether what’s studied is rightly understood. For example, what counts as “suicide”? Emile Durkheim, the founder of sociology, thought that anyone who undertakes an action foreseeing that it will cause his or her death commits suicide. By that measure, people who risk their lives to save their friends commit suicide. Speaking of heroes, pace Durkheim, what might a study tell us about one who acts on conscience? Not much unless we are clear about what conscience is. A last example: what might studies show about what is it to “love” someone? Not much unless we know what love is. Is it a strong affection, accompanied by physical attraction? Or is love the willing of the good of another and the steady acting to bring about that good? Or is it something else?
Note, too, that empirical studies have logical limits. Such studies don’t, because they can’t, provide evidence for logical truths. For example, such studies neither confirm nor deny the principle of transitivity: if x is greater than y, and y is greater than z, then x is greater than z. Nor can any such study give evidence for or against the criteria for evaluating a study. Such criteria, rather, are already in place.
Nonetheless, I remain a friend of empirical studies and of the inductive reasoning on which they are based. But my friendship is conditional. Such studies, for example, cannot give us fresh evidence that 7 + 5 = 12. Nor can they show me that I can trust my senses. I must rely on my senses to access the studies that would purport to do so. That said, I am pleased to report that I have observed of late that hummingbirds now visit both my front yard and back yard. Anecdotal evidence only? Not for me, it isn’t.