This third round of essays contained some doosies—ideas risky enough to make any scientist cringe. A few truly jarring essays overshadowed this section and the subsequent book club discussion. The contentious essays were (in some people’s opinions) weakly argued and dangerous, and definitely deserving of a rebuttal:
Gary Klein boldly claims that Evidence-Based Medicine must die. This is a worrisome idea. Evidence-based medicine came about to rigorously test medical interventions in a standardized method to determine what medicines work, what medicines may be dangerous, and what medicines don’t really do anything. Without evidence guiding medical decisions, we are back to days without antibiotics, vaccines, chemotherapy, and other life-saving inventions. Klein argues that evidence-based medicine is not perfect because not every study can be replicated and, in a few extreme cases, results of clinical trials have been faked. Well, as a fellow book-clubber exclaimed, this essay is “trying to throw the baby out with the bathwater.” Yeah, science is not perfect. Nothing is. That is no reason to kill it, but more a reason to demand increases in rigor and integrity. Asking to kill evidence-based medicine because it isn’t perfect is reckless, and could dismantle trust in a medical system that strives to save lives.
In a similarly questionable essay, Dean Ornish tries to nix the idea of Large Randomized Control Trials. He calls out how some studies are poorly designed, which some studies are. In my opinion, poorly designed trials must die. Ornish takes his opinion to the extreme in stating that all large randomized control trials must die. The loose logic of his essays seems intended to make small studies about behavioral interventions appear stronger. Maybe the studies he mentioned just need a better design.
Tom Griffiths tries to argue that bias can be a good thing in his essay, Bias is Always Bad. Bias is bad. It makes for bad research questions, fraudulent and incorrect data interpretations, and a milieu of societal problems. Griffiths, however, is talking about a different kind of “bias” where it is used as a way to process digital images. Basically, he’s arguing that bias isn’t bad because some people have a different definition of bias. It’s a misleading argument, which may be intended to ruffle feathers with a flashy title. But the substance of the essay is arguing semantics, not thought-provoking ideas.
The last essay to which I will openly dissent is Richard Nisbett’s Multiple regression as a means of discovering causality. He argues that a statistical technique, multiple regression, is limited. And it is. Multiple regression is designed to determine which factors are correlated. Nisbett has a problem with people misusing this statistical technique. Well, that’s pretty obvious…Correlation doesn’t equal causation. All scientists know this. Bad scientists do abuse this. But the misuse of a statistical tool does not mean the tool needs retired (as the content of his essay argues); it means the misuse needs retired.
This reading section did contain some wonderful essays, too! Jamil Zaki’s The Altruism Hierarchy was delightful. Basically, it argues that the back-and-forth surrounding the meaning of altruism is trivial. Not only is figuring out a hierarchy of altruism “logically self-negating,” but it is “morally self-negating.” Zaki expresses frustration about how the science surrounding altruism strips the humanity from it, saying how “it’s profound and downright beautiful to think that our core emotional makeup can be tuned towards others, causing us to feel good when we do.” Honest and emotional human insights coming from a scientist like Zaki can hearten fellow scientists and humans, and I’m glad I read his piece.
Ian McEwan questioned the Edge question in his piece entitled Beware of arrogance! Retire nothing! The short, humorous, and poetic prose elegantly frames how even bad ideas need preserved, because that’s how science progresses. We learn from our mistakes, and it is dangerous to negate old ideas as meaningless.
Lastly, Robot companions by Sherry Turkle challenges us not to fight the developments of Artificial Intelligence, but to truly consider how we want robots to serve us. Do we really want to create a machine intended for companionship or love? Turkle “see[s] us on a voyage of forgetting.” As we embrace technological advancement, we must consider how a genuine, tenuous, and beautiful human experience shapes our relationships with others. Although AI geeks love to argue the seemingly limitless potential of robots to mimic humanity, robots will never truly have humanity.
Once again, this section provided plenty to discuss at our meeting—with equal amounts of frustration and intrigue (although surrounding different essays). It’s interesting, and maybe terrifying, to see the logical shortcomings of supposed scientific thought-experts in our society laid so bare. Discussing blatantly radical ideas does force reflection, both on the ideas and the identity of the authors. At the end of the day, these esteemed thinkers are simply humans, with ideas with which we may refute. Only time will tell whose ideas will die, but I for one really hope that evidence-based medicine and large randomized control trials live.
– Sam Tucci