My Four Big Mistakes

Scientists hate making mistakes. Faced with the possibility that there is a mistake in our data or that someone else has tried and failed to replicate one of our published studies, we immediately become defensive. We search our data to reassure ourselves that nothing is in fact seriously wrong. We scrutinize the other laboratory’s paper, looking for signs that they are the ones who have made the mistakes.

Source: Shutterstock

When all those attempts at proving ourselves not mistaken fail, we struggle with what to do next. Should we send in a correction to the journal that has published our paper? How embarrassing is that? Too many corrections and maybe no one will believe anything we do is correct. Perhaps journals will refuse to publish any of our papers in the future.

Amidst this anxiety and paranoia, we imagine our scientific careers going down the drain. We are faced with turning ourselves in, something very few willingly do.

I had these thoughts in mind when recently, within just a 48-hour window of time, I was confronted with four instances in which I seemed to be mistaken, three of which involve scientific pronouncements I have made repeatedly. As none of them relate to any of my own data or to papers I have published, there is no need for me to issue any corrections or retractions. Rather, these putative errors forced me to confront the emotional upheaval a scientist faces when he or she is proven wrong.

Here are four of my pronouncements I subsequently learned may be wrong:

1. Following a serious traumatic event, post-traumatic stress disorder (PTSD) is not the most common psychiatric disorder to develop. Depression is more common.
2. In making decisions, we often default to the use of our “emotional” brain, sometimes called “Type 1 thinking,” and do not engage the “rational brain,” referred to as “Type 2 thinking.” Our biases are the results of less effortful emotional thinking trumping more fatiguing analytical thinking.
3. There are no health risks associated with genetically modified foods, often referred to as GMOs.
4. The 1951 New York Yankees beat the New York Giants in the World Series in four straight games.

Where did I go wrong?

First, a paper crossed my desk published in the American Journal of Public Health in which researchers described examining data from 811 survivors of 10 disasters. The paper reported that “PTSD was the most prevalent postdisaster disorder (20%), followed in frequency by major depression (16%) and alcohol use disorder (9%).”

Next, I received my monthly copy of the Journal of Neuroscience and noticed an article titled “Reason’s enemy is not emotion: engagement of cognitive control networks explains biases in gain/loss framing.” This functional magnetic resonance imaging (fMRI) study of decisions that involve risk showed that the conflict is not between emotion and reason. Instead, such decisions involve a network called the “resting brain” and not the network usually associated with emotion. Rather than being a battle between emotion and reason, the researchers show it is one of “less cognitive effort versus more cognitive effort.”

My third mistake came to light when I read that Dr. Paul Winchester of Indiana’s Rile Hospital for Children was about to announce findings that 90% of pregnant women in one obstetrics/gynecology practice had detectable levels of the herbicide glyphosate (Monsanto’s Roundup) in their urine and that the level of glyphosate was correlated with risk for preterm delivery, low birth weight, or both. Thus, GMOs may not be as entirely benign to human health as I had often asserted.

Finally, as a diehard New York Yankee fan, I love taunting Giants fans, whose greatest moment was the 1951 home run hit by Bobby Thompson, called the “Shot Heard Around the World,” that won the Giants the National League pennant in dramatic come-from-behind fashion. Big deal, I used to say, the Yankees destroyed them in four straight games in the subsequent World Series. Only it was a six-game series, it turns out.

The last of these errors is an obvious fulfillment of one baseball fan’s fantasy life and as such is entirely harmless and understandable. Hence, I will not apologize for it, but in the future will be more careful in discussing that 1951 World Series that the Yankees nevertheless did win.

But the other three are of course more consequential and therefore I tried mightily to figure out a way to redeem my previous pronouncements, even in the face of conflicting evidence. I am sure, for example, that I heard a very respected PTSD researcher say that her work showed that depression is more common after a traumatic event than PTSD. In the North et al study, the difference in incidence is very small (20% versus 16%), so another study might find the opposite effect. It is, after all, only one study.

Post-Traumatic Stress Disorder Essential Reads

Taming the Amygdala in PTSD

What Happens In the Brain During PTSD?

With respect to the Journal of Neuroscience paper, if it is correct, then I am in good company in being wrong, accompanied for example by Nobel laureate Daniel Kahneman. In fact, the authors acknowledge that “Psychologists have long described human experience as comprising two dueling modes of thought: one process of quick emotion-laden association and another of reasoned analysis” and cite no less a luminary than the father of modern psychology, William James. So maybe it’s the new paper that’s wrong.

And I only ever said that eating GMO foods is not harmful. Glyphosate is not a GMO but rather an herbicide. I never claimed that consuming it was safe. Maybe someone should tell farmers not to use so much of it, but isn’t that a different issue? And in any event, I could always cite the mantra that “correlation is not causation” and challenge the notion that the herbicide causes premature birth.

Whether these excuses are robust or not is beside the point: the amount of effort I put into avoiding acknowledging that I might be wrong is what’s most important here. What I could not face is that new science may be rendering old ideas incomplete or even wrong. And that is not a very healthy or scientific attitude.

The problem goes beyond my own potential loss of credibility. When science changes its mind, it leads people to think that scientists are not trustworthy. This fuels science denial. Why believe that vaccines and ECT are safe, that climate change is the result of human activities, or that personal gun ownership is dangerous when four things I firmly believed can be proven potentially wrong in the space of two days? If scientists could be so wrong about fat, rather than sugar, being a risk factor for heart disease, for example, then why not dismiss their reassurances that the level of bisphenol A (BPA) in plastic bottles is too low to be a health risk or that the recommended schedule of vaccines for children cannot possibly “overwhelm” the immune system. Perhaps these assertions too will turn out to be wrong.

I am sure I will never be forced to change my mind about vaccines, ECT, climate change, or guns and yet I may have to about PTSD, the parts of the brain that are active in decision-making, and even that GMOs are unassociated with any health risks, if further studies confirm the three new findings discussed here. I have certainly changed my mind about the number of games in the 1951 World Series.

The important issue, for which we have no immediate solution but will grapple to find one, is how do we convey to the public that science is a dynamic enterprise in which new data, often made possible by new technologies, may modify our previous conclusions? This does not mean that there are no scientific facts. Cigarette smoking causes cancer—that one will never change. Seat belts save lives—that’s a fact too.

The scientific community needs to figure out how to stop punishing its members when new information challenges their published findings. All of us need to stop being “married” to our data and recognize that we are prone to making mistakes and that new research may prove us wrong. That is why I am taking this step of unashamedly acknowledging I may be wrong about some of the things I have insisted on in the past. It actually feels pretty good.