Aiming for Accuracy (Even When the Truth Hurts)

John V. Kane

Books and movies often ask us to “suspend our disbelief.” If we don’t set reality aside for a moment, if we don’t indulge the fantasy a little bit, we’re not going to enjoy the story very much. We set the truth aside so that we can have a good time and not get too hung up on details. It’s a good approach to a work of fiction.

We need an opposite approach when learning about a new study’s results (or even just hearing someone armchair theorizing). We need to suspend belief. That’s because social science research isn’t like fiction. We’re trying to learn something real, not create a story or be entertained by one (though a silly pun here and there wouldn’t kill us, right?).

We can be highly motivated to either believe or disbelieve a particular claim or finding. But what we actually need to do is evaluate the evidence. We start by treating that claim or finding—no matter how much we may love it or loathe it—as a working idea only, a mere possibility and nothing more.

But as with many things in life, this is all much easier said than done. How can we resist the urge to reason in a particular direction? How can we avoid believing or dismissing a statement on a topic we really care about, simply on the basis of whether it comports with what we want to be true? The answer lies in motivation.

Why Wanting to Be Right Can Be a Good Thing

We’ve all met people who seemingly need to be right all the time (believe it or not, there’s more than a couple social scientists who are like this). Heck, sometimes we even do it ourselves (but that’s only because we’re actually right, am I right?).

Normally, such behavior is awkward and exhausting for everyone involved. But when it takes the form of wanting to be accurate about a particular claim, it can actually be a good thing.

This is the second fundamental kind of motivation that guides our reasoning: the motivation to be accurate. Sure, we can (and do) reason in a very biased fashion. But we also don’t like the idea of being knowingly incorrect about something, of being willfully wrong. And in that small bit of insight lies a giant ray of hope: If we desire to be factually correct, then perhaps there are limits on how directionally biased we can be. Perhaps not in every case, and perhaps with some resistance, but when the evidence becomes clear enough, many of us will often relent, however passive-aggressively it might come out—“OK, OK, fine. I was wrong! Happy now?!”

Accuracy: Just Remember

A motivation to be accurate is a great thing. So why does it seem like we’re still allowing the opposing motivation—directionally motivated reasoning—to dominate?

First, the context matters a lot. When we’re reasoning about topics that mean a lot to us, that are central to our worldview and who we are as a person (political and religious beliefs and identities, for example), directionally motivated reasoning can wield some truly gargantuan strength.

But as Flynn, Nyhan, and Reifler note, when (say) purchasing an appliance for our home, our motivation looks a lot different. In such a situation, we are motivated to choose on the basis of the facts, not our emotional impulses. When faced with that kind of goal, people will be “relatively evenhanded in their search for information and in how they evaluate the information they encounter.” Indeed, research has found quite strong evidence that prompting people to read information in an evenhanded, unbiased fashion can dramatically offset (and even eliminate) the effects of directionally motivated reasoning.

So, given the right context, the accuracy motivation can hold its own against the Goliath that is directionally motivated reasoning. Often, we just forget to do it. That’s it. Something catches our attention that epically aligns with—or crashes up against—what we believe, and it is in that turbulent moment that we simply forget to be accurate. Contrary to popular belief, we’re not doomed to be biased all the time. We can be accurate—we just need to remember to care about it.

Beat Back Bias: Vibe Checks, Thought Experiments, and a New Identity

The accuracy motivation is an indispensable travel item for our trip along the Road to Evidence (R2E). But in many cases it isn’t enough to prevent motivated reasoning. Evidence in social science is never perfect. The tests we do are limited in number and scope. If the two poles of the continuum are occupied by “unabashed nonsense” and “incontrovertible proof,” social science research will always live somewhere in-between (though hopefully closer to the latter).

This unfortunately opens the door to a big problem. Evidence is almost always “better than nothing,” yet it is also “not perfect.” This nebulousness allows motivated reasoning to resume control of the steering wheel. When findings comport with what we want to be true, any imperfections in the research methodology can be written off as trivial, as mere trifles that in no way undermine the larger merits of the study. And yet, should the same findings not sit well with us, our directional motivations (that is, our biases) may regard those same imperfections as disqualifying, as glaring flaws that throw the entire study into question.

What else can we do to beat back our biases?

First, we can focus on how something makes us feel. As Carl Sagan warned, “The more we want it to be true, the more careful we have to be.” When we care deeply about the topic, it won’t be hard for emotions to get activated once we hear a claim or read a result. Perhaps we feel validated, perhaps offended, perhaps vindicated, perhaps upset, perhaps enraged. And if you’ve ever tried to engage in careful reasoning or evaluation of information while in a highly emotional state, you probably know how it usually turns out (not great).

Again, we’re human beings, and feeling an emotional response is probably unavoidable. But we can avoid allowing the emotion to hijack our reasoning. We can take a few moments (or minutes, or days—it’s your life!) to identify how a claim or finding is making us feel before we start trying to evaluate it evenhandedly. Whether we’re involuntarily shaking our head in horror, barely able to refrain from shouting out a string of expletives, or vigorously nodding in agreement—it’s a big red flag. Our feelings are now in position to take over. We risk being too dismissive, or too accepting, simply because it feels good. As the economist and journalist Tim Harford writes, “Today’s persuaders don’t want you to stop and think. They want you to hurry up and feel. Don’t be rushed.”

In the interest of being accurate, once we recognize that a claim or finding is making us very happy or very sad or very angry, we can simply slow down a bit. The less emotion we bring to the table, the more we can treat that claim or finding like an appliance on a webpage, not something personal. As Sherlock Holmes himself said, “The emotional qualities are antagonistic to clear reasoning.”

Another line of defense is to engage in thought experiments that can help us reveal—to ourselves—whether we’re being biased. A particularly useful one for thinking about social science research is the “selective skeptic test.” (Seriously, I do this one all the time.) The idea here is to simply pretend that the results are the opposite of what they actually are. What this helps to ensure is that any reactions we might have about a study are not simply reactions to the results, masquerading as reactions to the scientific merit of the study. If we know that our skepticism would be just as strong if the results were in the opposite direction, we can be more confident that our skepticism is sincere, not directionally motivated.

Finally, because how we view ourselves matters so much in how we react to information, one of the most powerful antidotes to bias is to take on a new identity—the identity of scientist. Why would adopting such a label matter? Because this identity, as psychologists Packer and Van Bavel note, “comes with strong expectations that we will use empirical analysis to arrive at verifiable conclusions.” These expectations, or norms, have power—power that can offset the formidable temptations of bias.

In a similar vein, political scientist Philip Tetlock discusses people who adopt the identity of “superforecasters.” Such people pride themselves on keeping an open mind, on evenhandedly looking at information to accurately predict future events. For them, all that ultimately matters is predicting correctly—even when, personally, they might dread their prediction coming true. Similarly, Julia Galef advocates adopting a “scout identity”—that is, one whose goal is (like a scout) to form accurate beliefs, rather than reflexively defend existing beliefs, to see the situation as clearly as possible, even if it doesn’t make one’s “team” look good.

All of these various labels (scientist, superforecaster, scout) all have something in common (besides just starting with “s”). They imply a commitment to “getting it right” over “feeling good about it.” They imply an addiction to truth.

So choose whichever label you like (or make up your own!)—as long as you start seeing yourself through that lens, you are taking the single most consequential first step.

With practice, we can strengthen this new identity. This is crucial because, as Packer and Van Bavel write, “The more that someone identifies with a group, the more strongly he or she will tend to conform to that group’s norms.” To accomplish this, we can ask ourselves things like “OK, I personally don’t like this finding, but what would a scientist do in this situation?” Again, the goal is to adopt a new mindset so that we adhere to a new set of norms. It means changing not just what we want to do, but also who we are. If we are successful, we will offset the identity-based forces (e.g., the social and political groups we affiliate with) and personal forces (e.g., the beliefs we’ve held and expressed, perhaps even publicly, for a long time) that constantly tug us toward what we want to be true rather than what is true.

Adopting this new persona as a scientist, seeing ourselves as part of a larger community, charged with a mission to accurately reveal hidden truth in the social world—this is the safest way to begin the journey along the R2E. Once we’ve internalized such an identity, the act of changing our mind in light of new evidence transforms from an admission of error into a point of pride. It becomes a virtue, not a vice. A sign of strength, not of weakness.

Final Points

In any instance wherein we hear a claim being made, even by experts, the first step is asking for evidence. In many cases, we’ll find that the claim is based on essentially nothing. It was a well-disguised hunch; in the parlance of our current moment, it was a confidently delivered vibe. At other times, our wish may be granted. We will come face-to-face with some form of evidence. The question then becomes how to evaluate its merits.

But not all “evidence” is created equal. Once we are motivated to learn the truth—even when the truth might hurt—and once we have some strategies for keeping our own biases at bay, we will ultimately need to know what makes some evidence strong and some not. This, dear reader, we can learn only by taking our first brave steps along the Road to Evidence.

***

John V. Kane is Associate Professor at the Center for Global Affairs at New York University, where he teaches courses on research methods, applied statistics, and public opinion. He has written for top peer-reviewed journals, as well as for the New York Times National Public Radio, and elsewhere.