Here's how a terrible study decided that liking your coffee black makes you a psychopath

If you have a Google alert set up for “coffee” or “psychopath,” you might have seen some surprising headlines lately. The Independent announced, “How you drink your coffee ‘could point to psychopathic tendencies.’”

Jezebel proclaimed “Study Says People Who Take Coffee Black Have Psychopathic Tendencies.” Delish’s variation was, “If You Drink Black Coffee, Experts Say You Might Be a Psychopath.”

If you take your coffee black, these headlines may have alarmed you. Could your garden-variety narcissism, coupled with your coffee snobbery, actually be a sign of something much worse? Will drinking coffee black come to have such social stigma that you’ll have to start ordering vanilla lattes if you want people to trust you?

I have great news for black coffee drinkers, and everyone else: The study in question does not actually say that people who take their coffee black have psychopathic tendencies. The researchers who wrote the study didn’t actually ask anyone about whether they take their coffee black. And what the study does say about the relationship between taste preferences and personality traits is probably meaningless and worth ignoring.

The authors of the study, which has been accepted for publication in the journal Appetite, surveyed a total of 953 people recruited on Amazon’s Mechanical Turkplatform about how much they like specific foods (including coffee); how much they like bitter, sweet, sour, and salty foods in general; and how much they agreed with a total of 52 statements from various personality tests.

They found that people who said they liked bitter foods in general were very, very slightly more likely to have traits associated with psychopathy, narcissism, sadism, and Machiavellianism. However, there was no correlation between people liking specific bitter foods (e.g., coffee) and having these “dark” traits. How could that be?

Apparently, no one can agree on what constitutes a bitter food. When the researchers asked participants to assess the flavors of specific foods, they found that participants disagreed with the researchers over whether certain foods were bitter. The researchers thought cottage cheese, ginger ale, grapefruit juice, rye bread, and tea were bitter; participants did not. Whoops! Perhaps this explains why there was a less-than-50-percent correlation between participants saying they liked bitter foods in general and saying they liked the specific bitter foods on the list.

You might think that the lack of consensus over what “bitter” means would send the authors back to the drawing board to design a better study. Instead, the authors forged ahead, ignoring the fact that the majority of relationships between taste preferences and personality traits were statistically insignificant.

Even the strongest correlation touted by the study is pretty weak: The researchers found that liking bitter foods is predictive of—at the very best—19 percent of someone’s sadism. There was a much stronger correlation between saying you like bitter foods and saying you like sour foods than there was between saying you like bitter foods and having any particular personality trait.

We should also consider the possibility that the participants didn’t take the study very seriously. (And why should they? In accordance with Mechanical Turk’s ridiculously low rates, they were paid only $0.60 to $1 for their participation.) Self-reporting is notoriously unreliable, and the study included a bunch of personality-test questions that I, personally, would have no idea how to answer.

For instance, participants were asked to rate this statement on a scale from 1 (extremely uncharacteristic of me) to 5 (extremely characteristic of me): “I often find myself disagreeing with people.” What constitutes disagreement? What constitutes “often?” Does “people” include Donald Trump or just people I actually know and hang out with? If I were being paid $0.60 to answer 52 maddeningly vague questions like this, I’d probably just mark a number arbitrarily and move on as quickly as possible.

Should this study even exist? Far be it from me to tell any scientist that she shouldn’t pursue her dreams, even if those dreams involve trying to measure a relationship between taste preferences and personality traits that is probably not real. What I doknow is that the only reason anyone in the media is covering this study—and framing it as a referendum on black-coffee drinkers’ mental health—is that people love reading about the ostensible dangers of coffee, even if they’re not rooted in solid science. It’s enough to leave a bitter taste in your mouth.