The Internet Makes You Prefer Slender Women? That’s A Thin Premise

It’s long been said that beauty is in the eye of the beholder. But, according to a new study from the scientific journal PLOS ONE, the physical traits we find attractive may be influenced by the perceptions of the Internet-haver, with thinness in women and masculinity in men outweighing basically everything else. But, is this hard science, or a more insidious trend of misinterpreting data from scientific studies to legitimize foregone conclusions?

In this case, researchers surveyed 18- to 25-year-olds in the developing nation of El Salvador to determine if the digital divide has any impact on survey results regarding aesthetic preferences. They found that Salvadorans without Internet access were reportedly attracted to women with more body fat. Or, as Nature World News' headline put it, “people with Internet access find thin women more attractive.”

So, the Internet must be pushing people to prefer skinny women, right? Well, no. That wasn’t the point of the study at all. Rather, the researchers were aiming to show that in places where people don’t have Internet, scientists should be wary of issuing only an online survey, which can overlook opinions or data from those without Internet access — data that might change the study's results. Researchers tested Internet-havers online and in-person and found that their responses were the same regardless of survey medium: This group preferred more masculine men and thinner women. When they issued the same survey in-person to non-Internet-havers, however, researchers found that responses were markedly different: This group preferred both men and women with softer, rounder features. In other words, issuing only an online survey could exclude data from this segment of the population and skew results. Pretty important to know if you’re a researcher.

For reporters to pick up this study and imply a causal relationship between Internet access and attraction preferences not only confounds the work, but erases a multitude of factors and cultural differences and assumes only two cultures exist: Internet culture, and “other.”

Real talk: Everybody who uses the Internet is different. All you’d have to do is survey your Internet-having friends to know that Internet-havers are not a monolith: Maybe you like scrawny dudes, while your best friends like beefcakes and voluptuous ladies, et cetera. And, it’s not as if Audrey Hepburn and James Dean became global sex symbols only after the advent of 14.4 kbps modems. Moreover, Esther Honig’s Before & After project proves that all over the world, cultural values of attractiveness differ — the Photoshop jobs of 40 artists from different countries vary greatly, and they’re all using the same Internet.

In a world where misinformation abounds — a world where celebrities take to Twitter to prove they’re not dead — it makes sense that we’d turn to scientific studies for evidence-based claims. But, just because a scientist finds a relationship between A and B, that doesn’t mean one caused the other. In fact, some correlations — like this two-day study which concluded that kids who eat chicken on the bone are twice as disobedient and aggressive as other kids — are downright goofy.

The potential for misinformation lies in semantics: Headlines may use words like “linked” ("chicken wing consumption linked to disobedience and aggressive behavior”), but the “link” is merely a correlation, and could even be coincidental. Still, it’s possible that a parent may hear only about this purported link and, without doing her own investigating, forbid her child from eating chicken — or whatever the headline of the day suggests. In the case of the Salvadoran Internet study, the researchers wanted to find out if using the Internet might change the way people answered surveys about attractiveness, but the story got picked up and reported as findings regarding the way Internet influences attraction.

It seems a few researchers and reporters could use a reminder that correlation does not imply causation. There’s even a website full of spurious correlations to drive home this point. Did you know that divorce rates in Maine correlate with the U.S. per capita consumption of margarine? Or that the production of honey is inversely proportional to juvenile arrests for marijuana possession? These true but independent stats serve as great examples that just because data coincide, a causal relationship does not necessarily exist.

Merely observing these relationships yields no conclusive causal value. Saying the Internet influences attraction (and not crediting a complex hegemon of culture, class, and imperialism) is like saying, "If we could just get bees to squeeze out more honey, there would be fewer kids in juvy." Rather than interpreting the Salvadoran study to mean that the Internet influences people to find thinner women more attractive, a deeper look might allow one to hypothesize (not conclude) that in the developing nation of El Salvador, having Internet implies a certain degree of wealth, and that may mean living a more first-world lifestyle, including subscribing to first-world beauty standards.

So, the next time you hear a wonky claim legitimized by a study, employ a healthy dose of skepticism and read the analysis for yourself. It’s usually easy find these studies online, and the best place to start is the abstract, where the background research and hypotheses are introduced in plain English.

Keep your critical thinking skills sharp. Studies suggest this may be in your best interest.

This post was authored by Katie DeRogatis.