All posts by matt

The real winner of Tinder vs Vanity Fair is science education

Recently, Tinder has a public reaction on Twitter to an article in Vanity Fair. The brief gist is that the Vanity Fair writer, Nancy Jo Sales, wrote about Tinder as contributing to a negative hookup culture. Tinder wrote a long series of tweets in response about all the good, positive interactions people have on their platform. Notice how I have left out all the judgment words (don’t worry, it won’t last). But I did so deliberately because regardless of what you think of Tinder, hookup cultures, or Nancy Jo Sales’ reporting, the real winner here should be science education.

Judgement time. The Vanity Fair piece was actually not very good (despite Wired referring to it as “excellent”). The Tinder response was also not very good. But they were both not very good for the same reason: selection bias. And focusing on how that bias plays out is actually where some use might come out of this public meltdown.

In the VF piece, the interviewees seem to almost all be young men and women who are actively using Tinder. The women seem to think it contributes to a hookup culture that devalues them; the men seem to agree. Yet therein lies the key problem: by talking only to people who are currently using the app, Sales is actually interviewing the “failures”, also known as a negative selection bias.

If the point of Tinder is for people to find long-term romantic relationships, than the majority of the people on the app will be people who are not yet successful at doing so. Thus, if you want to explore the potential effect of the app on finding a mate, you have to interview people who are no longer on the app and find out why they left. Is it because it was all hookups and they tired of it? Or is it because they no longer need the app because they found someone?

Tinder seems to be trying to point this out in their Twitter rant, but they go too far and end up falling victim to the other end of the spectrum, the positive selection bias (sometimes referred to as the survivorship bias). They mostly focus on people who should have theoretically left their platform: marriages. A “shit ton of marriages”, according to them.

If you throw enough people at a situation, usually some of them are going to come out with positive results. But if you only look at those with positive results, you ignore the disappeared failures who either left and didn’t have positive results or are still around and failing. Tinder specifically calls out how they are relying on the stories users tell them, which they call #SwipeRight stories.  Given the interaction pattern of the app, that presumably means just the positive ones.

The real pity is that avoiding the positive and negative selection bias should have been relatively easy. You could interview people in all four quadrants (current user/positive, current user/negative, ex user/positive, ex user/negative). Even this is a sort of selection bias, however: there are plenty of people who get married or have bad relationships entirely without the aid of technology. Tinder is taking credit for the marriages, Sales is blaming them for the hookups, but in reality both happen in populations that have never touched Tinder. So what you really want is to understand how the populations that use Tinder and those that don’t differ. And even then, you can’t blame that difference entirely on Tinder. After all, they might have been different to start with in a way that made them choose to be on the platform or not.

Now clearly, journalism isn’t science and Twitter certainly isn’t science. And we can debate whether either of those two mediums have a responsibility to uphold some sort of ethics. But science education means that we don’t have to rely on the ethics of others. All the spectators reading the article and Tweets, with a bit of STEM education, can actually recognize the biases for themselves and avoid perpetuating them. And I sincerely hope that someone uses this debacle as a reason to increase science education funding. Wouldn’t that be lovely?

Side note: It is always amazing to me how quickly people treat changing sexual and relationship norms as if they are the end of the world. In the 90s, divorces became more common. And yet people still fall in love and we actually just worked really hard as a country to make sure that all people have the right to marry. It is far from the death of marriage. I don’t know if hookup culture is meant positively or negatively (probably depends on whether you are actually hooking up or not), but there is one thing I know: it is not the end of love.