Studying Society Through a Keyhole

The problem of methodological individualism when studying societal phenomena

My colleague recently wrote about social media and well-being in our social science newsletter, which got me thinking more deeply about the conclusions researchers are drawing and how they study this topic.

As is often the case, experts disagree about the effects of social media. Some, like Jonathan Haidt, believe social media is harming mental health, especially among teens growing up immersed in the technology. Others object to that claim, saying Haidt’s conclusions are “not supported by science.”¹

I don’t know which view is correct. I think it’s likely that social media (and other digital technologies) impact different people in different ways, and that the impact depends on how people use social media and what they would be doing otherwise. But something deeper about this research has been bothering me. Social media use is a societal phenomenon, but it’s primarily being studied as an individual one.

Typical studies about the effects of social media use ask people to forego using social media for some period of time, or they longitudinally track changes in an individual’s social media use and well-being over time. I don’t disagree with those studies—they are a great way to assess the impact of an individual’s use of social media on well-being. However, The effect of an individual’s use of social media is subtly distinct from the effect of social media on an individual. I believe those studies are missing the point if we want to know the effects of social media on individual well-being.

A social media detox prevents you from directly observing things on social media, but it doesn’t free you from social media’s influence. If you stop using social media and go to the park, it does not bring other people to the park. It doesn’t cause people to make eye contact with you or greet you with a smile. People will still walk down the street staring at their phones. Friends will still be on their phones at dinner, or not answer your call because they’re mid-YouTube binge. The fact that you have given up social media does not make the world around you more present or connected. You still experience the “secondhand smoke” of social media.

And that, I think is where the main effect of social media lies. It is not an individual’s use of social media, it is that everyone is using social media. But that’s not what I’ve seen being studied most frequently in research about social media.²

I recognize that studying the effect of our collective social media use is challenging, if not impossible with many methods of social inquiry. And I believe studying individual-level use of social media produces valuable knowledge. But we shouldn’t mistake studies about individual-level social media use as the whole picture.

The research about social media use feels like an example of methodological individualism—treating individuals as the fundamental unit of analysis—causing a theoretical individualism. Because we mainly study individuals, we mainly see individual-level causes for behaviors, attitudes, happiness, and other mental states.³ But are those the the most powerful factors shaping our experience?

I suspect we will struggle to understand the effect of most social phenomena if we continue to study them primarily at the individual-level.

1. I get annoyed with atheoretical arguments about things being supported or not supported by science, especially with new phenomenon where all research is relatively recent. Scientific discovery is a continuous process, and there are numerous examples of scientific understanding evolving over time. We used to think that moderate alcohol consumption was good for your health. We compared non-drinkers with light-drinkers and saw that the light-drinkers were healthier! But it turns out that many people who don’t drink had to stop drinking for health reasons. Their health outcomes weren’t worse because they abstained from alcohol, they abstained from alcohol because of their health problems. See Science Vs for an informative and entertaining review of this literature.

2. Me not seeing it does not mean it isn’t being studied. Please let me know if you have seen good research on that subject!

3. Many researchers go beyond these individual-level causes (social psychology is an entire field of study), such as the excellent work of Betsy Paluck.