Social Media: In Our Feelings
Social Media: In Our Feelings
Techno Research #1. Science about humans, for humans
Thanks for reading Techno Sapiens! If you haven’t already, please consider subscribing. You’ll get posts delivered straight to your inbox and join a growing community of amazing humans.
This week, I’m trying out a new series called Techno Research. I’ll take a recently published scientific study about humans and tech and summarize it in the least boring way possible.1 We’ll start with one of my own studies. As usual, important commentary can be found in the footnotes, including tales of rejected overtures in online stats forums and a historic Twitter spat over cute puppies.
10 min read
When I think back on my time spent on Twitter over the past month, I can identify a full spectrum of emotions. Excitement at getting likes or retweets. Frustration at tweets with which I disagreed. Fear of Missing Out (FOMO) at seeing photos of friends together at a conference I missed. Some combination of happiness and imposter syndrome at seeing others’ job promotions and fancy awards. Unbridled joy at WeRateDogs2.
Our emotional responses to social media play an important role in how it affects our mental health. Perhaps you’re someone who finds themselves frequently feeling happy at your posts’ positive comments, or, alternatively, upset by the pressure to look a certain way in your photos. Maybe you’re someone who just doesn’t really care either way. Or maybe you’re somewhere in between.
Now imagine how this plays out for teens. Let’s picture two teens, in two different homes, in two different towns, who are opening up Instagram, getting ready to post. Each scrolls carefully through recent photos, selecting a picture of a recent night out with friends, applies a filter or two, composes a clever caption, and posts. Then they wait. As they check back in on the photo’s status, each begins to realize that only a few “likes” are trickling in. A few hours later, the like count hasn’t budged.
How do they react?
One teen shrugs her shoulders, closes the app, and heads out to soccer practice, unbothered. The other teen gets a pit in her stomach and sits on her bed, ruminating over what was so wrong with the photo to make her peers despise it or—worse—ignore it completely.
The same social media experience can affect teens very differently.
In a recent study, my colleagues and I set out to learn more about how these different emotional responses to social media might impact teens’ mental health.
What We Did:
We went into high schools in North Carolina and, after getting permission from teens’ parents, had teens complete a series of surveys. In the end, we had almost 700 teens (average age: 14) answer questions about:
How frequently they have negative emotional responses to social media (e.g., feeling bad about getting too few likes, feeling left out or excluded)
How frequently they have positive emotional responses to social media (e.g., feeling happy because of a positive comment, feeling less alone).
Their symptoms of depression.
We did this once in 2018 and again, one year later, in 2019. Then, we analyzed the data.3 We looked at both directions of effects. In other words, we tested whether greater emotional responses to social media were associated with4 more depressive symptoms one year later, and whether higher depressive symptoms were associated with more emotional responses to social media later on. We also looked at how these effects might be different for boys versus girls.
What We Found:
Teens who reported more positive emotional responses to social media were actually more likely to experience depressive symptoms one year later. Why would this be? We suspect that teens who are more emotionally invested in their online experiences – relying on, for example, high numbers of “likes” to feel happy – may be at greater risk over time. This was especially true for teens who reported feeling lonely. Maybe these teens aren’t having positive social connections in their offline lives, and they’re relying—unsuccessfully—on social media for those rewarding social experiences.
Teens who were already experiencing depressive symptoms were more likely to report negative emotional responses to social media. Teens who are feeling depressed might be at higher risk for these difficult emotions when using social media – feeling bad after a negative comment, or feeling pressure to show a perfect version of themselves.
Compared to boys, girls were more likely to report both positive and negative emotions in response to their social media use. It might be that girls are more emotionally sensitive to the experiences they have online.
Problems:
Every academic journal article contains a section called “limitations,” in which the authors must list all the problems with their study. The idea is to increase transparency and temper any overstated conclusions, as well as provide some ideas for what future studies could improve on. It might also serve to keep researchers’ egos in check, and to make them feel sad.
Here are this study’s main limitations:
Self-report measures. The concept of emotional responses to social media is hard to capture in a self-report measure. Is this measure getting at what we hope? Here’s the full measure in case you’re curious.
Time-frame. Is a year the appropriate amount of time to look at social media experiences in relation to depressive symptoms? Or should it be much shorter (like, within the same day)?
As with all studies, these limitations don’t mean we should throw away the study findings. They just mean that we should be careful not to draw definitive conclusions from just a single study.
Why This is New:
In the early days of social media research, most studies would just look at how often teens were using social media. Now, studies have started to get more nuanced, looking at things like what specific behaviors teens are doing online. Even so, very few studies have looked beyond what teens are doing online, and started asking how they feel when they have those experiences. This study is one of the first to do that, and it shows that these emotional responses to social media are pretty important.5
So, what does this mean for me?
Here are some takeaways and conversation starters, for parents, educators, and mental health professionals who work with teens:
We can work with teens to recognize how social media impacts their moods and feelings. This awareness is an important first step for them toward managing the complicated emotional dynamics that play out online. Asking teens questions can help encourage introspection and prompt more mindful use of social media. Questions tend to be more effective than lectures:
“How did it make you feel when Sarah made that comment about you on TikTok?”
“How do you think posting on Instagram affects your mood?”
When it comes to social media use, all teens are not the same. Teens can have very different emotional responses, even to the exact same situation on social media. Pay attention to how your teen seems to feel before, during, and after their social media use. Do they seem stressed before they post? Super excited after a positive comment? Upset when they don’t get a lot of likes? Having a range of emotions around social media use is normal, just as it is for in-person social interactions. But if those emotional reactions are so intense that they’re consistently getting in the way of the rest of teens’ lives—say, preventing them from focusing on homework or discouraging them from spending offline time with friends—it might be time for a change:
“I’ve noticed that sometimes you feel stressed after you post on Instagram. Is that true? Why do you think that is? Is there anything you think we could do to help you feel less stressed?”
“Sometimes I feel a little down when I see everyone else’s perfect posts on social media. Does that happen for you? What helps you feel better in those moments? Sometimes taking a break from scrolling helps me. How about you?”
We should pay special attention to teens who are struggling, whether with depressive symptoms or loneliness. We shouldn’t stop them from using social media altogether (they might be accessing important social support on there), but we should help them identify fun, rewarding, and positive activities offline to add to their social repertoire.
“I’m so sorry that you’ve been struggling. What are some things that you’ve found help you feel better when you’re having a tough time?”
“You’re such a great [artist/singer/soccer player/actor]. What do you think about trying out [activity] and seeing if you like it?”
Thank you to all my study co-authors, including: Drew Rothenberg, who provided critical stats wizardry and (even more critical) emotional support when I was ready to throw my laptop out the nearest window; Alex Bettis, Maya Massing-Schaffer, and Kara Fox, who assisted with idea generation, analyses, and writing; Eva Telzer, Kristen Lindquist, and Mitch Prinstein, who provided the funding and resources to collect this data; and a large team of undergraduates and research assistants who made data collection possible.
What did you think of Techno Research? What do you want to read about in the future? Reply directly to this email!
One of the biggest problems with academic research, to me, is the inaccessibility of research findings. When researchers publish a study, it goes into an academic journal and then hides behind a publisher’s paywall such that even they—the authors themselves—have to pay to read it. I’ll save my thoughts on the horrors of academic publishing for another time. But even if people could get to these articles, they’re inaccessible in another way: they’re long, jargon-y, and, often, unbelievably boring. This is not a criticism of other scientists. All articles, including my own, are like this. They have to be in order to get published. That said, the goal of Techno Research is to help us all stay a bit more informed on the current research, without spending all our money or involuntarily falling asleep.
For the uninitiated, We Rate Dogs (@dog_rates) is a beacon of hope in the sea of angry shouting and cringey humble-bragging that is Twitter. People DM photos of their dogs, which are then posted by the account owner with a brief comment and rating out of 10. All dogs receive a rating that is higher than 10. An important moment in the history of the account—and in my personal opinion, the history of the Internet—was this exchange from 2016. A user named Brant accused We Rate Dogs of having a “flawed” rating system, since every dog receives a score above 10. Chaos ensues, with multiple intentional misspellings of Brant’s name (Brint, Brönt) and impassioned defense of the rating system by devout followers. Shockingly, Bjornt fights back. In the end, Broont is clearly wrong and all dogs are good dogs.
When I run analyses like the ones for this paper, I use a software program called MPlus. An odd quirk of MPlus is that customer support mostly takes place on a series of public online forums. The social dynamics that play out on these forums are truly wild. Just imagine an unwieldy Facebook comment thread (some forums contain over 10,000 posts, dating back 15 years), where every poster is patiently awaiting a response from just two people (in this case, the husband and wife team (!) who co-created this software). Somehow, the software creators appear to have responded to every single post over the past 15 years.
So, I was disheartened to find this short thread, in which a user named Sean asks about potential future directions for the software. The creators show interest, asking whether he has any specific ideas for new features. Sean responds with some ideas, ending with: “Maybe in time, we could collaborate and my conceptual piece could be the pedagogical paper you use to show how MDS [multidimensional scaling] works in Mplus??? Lol.” And there the thread ends. No response. A tentative proposal, summarily denied. Was it the forwardness of the request? The request itself? The use of “lol”? I can’t help but wonder how Sean feels about this rejection. Sean, if you’re out there, I hope you found a pedagogical collaborator who will respect you, who will look at your conceptual piece like it’s the only one in the room, who will laugh out loud with you. Keep doing you.
You might be thinking that an easier way to say “was associated with” would be “predicted,” as in, “more emotional responses to social media predicted more depressive symptoms.” This would definitely be easier. However, researchers are allergic to using the word “predicted” in a non-experimental study like this one, just as they will refuse to use the word “caused.” This is because, even though we know these two variables are correlated with each other over the course of a one-year period, we still don’t know that one caused the other. I am trying to avoid jargon, but I’m also trying to be accurate here, lest the angry scientist mob come after me on Twitter (to which my emotional response would be scared).
Full Citation: Nesi, J., Rothenberg, W.A., Bettis, A.H., Massing-Schaffer, M., Fox, K.A., Telzer, E.H., Lindquist, K.A. & Prinstein, M.J. (2021) Emotional Responses to Social Media Experiences Among Adolescents: Longitudinal Associations with Depressive Symptoms. Journal of Clinical Child & Adolescent Psychology, doi: 10.1080/15374416.2021.1955370