Mind & Body

Tweeting About Negative Feelings Might Make You Feel Better

If you ever wonder whether people are sad, check Twitter. A wildly popular account, so sad today, tweets sentiments like "i don't like anything related to reality." Depression memes (and depression memes about memes) abound. Even the jokes are about crying and tragic sandwiches. Are we all just getting sadder and sadder online — or does the tweeting somehow help?

Related Video: How to Comfort a Friend in 3 Easy Steps

Inside #FeelingsTwitter

A recent study explored this by combing through almost 75,000 Twitter accounts. Researchers specifically trawled the social media platform for accounts that tweeted straightforwardly about feelings — "I feel really bad," for instance, or "I'm feeling awesome." Tweets like this are instances of "affect labeling," and previous psychological research suggests that it has a therapeutic effect, especially if the affect labeled is negative; just putting dark feelings into words seems to lessen them. But would that hold true online, or was that just a product of face-to-face interaction?

To find out, they processed the accounts that had posted affect-labeled tweets through a sentiment analysis tool called the Valence Aware Dictionary and Sentiment Reasoner (VADER), which rated the positivity or negativity of the tweets posted before and after each #feels tweet.

It isn't a perfect tool, of course. For one, it collapses the huge spectrum of human emotions — sadness, jealousy, frustration, hope, happiness, gratitude — onto a simple positive-negative spectrum. For another, it's fluent for an algorithm, but it's not for a person. It seems like VADER could be tricked by sarcasm, or the way "I'm crying" means anything and everything except "I'm literally crying" online.

Still, it captured some intriguing macro trends in how we emote on the web.

Express Yourself?

Researchers found that "affect labeling" worked differently for positive and negative emotions. First of all, negative affect labeling usually brewed for a long time; a tweet like "I feel really bad" usually came after a long string of tweets with negative VADER scores. Positive affect labeling, on the other hand, came on suddenly, with much less windup. (It's no wonder we talk about "bursts of joy"!)

The tweets themselves had different effects, too. Tweeting "I feel really bad" typically made people feel better — or at least, the tweets that came after a straightforward expression of unhappiness like that had higher VADER scores. Tweeting "I feel amazing," on the other hand, had the opposite effect. It seemed to dissipate the user's joy since the tweets that followed tended not to be as glowing.

In a way, though, these are both facets of the same phenomenon: Expressing or "labeling" an emotion tends to diffuse it and bring you, or at least your tweets, back to a neutral emotional state.

It's unclear, though, if tweets are a good way to measure someone's genuine emotional state. People feign happiness on social media all the time, and even when we express sadness, we do it in careful and curated ways. This study may really be evidence that people only tweet about their darkest emotions sporadically, for fear of seeming like a drag.

It's also unclear if affect labeling actually causes feelings to dissipate. It's more plausible with negative emotions — if people respond with a flood of support, that could understandably cause an uptick in your mood. But why would tweeting about joy deplete it? Instead, it might just lessen over time because that's what all feelings do.

Get stories like this one in your inbox or your headphones: Sign up for our daily email and subscribe to the Curiosity Daily podcast.

For more about how Twitter toys with your emotions, check out "The Happiness Effect: How Social Media is Driving a Generation to Appear Perfect at Any Cost" by Donna Freitas. We handpick reading recommendations we think you may like. If you choose to make a purchase, Curiosity will get a share of the sale.

Written by Mae Rice February 1, 2019

Curiosity uses cookies to improve site performance, for analytics and for advertising. By continuing to use our site, you accept our use of cookies, our Privacy Policy and Terms of Use.