Skip to Content

Billion-Tweet Study Proves We Write Happier Messages When the Weather Is Good

Everyone’s mood improves when the weather is good, right? Actually, the large-scale evidence to prove this has never been gathered—until now.

How does the weather influence our emotions? At first glance, the answer seems trivial—there is plenty of anecdotal evidence to suggest that humans prefer warm sunny days rather than cold wet ones and that we have a sunnier disposition as a result.

But when it comes to hard empirical evidence, there is little in the way to back this up. This evidence is hard to gather because it is difficult to measure mood on the scale necessary to come to statistically significant conclusions.

Today that changes thanks to the work of Patrick Baylis at Stanford University and a few pals who have carried out the largest investigation into the relationship between meteorological conditions and mood. And they say the weather has a significant impact—the first time this has been observed on this scale.

The team’s method is straightforward. They begin by assuming that a reasonable proxy for human mood is the way we express ourselves over social media networks such as Facebook and Twitter. In other words, we use more positive phrases when we are happy and more negative ones when we are down.

In recent years, it has become straightforward to measure the sentiment of social media posts by counting the number of positive and negative words they contain. So Baylis and co simply measured the sentiment in geolocated social media posts and then studied how this varied with the weather in these places.

What gives this study gravitas is the sheer scale of the undertaking. Baylis and co measured the sentiment in 3.5 billion social media posts from tens of millions of individuals on both Facebook and Twitter between 2009 and 2016. These individuals were all from one of 75 metropolitan areas in the U.S.

They measured the positive or negative sentiment expressed in each message and then compared this to the daily meteorological data from each location.

The results make for interesting reading. The team found a significant increase in negative sentiment when the weather is both too cold or too hot or when it is too wet, too humid, and cloudy.

And the size of the effect is significant. To find out just how significant, the team measured the change in sentiment associated with specific events, such as the August 2014 earthquake in Oakland and San Francisco, the San Bernardino terror attack in 2015, and so on.

They then compared this change in sentiment to that associated with freezing weather and found that they were similar in magnitude.

That’s a fascinating result. “We find substantial evidence that less ideal weather conditions relate to worsened sentiment,” say Baylis and co. “To the extent that the sentiment of expressions serves as a valid proxy for underlying emotions, we find some observational evidence that the weather may functionally alter human emotional states.”

That has implications for all kinds of people. The findings could be used to fine-tune marketing campaigns to take into account changes in weather-related mood. It may also allow content providers to better match their output to the emotional state of their audience. And intelligent assistants such as Siri, Alexa, and Google Assistant could use it to adapt their messages to their owners’ needs.

Interesting stuff!

Ref: arxiv.org/abs/1709.00071 : Weather Impacts Expressed Sentiment

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.