THIS ARTICLE/PRESS RELEASE IS PAID FOR AND PRESENTED BY NTNU Norwegian University of Science and Technology - read more
Our mood fluctuates in rhythm with the day and week
People’s mood on Twitter varies according to more or less fixed patterns. Guess when we’re happiest.
Are you in a better mood as the weekend approaches? Do you get more depressed as Monday morning looms? Is your job not that much fun? You’re not exactly alone.
New analyses of 25 million messages confirm that the mood in Twitter messages follows the fixed rhythms of the day and week pretty closely.
“People’s moods vary at certain times during the week and each day, and this follows more or less fixed patterns. This is also reflected in what people tweet,” says Lucas Bietti, a senior lecturer at NTNU’s Department of Psychology.
Twitter is not only useful for people who want to control a superpower or to yell at faceless bureaucrats. Social media is also a source for researchers studying human behaviour.
Bietti collaborated with Eric Mayor, a researcher at the University of Basel in Switzerland.
Stronger emotions outside of work
The researchers distinguished between messages that people posted about themselves, and messages that were about other people and other topics. They analysed seven million self-referencing messages and 18 million others.
It turns out that both the day and the week affect our shape and mood.
“On weekdays, people are in the best mood between 14:00 and 23:00. They use words like 'sweet', 'love' and 'nice' more often. People’s moods are worst late in the late morning, maybe while they’re working most intensively,” says Bietti.
“We think the positive emotions between 14:00 and 23:00 may be due to people getting reenergized from what they’ve eaten, or that they start to think about the evening approaching, or they’re busy with their leisure activities. But negative emotions also increase at this time, which suggests a complex picture that should be examined more thoroughly,” says Mayor.
“We also see an increase in positive emotions when people get up in the morning before they go to work. But this positive trend ends around 10:00,” says Mayor.
The week affects us, too
Weeks follow a different pattern.
Bietti summarizes it this way: “We’re the most nice towards ourselves starting Thursday afternoon, and think the worst about ourselves from Sunday afternoon to Thursday afternoon,” he says.
The rhythms in people’s tweets don’t follow their work days and days off exactly. People’s moods worsen again as the weekend draws to a close, when they may begin to dread going back to work.
“Positive words are used less in tweets starting on Sunday afternoon. This lower mood lasts until Thursday afternoon,” says Bietti.
In other words, the mood picks up again when the weekend is around the corner. (Hooray!)
But life isn’t just about work and weekends. Biology also helps to explain our variable moods.
The short 24-hour rhythms are called circadian rhythms. These rhythms are probably familiar to most of us, because we’re just not quite the same in the morning as in the evening, are we? But the changes may be bigger than you thought.
Our body’s cells work differently throughout the day, and our behaviour can vary. Our internal clock responds to variations in body temperature. This temperature is highest in the middle of the day and lowest at night.
We might have developed these rhythms in response to variations in light and dark, but we don’t know for sure. In any case, the rhythms are important for our sleep, and likewise for our mood and feeling upbeat. When we throw these rhythms out of whack, like on long flights, we may experience jet lag.
So circadian rhythms are normal for most of us. But the fact that weekly variations, the circaseptan rhythms, are also so common and so clear, is perhaps a little more surprising. Or maybe not. Most people apparently like their time off best.
Also analysed emojis
Previous research has already suggested these kinds of connections between time and mood, but the basis for the conclusions wasn’t always that solid.
This earlier research has concentrated on text. But since both text and emojis can express emotions, researchers took both into account. The analyses also take the use of slang into account. Different methods were used to interpret people’s moods.
“This is the first survey that uses several of these different methods to analyse tweets,” says Bietti.
Mayor and Bietti found roughly the same trend, although the results vary somewhat depending on the method. This helps to make the results more reliable.
The results of the study have been published in an article in the academic journal Royal Society Open Science.
Eric Mayor & Lucas M. Bietti: Twitter, time and emotions. Royal Society Open Science, 2021.