We attribute our changing moods to many things — bad news or good news, ugly weather or beautiful sunshine, hormones and pheromones, and so on. While all these factors may, in fact, have something to do with the mystery of mood changes, a new study tells us that moods also may fluctuate according to some universal schedule that governs people all over the globe, regardless of location or events.1
The study, conducted by sociologists Michael Macy and Scott Golder of Cornell University, analyzed over half a billion messages posted on Twitter by 2.4 million English-speaking individuals in 84 countries. There is much to be astonished at in that fact alone — the sheer volume of data, for one thing, and also the fact that all that data (your Tweets and probably Facebook posts) can be accessed by researchers, and even the fact that social media sites now are being looked at as repositories of good information about global trends.
If you’ve been on a safari to the Moon for the last few years and don’t know what Twitter is, it’s an online community where people can post up to 140 characters at a time (mostly from their cell phones) letting the world know what they’re up to. For perspective, the previous sentence contains 246 characters, meaning that “Tweets” (the name assigned to Twitter posts) are pretty puny in length. Although the overwhelming number of Tweets are insane, Tweeting played a major role in coordinating the several revolutions of the Arab Spring;2 and again, as pointed out above, this mountain of banality can now be mined for gems of scientific research.
To do this, the scientists fed the messages into a computer program and searched for about 1000 words that had either negative or positive connotations. They then evaluated the messages by hour of the day, looking for those value-laden keywords. According to the system, words like “awesome” and “fantastic” denote a happy mood, while words like “annoyed” and “afraid” indicate a down mood.3 Upon analysis, the results showed a clear pattern: people everywhere in the world used their happiest language and ostensibly were in their brightest mood first thing in the morning, typically between 6:00 and 9:00 am. Then, there was a slow decline, with people getting in progressively worse moods until 3:00 or 4:00 in the afternoon. Moods started climbing up the happiness scale again after that, peaking a bit after dinner. The analysis found that people were in their worst moods on Monday, with moods getting sunnier by the day until peaking on the weekend.
Logic would seem to point to the horrors of the workplace as being the cause of the afternoon slump — a few hours at work and people lose all hope and joy. The same thing would apply for the Monday blues, when people have to go to work after enjoying time off. Unfortunately, under scrutiny, that logic doesn’t seem to hold up. The fact is that the pattern holds up on weekends too, when work is not a factor, with happiness peaking at nine in the morning and nine at night, and late afternoons bringing the doldrums. The weekend highs run a few hours later than the weekday highs, probably because people sleep later Saturday and Sunday. But even on weekends, people suffer the same afternoon slump.
As Dr. Golder remarks, “This is a significant finding because one explanation out there for the pattern was just that people hate going to work [as we already mentioned]. But if that were the case, the pattern should be different on the weekends, and it’s not. That suggests that something more fundamental is driving this — that it’s due to biological or circadian factors.”
Suicide statistics bear out this pattern, by the way. A University of Minnesota study found the fewest suicides taking place mornings from 4:00 am to 8:00 am, and the most on Mondays, with lunar phase playing no role.4 A 2001 Italian study found that young people tend to commit suicide between 3:00 and 6:00 pm; but older adults tend to jump the gun and choose late morning to early afternoon. So much for impetuous youth!
The researchers do note that although the bell-curve remains consistent no matter the day of the week, with mornings always being the sunniest time, the happiness baseline remains higher on weekends, meaning that people start out happier on a Saturday than they do during the week and don’t usually get as low. The same thing holds true for holidays and days off, even when those days don’t fall on a weekend. Dr. Macy suggests that the fact that happiness remains elevated on days off may have to do with the fact that people are better rested, since they can sleep as late as they want.5
The question raised by this study is whether, in fact, we humans really do come equipped with standard-issue, built-in time clocks that regulate mood. This study makes it look like indeed we do — whether we fish for a living in the Arctic or trade futures on Wall Street, whether we’re of Sudanese or Taiwanese or German descent. It would be interesting to find out if animals, too, get grumpy in the afternoons and feel best in the morning; in other words, if this is a biological blueprint that all mammals share. If only we could find a large population of chimpanzees into Tweeting!
But assuming that there is some validity to this study, that circadian or biological factors rule our moods, it’s certainly also true that other factors play a big role. There are the individual events in our lives, of course, which can make us feel giddy or miserable no matter the time of day, although perhaps unhappy events make us feel worse in the afternoon than they do in the morning. And then there’s the influence of global events. Another study that analyzed Tweets, out of the University of Vermont, found that world events influence happiness trends worldwide, even when the people posting had no direct involvement in those events. The earthquake in Japan triggered the second greatest mood slump on record. But interestingly, the lowest point in global mood on record came with the death of Osama bin Laden. Go figure. And perhaps, not surprisingly, given the state of the world economy, overall, happiness in the world has been declining in the past year.
Then there’s the obvious fact that diet influences mood. It’s interesting that people are happiest before eating in the morning and after digesting in the evening. Could it be that lousy food choices worldwide put people in a slump, so that by midday they’re depleted and nasty from the fallout? Is it connected to the insulin spikes caused by high glycemic meals and snacks that leads to a “high,” followed by the subsequent crash as the sugar is metabolized and insulin levels plummet? Could it be that the tendency to eat junk food in the mid to late afternoon is a misguided attempt at self-medication, wanting sugar or caffeine or alcohol to stimulate the drooping mood? The possibility gives an interesting spin to the pre-dinner institutions of tea-time in Great Britain and Happy Hour in the US. Certainly bolstering the system though healthy diet, supplementation, and natural energy boosters would be a wiser fix than trying to patch the slump with a doughnut and coffee — not to mention a shot of alcohol depressant at the bar.
1 Carey, Benedict. “Twitter Study Tracks Where We Are” 29 September 2011. The New York Times. 6 October 2011. < http://www.nytimes.com/2011/09/30/science/30twitter.html?_r=1>
2 Philip N. Howard, Aiden Duffy, Deen Freelon, Muzammil Hussain, Will Mari, Marwa Mazaid. “Opening Closed Regimes: What Was the Role of Social Media During the Arab Spring?” PITPI Working Paper 2011.1. <http://dl.dropbox.com/u/12947477/reports/pITPI_datamemo_2011.pdf>
3 Mann, Denise. “Twitter Tracks Mood Swings.” 29 September 2011. WebMD. 6 October 2011. < http://www.webmd.com/balance/news/20110929/twitter-tracks-mood-swings>
4 Maldonado, G and Kraus, JF. “Variation in suicide occurrence by time of day, day of the week, month, and lunar phase.” 1991 Summer. PubMed.gov. 6 October 2011. <http://www.ncbi.nlm.nih.gov/pubmed/1887454>
5 Peck, Morgen E. “Scientists Chart Worldwide Moods Using Twitter.” 29 September 2011. Innovations Daily. 6 October 2011. < http://www.innovationnewsdaily.com/scientists-chart-worldwide-moods-using-twitter–2289/>
How did the University gain
How did the University gain access to the Twitter postings? I’ve never seen these — are they all in the public domain?
Yes, Twitter is different
Yes, Twitter is different than Facebook.
“Then, there was a slow
“Then, there was a slow decline, with people getting in progressively worse moods until 3:00 or 4:00 in the afternoon. ”
It’s ancient lore. There’s a reason for the tradition of the Siesta.