In a novel new study from Cornell University which is published in the journal of Science this month, researchers used text analysis to track people’s daily mood fluctuations and patterns. While there have been many studies using online data mining looking at blogs, chat rooms, social media sites and all manner of internet activity, most have focused on more of a long term picture or stayed within a local time zone or holiday period

The daily study, the first of its kind that crosses cultural and geographical boundaries appears to show that people’s daily mood patterns maintain a close correlation to biological patterns.

Peter Sheridan Dodds, a researcher at the University of Vermont who was not involved in the new research was excited about the new information :

“There’s just a torrent of new digital data coming into the field, and it’s transforming the social sciences, creating new lenses to look at all sorts of behaviors …… [its] very exciting, because it complements previous findings and expands on what is known about how mood fluctuates.”

The research gathered information from more than two million people in 84 countries and concluded that in general people’s emotional tone in their Twitter messaging moved according to the time of day, starting off positive in the morning, declining to a low in the late afternoon and lifting again as most people’s bedtime approached. They also looked at longer term seasonal patterns and concluded that our moods are influenced by the biological rhythms of night and day, winter and summer.

The text of each message was assessed with a standard computer software that gives words like ‘awesome’ and ‘agree’ a positive score and others, such as ‘annoy’ or ‘afraid’ with a negative score. There was a clear pattern showing people’s mood elevated from 6 to 9am and hitting a low point from 3 to 4pm and improving after dinner time.

Unsurprisingly people’s moods were lower at the start of the work week and raised at towards the end, and data assessment was adjusted to account for countries where Saturday and Sunday are not considered days off work. The weekend data also showed a shift in the good to bad mood peak and trough of about two hours, as people tend to sleep later and stay up later on the weekend.

Researchers not involved in the project added a note of caution Dan Gilbert, a Harvard psychologist thought that :

“Tweets may tell us more about what the tweeter thinks the follower wants to hear than about what the tweeter is actually feeling,” said in an e-mail. “In short, tweets are not a simple reflection of a person’s current affective state and should not be taken at face value.”

He was however interested by the weekend shift, because it would be easy to assume people’s moods become slightly depressed towards the end of the work day, but the weekend followed the same pattern as the weekday.

“This is a significant finding because one explanation out there for the pattern was just that people hate going to work ….. 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.”

Another interesting indicator drawn from the data showed negative messages just as common in the winter as the summer, but there was an apparent swing as the lengh of the day increased and decreased around the spring and autumn equinox in March and September. This led researchers to conclude that the apprehension of the changing season and amount of day light, had more of an effect than the actual period of more or less daytime.

One thing on which all the researchers agree on is the need to improve and fine tune the software which is still only in rudimentary stages. Dr. Dodd illustrates his point :

“I suspect that if you counted the good and bad words people said during intercourse, you’d mistakenly conclude that they were having an awful time,” Dr. Gilbert said.

Rupert Shepherd