The truth about tweeting

Having breakfast, need coffee, need shower more…..sorry, you caught me in the middle of a “tweet” there. I’m not alone of course, millions upon millions of people use the social media site Twitter to broadcast in 140 characters or less everything from concepts raised in a lecture to their current plans for ablutions. In 2010 there were 25 billion tweets and 100 million new Twitter accounts registered. Tweeting (also known as microblogging) is a major social phenomenon and it is being used as a tool to analyse other social phenomena. Such was the case in a recent Cornell University study but we do need to be careful how we evaluate the information that Twitter provides.

The researchers looked at Tweets from more than two million people across 84 countries. The text of each message was assessed using software that gave words like “awesome” and “agree” a positive score, and words like “annoy” or “afraid” a negative score. These results were matched against the time of day that the tweet was posted and a clear pattern of mood across countries emerged.

The analysis showed that people were in a happy mood between 6am and 9am but that mood hit a low point between 3pm and 4pm before picking up again in the evening. Overall mood also was lower at the start of the working week and lifted toward the weekend. Interestingly, the daily fluctuation in mood held true for weekend days suggesting that the change is not purely related to feelings about work, although the pattern did shift back by about two hours reflecting later rising times on the weekend.

The data also showed although negative messages were equally common in summer as in winter, there was a swing in message content around the autumn and spring equinoxes. This suggests that perceived changes in amount of daylight can also impact mood.

The researchers concluded that this data shows that our mood does vary with biological rhythms of day and night and also the seasons.

There are a couple of things to keep in mind though before accepting the conclusions suggested by the data. It is highly likely that Tweets tell us more about what the Tweeter thinks people want to hear rather than what they are actually feeling. So using Tweets as a reliable guide to mood is problematic.

Then there is the issue of whether the words chosen to indicate mood, do actually do it accurately. As one psychologist has noted, if you counted the good and bad words that people used during sexual intercourse and used bad words as indicators of bad mood, then you would often conclude that people having sex are having a bad time.