Is Twitter Able to Determine Your Mood?

| No Comments

I was casually reading the newspaper a few days ago, and I came across a very interesting article in the Pioneer Press (posted above). The article is based on a study that tries to connect a person's mood and how it changes over the course of a day, week, or season. How they planned to do this you ask? The answer was Twitter. The study conducted by Scott Golder and Michael Macy, sociologists at Cornell University, seems to repeat what we may already think is common sense. According to the article, some previous studies have tried to measure the average person's mood on social media sites and "elsewhere on the internet, but (these studies) looked at collective moods over time, in different time zones or during holidays." However, this study was different because it went across cultures, using over 2 million tweets from people in 84 different countries. The results of the study showed that positive posts crested during the times of 6-9 a.m. and gradually fell throughout the day until 3-4 p.m. After that, it slowly went up again, with a sharper increase after dinner. This follows the previous studies in that people's moods were lowest on Monday and Tuesday, and rose as the week went on with peaks on Saturday and Sunday. What makes these results interesting is that the same trends were found on the weekends, only shifted a few hours later. This is making the researchers to believe that our mood could be biologically influenced due to the time of day.
However, there are plenty of possible confounds in this study. The first one is that the article never said if the tweets were randomly selected. That could affect the validity of these results in a negative way if random selection was not used. Also, this study doesn't have the characteristics of an experiment. Those characteristics are random assignment of participants to conditions and manipulating an independent variable. That is important because we can only draw correlations from this evidence and can't rush to assuming causation without further tests. Another possible shortcoming of this study is how they measured positive and negative moods. Sarcasm can't be easily found in text, so it is hard to truly determine if a person feels good or not through a tweet. Something that the article mentioned, which I completely agree with, is that the tweeter's motives for posting the tweet could be clouded. They could be posting the tweet expressing their true emotions at the given time, or they could post a tweet to tell their followers what they want to hear. This type of tweet could mislead the scientists from knowing what mood the tweeter was in. Overall, some confounds are present here, and the study seems to have been conducted in a very meticulous manner, but we have to make sure that we don't fall victim to mistaking correlation with causation.

Leave a comment

About this Entry

This page contains a single entry by sabe0047 published on October 3, 2011 12:02 AM.

Phantom limb-writing 1 was the previous entry in this blog.

"Unnecessary Illusions and the Truth about Procrastination" is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.


Powered by Movable Type 4.31-en