#currentmood is a popular hashtag among social media users. It typically accompanies an amusing image depicting the user’s mood. But behind the tongue-in-cheek memes and grumpy cat pictures, there’s a huge amount of data that can provide insight into the human condition.
Previous studies have shown that those with mental health conditions use language differently than the general population. But that research has been limited to Twitter data, which is problematic for two reasons: first, tweets are short-form posts that may restrict users from discussing topics at length; and second, Twitter is not anonymous. This lack of anonymity is a drawback because previous research has found that anonymity is more likely to facilitate open and uninhibited discussions of mental health.
Dr. Frank Rudzicz (TRI Scientist) addressed these shortcomings by using data from Reddit, which is a long-form and anonymous social media forum. He focused specifically on how mood is affected by the season or length of day—with the aim of shedding light on those suffering Seasonal Affective Disorder, a subtype of major depression.
In the study, Dr. Rudzicz and his research student, Kawin Ethayarajh, examined comments from more than 100,000 Reddit users over the course of one year. They used Linguistic Inquiry and Word Count, a language analysis tool, to assign words into specific psychological categories such as anxiety, anger or sadness to determine mood.
After relating these data to the length of day on which the comments were posted (calculated based on the city of the user), they found that a subpopulation of users’ moods changed depending on the length of day.
“We are the first to demonstrate that Reddit comments can be used to study mood and mental health,” explains Dr. Rudzicz. “Our results provide a new direction for future research on the ways in which social media can be leveraged to benefit health care.”
This work was supported by the Toronto Rehab Foundation.
Ethayarajh K, Rudzicz F. The effect of photoperiod on the mood of Reddit users. Cyberpsychol Behav Soc Netw. 2017 Apr. doi: 10.1089/cyber.2016.0487.
Dr. Frank Rudzicz is a Scientist at UHN’s Toronto Rehabilitation Institute, where he studies language, artificial intelligence and robotics within the context of health care.