8 March 2017
This blog post comments on one of my classmates, Mielle’s presentation on her review of a study titled GIS Analysis of Depression Among Twitter Users. This study aimed to develop a method to diagnose MDD (major depressive disorder) using posts published on the social media application, Twitter. According to Mielle, the authors of the study cite several other studies that prove that social media can be used as a method to diagnose MDD (major depressive disorder). This is because users of social media sometimes share information about their mental state and well-being on applications such as Twitter. As such, the authors of the study uses a computer program to randomly select tweets from Twitter that contained the word “depress”. Other factors such as the context in which the word was used and the frequency of word use were considered in order to fulfill the criteria listed in the DSM-IV for the diagnosis of MDD.
I liked Mielle’s presentation because she did a great job pointing out the flaws of the study. For example, Mielle pointed out the fact that the use of the word “depress” does not necessarily point to MDD because there are many colloquial uses of the word. She also pointed out that the study diagnosed MDD based on only one indicator trait out of the nine listed in the DSM-IV. One last very important point that Mielle made was the issue of privacy and ethics. While the tweets are public, it is safe to assume that users did not intend for their tweets to be used for their diagnosis of MDD.
Mielle’s review made me think about how social media is reshaping the ways by which our society operates. The use of social media data to diagnose medical issue is very new and unconventional and seems to be a move (or an attempt to move) into a more automated world.