The next session at "Compromised Data" is the last I'm going to be able to liveblog, as I'll have to go to the airport this afternoon to head to my next destination on this trip (apologies to the presenters in the final session, whose papers I'll miss). We start with Gavin Adamson, whose interest is in the circulation of mental health news on Twitter. Generally, the journalistic coverage of mental illness in Canada and elsewhere is poor: mental illness is covered mainly in the context of (as a reason for) crime and violence; there are few good news stories being covered.
Do social media amplify or redress that problem? Gavin took a mixed-methods approach which builds on recent research to show that in five years of news coverage, in some 90% of articles nobody with a lived experience of mental illness was quoted; 75% don't even quote healthcare professionals! opting instead for police or people in the justice system. Other studies have shown an overemphasis on risk-based coverage (discussing escapees with mental illness, etc.), and a reliance on the justice system as a frame for mental illness coverage.
This is in the context of medical evidence that shows that one to two thirds of people with schizophrenia show significant improvement or recovery over their lifetime, and that there is no statistical evidence for increased violence in populations with mental illness as compared to populations without it; indeed, sufferers tend to be victims of crime instead.
Crime is also the most commonly tweeted-about theme for most news organisations. Does this mean that Twitter simply amplifies the distribution of stories about the relationship between crime and mental health?
Gavin tracked social media activities around the Globe and Mail and The Star for three months from October 2010 to January 2011 to explore this hypothesis, and found 142 articles on their sites during that period. Topics and sentiment in these articles were coded. He also attempted to find any tweets which referenced these stories.
Stories relating ”mental health" to crime and violence made up one quarter to one third of all articles; for "mental health" between one third and 70%, which supports previous research findings. For the tweets, however, only just over one quarter of all messages referring to the articles related to crime and violence; this is especially the case for articles discussing "mental health" rather than "mental illness".