I am wondering whether there are points of fruitful interaction to be had between a proposed approach to research within the Ontario Health Study and the work on cultivating citizen volunteers among users Twitter in responding to disasters in Chile.
The latter work I learned about in a meeting yesterday of students and researchers working under the supervision of Gonzalo Bacigalupe in Santiago and in Boston. As I understand the work, millions of tweets have been filtered down to a small set of twitter users in South America concerned about responses to earthquakes, tsunamis, and other disasters. The team is aims to characterize a sample of users as a first step towards informing the ways that national response organizations employ social media. (The results are not yet in, so I cannot provide examples of what these characteristics might be.)
As I listened, I thought first about how the millions of tweets needed to be filtered down in part because social media is a Wild West, in which all kinds of comments are made with varying degrees of civility, cogency, basis in evidence, and interest in being constructive. I wondered whether national response organizations (with which Gonzalo works) would ever support some kind of team whose goal was to bend the twitter discourse in more thoughtful and constructive directions. Such a team might affirm helpful tweets and gently, not agonistically, provide thoughtful, fact-filled replies to selected other tweets, while ignoring ones that showed were at the low end of the civility, cogency, etc. spectrums.
This thinking, in turn, reminded me of an attempt by a certain group epidemiologists and other population health researchers to get involved with the Ontario health study. This study has recruited a quarter of million people online to provide information about aspects of their health from time to time, not more often than once every six months. The information requested would depend on the particular research team authorized to collect data for their questions. The one group that I had a connection with wanted to examine the effects of well–designed interventions, which might be as simple as having selected office workers use the stairs rather than elevators, comparing some health outcomes to those for a control group not given instructions. The group’s thinking was not to wait for years to learn from non-interventional studies what characteristics were correlated with healthy outcomes.
It was this sense of bringing interventions into the research itself, as well as the subjects being online, that made me think of the possibility of fruitful interaction between the two endeavors.