It has long been questioned whether data found on Facebook is useful in real life settings, perhaps most notably when it comes to product marketing and individual Klout scores. Although research has overwhelmingly indicated that Facebook data is not useful within this avenue, there has been limited research on whether Facebook data can be related to health status. Now, a study suggests that ‘liking’ activity-related or sedentary-related interests on Facebook are associated with obesity prevalence.
The study, which was published in PlosOne, was a cross-sectional study that compared Facebook user data with data from well-known health surveys. Specifically, the researchers focused on the types of ‘likes’ that individuals had on their Facebook to see if that was positively or negatively correlated with obesity in that area. The key findings indicated that individuals, who ‘liked’ TV shows and had other sedentary related areas, also resided in areas where obesity was high. Similarly, individuals who ‘liked’ outdoor activities were found to reside in areas with lower levels of obesity. This led the researchers to conclude that social media could be a useful tool for mapping health statuses and for creating interventions.
Although it was not specified within the publication, one can infer that the sample within this study was substantial. However, the time when the Facebook data was gathered was not the same time that the survey data had been gathered (they were two years apart). As such, it is not clear whether any substantial changes could have occurred to affect the results of the study. However, given that the study was concerned with several areas in the United States, it seems unlikely that the trends would have changed so rapidly within two years. Moreover, the researchers can be applauded for trying to amend this issue by comparing their data to previously known historical trends.
This study was a good initial study. However, there are several areas that future studies could improve if social media is to be taken seriously as a research tool. Perhaps the most obvious one comes from the fact that the comparisons were made on a geographical rather than an individual level. As such, ethnicity was ignored despite the fact that obesity, as defined by BMI, varies between races. Therefore it is plausible that some neighbourhoods that had more Asians were misrepresented. Moreover, it would be useful to compare data from social media with data from the built environment as it could be argued that the environment in some areas restricts the opportunity for activity-related interests.
Although it is clear that these types of studies are in their infancy, it can also be argued that the use of social media in research may not be an idea to abandon as of yet. If used correctly, then social media can circumvent previous problems within research such as cost, lack of substantial sample sizes and slow pace of data gathering. Perhaps, rather than throwing out the baby with the bathwater, studies should focus on figuring out where information from social media fits in with research and how it could be utilised to create interventions.