New data sources
Measuring Cultural Engagement amid Confounding Variables
IssueSpecial Issue: Arts participation and engagement
Arguing that current practices are unsustainable, some speakers outlined how “big data” or “organic data” could be combined with sample surveys in order to better track and study human behaviour. Organic data can be thought of as data that are captured as part of other processes (e.g., Google searches, data scraped from websites, tweets, retail scanners, credit card transactions, etc.).
It was argued that this type of data, which can be nearly real time, is better at capturing people’s behaviour than their values and opinions. In addition, organic data only cover one fleeting point in time, and researchers would not know many characteristics of the people for whom organic data activities have been collected.
Challenges related to organic data include privacy concerns, the high computing power required to “scrape” records, and the current lack of data mining techniques that could combine imperfectly coordinated datasets. There are also questions regarding the representativeness of alternative data sources (which cover only some portion of the overall population). The entire public is not active on even the most popular social media sites (Facebook and twitter).