Virtual Field Research with Social Media:A Pilot Case of Biometeorology

In social sciences, �eld research is an important approach for analysing individual and social behaviour. In doing so, researchers are moving in natural environments of individuals to observe unbiased behavioural patterns. However, systematic �eld work is quite time consuming and expensive. Due to this, sample size often remains small, leading to a problem of external validity. In most of the cases, long term studies are not feasible. Also, there are hypotheses which are not veri�able by observation, for instance opinions or behaviour in rare situations. Our contribution aims to demonstrate the potential of SocialWeb for gaining comparably deep insights into human behaviour with signi�cantly lower overhead. In fact, Social Media increasingly o�ers rich metadata, such as temporal and geospatial attributes. This information allows us to integrate additional environmental knowledge - such as weather data - and to interpret contents of Social Web as a replacement for traditional observations in �eld research