Methods of collecting facebook data and their effects on later analysis (original) (raw)

In this paper, we discuss how methods of archiving data on social network sites such as facebook necessarily reflect deletions (or selections) brought on by technical arrangements and theoretical interests. While collecting facebook data can be accomplished in a number of ways, eg. user diaries, observational notes etc., we discuss in particular the process of archiving the data from the web for analysis and documentation. Web data created in and by the process of archiving can be considered 'reborn' digital material, ie. an unique version of the material which we can never expect to find in the form it actually took on the web (Brügger, 2010a, 2011). Web data created in and by the process of archiving can also be considered both "found", ie. created through unobstrutive measures, and "made" insofar as they require a series of methodological choices and considerations to be meaningfully implemented in a research process (Jensen, 2012; Lomborg, 2012, cf. Brügger 2005). Furthering this body of research, we show in what ways four specific methods of web archiving each in its own way provide both limited and privileged access to the original data and, accordingly, frame what research questions can be answered. The four methods are 1) still image, 2) screen filming, 3) harvesting via API, and 4) web crawling. In addition, it is discussed how broad webarchiving in national web archives affect the possibilities for writing historical studies of, for instance, Facebook in the future, just as the interplay between internet and television archives is debated. In order to illustrate and feed the discussion, we build on an ongoing case study in which we have used all three methods. The case study is about the interplay between real-time internet and live television in the Danish talent programme Voice (Laursen, Sandvik, & Brügger, 2012).