What do we get from Twitter - and what not? A close look at Twitter research in the social sciences

Abstract

The state of Twitter research in the social science domain is investigated based on a set of 25 highly cited papers, identified with the Scopus database out of 370 social science publications on social media research. The analysis shows how social media research in the social sciences has risen since 2007. The selected top cited papers are analyzed concerning their domains, the applied meth-ods and the underlying data in use. It is shown that different methods, both experimental and analytical are applied, and that some papers have started to combine different modes of analysis. The size of the datasets used for studying Twitter varies considerably across studies. Furthermore, central advantages of studying data collected from Twitter are pointed out and open challenges in working with these particular data are listed. Challenges include, for example, data access via the Twitter API or via third party tools, representativeness of datasets and sampling strategies and ethical issues. (author's abstract)