New Working Paper: Automated video analysis for social science research

By Dominic Nyhuis, Tobias Ringwald, Oliver Rittmann, Thomas Gschwend and Rainer Stiefelhagen

The widespread digitization has profoundly impacted research practices in the social sciences. The ubiquity of digital data has enabled research projects on scales that were almost inconceivable a mere thirty years ago. As human coding is often no longer a viable option to deal with the immense amounts of data, scholars have begun to embrace methodological innovations that aim to automate the transposition of digitized information into data points. Among the most successful and most widely employed techniques is the automated analysis of text, which has become a staple of social science research (Grimmer & Stewart, 2013; Lucas et al., 2015; Wilkerson & Casas, 2017). But even though text mining has yielded crucial insights in a number of disciplinary subfields, scholars have yet to appreciate the full potential of the universal digitization for the social sciences. 

read paper