400020 SE Content Analysis: Manual and Automated Approaches (2018W)

The seminar offers an in-depth introduction to quantitative content analysis application and various statistical and computational approach of text-as-data for social science research. We first briefly review traditional, manual content analysis techniques, focusing its application as an interpretive device for social science-type inferences, and various issues arise from such applications (e.g., codebook development, coder training, dictionary development, software handling, and reliability assessments). We later then shift our focus to recent developments in automated, computer-assisted content analysis of digital material and advanced topics in text-as-data approaches, such as (un)supervised learning and topic models. It discusses the application of these approaches to different types of materials, including traditional newspapers, social media, party manifestos and many more. The goal of the course is to provide students with an overview of the literature while developing an understanding of what is possible. Students will learn about tools for analyzing texts quantitatively and intuition for why the tools are useful. As a final assignment, students will write a research proposal, ideally linked to their PhD project, focusing on methodological aspects and data collection instruments.