Reusing Qualitative Data for Automated Text Analysis. A Dictionary for Federalism Research

The Federal Entities Dictionary (FeNDi) is a lexical resource for automated text analysis. It captures linguistic references to Germany’s federal arena, including its institutions, actors, and coordination bodies. FeNDi builds on an extensive qualitative content analysis of parliamentary debates and was validated using “StateParl”, a large-scale corpus of debates in the German state parliaments since 2000. By providing a reusable and transparent tool for identifying references to federal entities, FeNDi contributes to the research on parliamentary discourse, intergovernmental coordination, and the dynamics of territorial and party competition in federal and multi-level systems. FEnDi was created in the research project “Reusing qualitative data for automated text analysis. A dictionary for federalism research” that was conducted at Freie Universität Berlin in cooperation with GESIS and was funded by the VolkswagenStiftung. FEnDi captures linguistic references to Germany’s federal arena and its entities, including institutions, actors, and coordination bodies. Its purpose is to provide a reliable tool for identifying how federalism is invoked in political discourse, enabling consistent analyses across research projects. FEnDi comprises 228 entries. For each entry, it contains the substantive term, its representation as a glob-style pattern for the use in the R package “quanteda”, and a corresponding regular expression (Regex). FEnDi is provided in a CSV format to ensure accessibility and extensibility. This choice allows scholars to easily inspect, adapt, and extend the resource, while keeping it compatible with a wide range of text analysis tools and programming languages. In addition, a Workbench Note documents the conceptual and methodological development of FEnDi.