Allgemeine Informationen
Veranstaltungsname | Blockveranstaltung: 08-26-GS-4 Quantitative Text Analysis in the Social Sciences |
Untertitel | |
Veranstaltungsnummer | 08-26-GS-4 |
Semester | SoSe 2025 |
Aktuelle Anzahl der Teilnehmenden | 15 |
erwartete Teilnehmendenanzahl | 20 |
Heimat-Einrichtung | Politik |
Veranstaltungstyp | Blockveranstaltung in der Kategorie Lehre |
Erster Termin | Montag, 14.07.2025 09:00 - 16:00, Ort: UNICOM 7.4680 (SOCIUM - Mary-Somerville-Str. 7) |
Art/Form | |
Teilnehmende | Texts have long been an essential data source for the social sciences. For a long time, however, the analysis of texts (party manifestos, speeches, parliamentary debates, newspaper reports, etc.) was limited to the qualitative analysis of text content. The consequence of this was that scientific studies could almost always only include a small selection of potentially relevant texts, which were then subjected to a more detailed analysis. More recently, however, tools have become available to analyze large volumes of text. On the one hand, statistical methods are used to identify topics addressed in texts or political positions of the authors based on word frequencies and/or word distributions in texts. On the other hand, large language models are used to automate text annotation procedures used in qualitative research in order to analyze developments and structures of debates. The goal of the seminar is to provide a practical introduction to various techniques of quantitative text analysis based on both statistical methods and qualitative annotations. |
Voraussetzungen |
In order to participate in the seminar in a meaningful way, basic knowledge of R is required. If you have not worked with R before, please check out one of the many available online tutorials (e.g. the free Datacamp introductory course in R: https://campus.datacamp.com/courses/free-introduction-to-r) and familiarize yourself with the basic functions of R and the R development environment Rstudio. For working with large language models we will also use Python. |
Englischsprachige Veranstaltung | Ja |