Kommentar |
Many research questions in the social sciences can only be adequately addressed using clustered data. For instance, studying institutional effects on individual behaviour requires data from individuals nested in different institutional contexts (voters nested in different electorates, pupils in different schools). Likewise, panel data (multiple observations nested the same units of analysis) and dyadic data (e.g., siblings nested in families) represent forms of a clustered data structure. The course provides an applied introduction to regression models suited for clustered data using STATA. Emphasis will be given to multilevel models in cross-national research designs and longitudinal data analysis (fixed and random effects panel regression, dynamic models). The course reviews methods for both metric and categorical outcome variables.BGSS doctoral researchers can obtain up to 6 credit points (ECTS) for the course. Masterstudierende können nach Absprache mit dem Dozenten und dem Akzeptieren der Leistungsanforderungen teilnehmen. |