sow944 - Quantitative Research Methods

After completing this module, students will have a working knowledge of a range of research designs and statistical methods that are commonly applied in social science research. Students will thus be able to understand the methods used in a wide range of quantitative empirical research articles. This entails not only an understanding of the achievements of these studies but also to comprehend their potential limitations. Overall, thus, students will be able to critically discuss existing research and to generate ideas for improving on existing research designs and analyses. This also entails that students will be able to recognize the limitations of existing research with regard to claims about causal relationships between variables, an important skill not only for assessing the contribution of research for scientific advancement but also for evaluating its usefulness for interventions and policies. In this class, students will also learn to master the statistical software and programming language R (with RStudio) that allows them to conduct a wide range of data analytics tasks. After successfully completing the class, students will know how to import and export data into/from R, conduct common data management and visualization tasks, specify, run and evaluate a range of basic and advanced statistical models, and write up and prepare their results for presentation and publication.

The two seminars combine lecture-style presentations of research designs and statistical methods with applied lab sesssions in which students work on hands-on exercises that help them achieve the necessaryuseful data analytics skills for their studies and career. Given that students enter this MA program with diverse methodological backgrounds, the first seminar is designed to bring participants to a comparable foundational level. For some participants, this will serve as a refresher of previously covered content, while for others it will provide an introduction to essential material. We will start with an introduction to the statistical software and programming language R (with RStudio) that is used throughout this class. We will then have a few (hands-on) sessions on data import/export, data management, and visualization. Then, after briefly reviewing the basic linear regression framework and its extensions, we will build on it by introducing generalized linear regression models that allow to extend the regression approach to dependent variables that are not metrically scaled (e.g. count, logistic, ordinal logistic, and multinomial regression). The second seminar will build on this foundation by discussing the limitations of the standard regression framework with regard to causal analysis and by covering research designs, aspects of model specification, and advanced statistical methods that allow to improve causal claims (e.g. directed acyclical graphs, difference-in-difference, fixed effects, and panel regression). In both seminars, we will also cover how to present and write-up statistical results and how to produce publication-ready output from statistical models.

• Best, Henning, und Christof Wolf, Hrsg (2015) "The SAGE handbook of regression analysis and causal inference" Los Angeles [Calif.]: SAGE Reference.
• Huntington-Klein, Nick (2022) "The effect: an introduction to research design and causality" Boca Raton: CRC Press. (Free HTML-version at: https://theeffectbook.net/)
• Wickham, Hadley, und Garrett Grolemund (2016) R for data science: import, tidy, transform, visualize, and model data. Beijing: O’Reilly. (Free and more recent HTML-version at: http://r4ds.had.co.nz/)

Portfolio consisting of 2 partial performances:
2 short tests