Lehrende: Daniel Gotthardt
Veranstaltungsart:
Vertiefungsseminar
Anzeige im Stundenplan:
V-SEM
Semesterwochenstunden:
2
Credits:
6,0
Unterrichtssprache:
Englisch
Min. | Max. Teilnehmerzahl:
10 | 25
Weitere Informationen:
B.A. Soziologie Hauptfach: Vertiefungsmodul Spezielle Soziologien
B.A. Soziologie Nebenfach: Vertiefungsmodul Spezielle Soziologien
Lehramt: Teilstudiengang Sozialwissenschaft
Kommentare/ Inhalte:
Simulations are imitations of real and theorized processes and have spread in the modern world from purely mathematical computer simulations, over flight simulators for pilots in training to video games like the Sims and SimCity. In the Social Sciences computer simulations can also take various forms for different goals. One of the most influential studies in social segregation was a relatively simple simulation from Schelling (1971) where he showed that a small preference for similar neighbors can lead to strong segregation over time. Simulations can also help us understand statistics. So called Monte Carlo experiments have become the standard of methodological evaluation of assumption violations like outliers as well as properties of newer models under semi-realistic circumstances (Carsey & Harden 2013). But even standard empirical quantitative studies can improve upon interpretatation and visualisation of their results with simulations (King et al. 2000).
In this seminar, we will explore the foundation of simulations for Social Science and learn the tools to implement them ourselves. How can we use artificial simulation of random processes and dynamics in a programming language like R to understand social processes, evaluate statistical models and interpret and visualize the results of our empirical resarch?
Lernziel:
• Conceptual understanding of the different uses of simulations in social science
• basic knowledge of the programming language R
• improved understanding and intuition of statistical concepts like distributions and inference
• ability to program simple simulations and use them for visualisation and interpretation of empirical analysis
Vorgehen:
After a short introduction to the concept of simulations in the seminar, the first part of the seminar will be a tutorial for R and how we can specify random variables and distributions to reconstruct (social) data generation processes. This will also allow us to understand how statistical inference actually works and get a more practical insight into different statistical procedures. With these tools, we will learn how to do our own small Monte Carlo Simulation studies as well as use simulations for plotting results and interpret empirical results beyond regression tables in the second part of the seminar. In part three we will discuss potential but also limitations of some of the most influential generative simulation models for social processes.
In the end, students will run their own simulation to either answer a methodological question, get greater insights into the results of an empirical study or check and expand upon an existing simulation study. Afterwards they will describe and discuss their approach in a short paper.
Literatur:
• King, Gary, Michael Tomz, and Jason Wittenberg. 2000. „Making the Most of Statistical Analyses: Improving Interpretation and Presentation“. American Journal of Political Science 44(2):347–61.
• Carsey, Thomas M., and Jeffrey J. Harden. 2013. Monte Carlo Simulation and Resampling Methods for Social Science. 1st Aufl. Los Angeles: SAGE Publications.
• Schelling, Thomas C. 1971. „Dynamic models of segregation“. The Journal of Mathematical Sociology 1(2):143–86.
Further material will be provided at the start of the semester.
Zusätzliche Hinweise zu Prüfungen:
Type of exam: short paper
Grading scheme: RPO (graded)
Scope:
approx. 10-12 pages (in addition to programming code)
Deadline: March 30th
Output of the assessed examination performances: by e-mail
Further study achievements accompanying the event (ungraded): homework, active participation
|