| title | course | output | term | authors | ||
|---|---|---|---|---|---|---|
Introduction to Computational Social Science |
Introduction to Computational Social Science [WP 7.1] |
pdf_document |
Winter 2022 |
|
Instructors
- Maximilian Haag ([email protected])
- Constantin Kaplaner ([email protected])
This is the repository for the course 'Introduction to Computational Social Science' held at LMU Munich, Winter term 2022-23. The course is held in double sessions (2 x 90 minutes) every other week. Session are structured as follows: The first part of a session will include a presentation of the session's topic by the lecturers to set the theoretical foundation. In the second part of the sessions, students will be able to explore the topic themselves using structured coding exercises. The material provided will be written in the R programming language. While students are encouraged to engage with the provided R material, they can also choose to follow the course in a programming language they are familiar with.
The course 'Introduction to Computational Social Science' provides a first glimpse into the field of computational social science (CSS) for political scientists. The advent of computational techniques has opened a window to the use of large-scale data collection and analysis. Examples of CSS in political science include the automated collection and analysis of legal texts, parliamentary speeches, and social data employing web scraping, machine learning, simulation, and network analysis techniques.
We will introduce participants to the necessary context and tools to apply CSS concepts to their own research. Thus, the focus is twofold; First, the course will give an impression of the relevant conceptual and theoretical developments in the field. Second, we will engage in hands-on coding sessions to enable students to apply the presented concepts and techniques.
Participants of the course are welcome to develop and work on their own research projects employing CSS methods in the course. Basic prior knowledge of quantitative data analysis and/or programming is a plus but not required.
Students will have to fulfill the following requirements:
- Read the required readings assigned for each week
- Participate at the sessions of the seminar
- Prepare and present (~10 min) a short research idea/design that outlines the research question and planned approach for the research paper
- Write a research paper (~ 20.000 characters) on your chosen topic (in agreement with the instructor).
Grades for the seminar will consist of the following elements:
- 50 % term paper
- 50 % presentation in class
Term paper deadline: 13.03.2023
At the end of some session's coding exercise you will find additional exercises and homeworks for yourself to try out and engage further with the material. These exercises are completely voluntary. If you decide to do them, you can hand them in to the lecturers to receive comments and feedback.
In accordance with the study regulations, attendance is not mandatory. However, since many of the course's theoretical and practical materials build upon each other, it is highly recommended to attend all of the session and to self-study session materials for sessions you might have missed.
If you cannot attend a session, please inform one of the lecturers beforehand. This helps us plan the session accordingly.
Students are encouraged to visit the instructors' office hours for help with understanding the theoretical and practical course materials.
Offline
If possible, please write an e-mail beforehand if you plan to attend the office hours to allow for better scheduling.
- C. Kaplaner: Room GU105, Tue 15:00 - 16:00 or by arrangement (via e-mail)
- M. Haag: Room GU102, Wed 12:00-13:00 and by arrangement (via e-mail)
Online (via Zoom)
Please schedule a meeting via e-mail.
To follow the course, you will need to have access to an RStudio installation. You can either install RStudio on your own computer or use the university computers. Your personal workspace on university computers is available via Remote Desktop.
Additionally, we offer a RStudio Cloud workspace for working for through the session the course and in your own time. This is an external service offered by RStudio. You will need to register for a free account in order to be able to work on the course materials on your own. The usage of the service within the free tier and for work in the seminar workspace is free to course participants.
Please chose a method of access to an RStudio Cloud installation and familiarize yourself with the access to your RStudio installation.
Please note: We will not be able to provide access to our RStudio Cloud workspace during the seminar paper preparation phase due to time and cost constrains on our end. However, you will have ample time to work through the course materials during the sessions or in your own time on the RStudio Cloud.
- Elicit
- ...
- Getting to know each other
- Course overview
- Terminology
- Tools
- Introduction to RStudio & R
- Lazer et al. (2009) Computational Social Science. Science 323 (5915), 721-723. DOI: 10.1126/science.1167742
- Data & working with data
- Explorative, descriptive, confirmative research
- Types of data & sources
- Working with different data structures
- Working with datasets
- Web scraping / data collection
- Example: scraping text and metadata of e.g. news websites, parliamentary speeches etcetc
- Example: using an API
- Hox, J. J. (2017). Computational Social Science Methodology, Anyone? Methodology, 13(Supplement 1), 3–12. DOI: 10.1027/1614-2241/a000127
- How can we use text as data?
- Which methods can be used to answer political science question?
- What are the limitations of text data?
- Working with text data on the example of legislative speeches
- Dictionary approaches
- Scaling models (Wordfish / Wordscores)
- Topic modelling
Examples:
- How often do MPs in the Bundestag talk about environment?
- How do parties position themselves in speeches?
- What are the main topics of debate?
- Grimmer, J. (2013). Text as data: The promise and pitfalls of automatic content analysis methods for political texts. Political Analysis, 21(3), 267-297. DOI: 10.1093/pan/mps020
- What is the basic idea behind machine learning approaches?
- What does a typical machine learning workflow look like?
- How do we assess prediction error?
- Basic machine learning workflow in R
- Using machine learning for (text) classification
Examples:
- How polarized are MPs in the german Bundestag?
- Peterson, A., & Spirling, A. (2018). Classification Accuracy as a Substantive Quantity of Interest: Measuring Polarization in Westminster Systems. Political Analysis, 26(1), 120-128. DOI: 10.1017/pan.2017.39
Guest Talk: Franziska Quoß (ETH Zürich) - The impact of political business cycles on the environment
- Introduction to geospatial data
- Data acquisition and preparation
- Spatial Analysis
- Examples from applid research (F. Quoß guest lecture)
- Basic concepts and definitions of network analysis
- Applications of network analysis in Political Science
- How to create a graph in R using
igraph - Example: Bundestag MPs on Twitter
- Borgatti, S. P., Mehra, A., Brass, D. J., & Labianca, G. (2009). Network Analysis in the Social Sciences. Science, 323(5916), 892–895. DOI: 10.1126/science.1165821
- Research design basics
- Research in CSS
- CSS Workflows: best practices & possibilities
- Writing term papers with RMarkdown