Course Coordinators: Michel Avital and Jason Burton, Department of Digitalization (DIGI)

Faculty

Assistant Professor Jason Burton 

Department of Digitalization 

Aims and Objectives

This course is designed for doctoral students who are interested in applying computational research methods to social science research. The overarching objectives of the course are to (1) familiarize students with key concepts in the field of computational social science, and (2) equip students with the practical know-how needed to apply computational research methods to their own research interests. Special focus is given to the collection and analysis of digital trace data and agent-based modeling and simulation.

Learning Objectives

At the end of the course, students should be able to:

• Critically discuss the emergence of computational social science as a field of research;
• Reflect on the opportunities and challenges of applying computational research methods to their own domain of interest, and social science at large;
• Collect digital trace data via APIs and web scraping;
• Statistically analyze digital trace data and interpret results;
• Build agent-based models of social systems and run simulations;
• Develop a proof-of-concept study that applies computational research methods to their own domain of interest.
  
  

Structure and Format

The course proceeds in two blocks. The first block consists of one half-day session delivered online as a series of pre-recorded videos, which serves to introduce the course, basic concepts, and help students assess their readiness for the course’s programming exercises. The second block consists of three full-day sessions delivered in person at CBS. Each full-day session includes lectures, group discussions, and hands-on exercises with R.

Course Project

The course project is designed to assess each student's understanding of the topics covered in class. The course project requires each student to develop a research study that applies one or more of the computational research methods covered. The course project report should be up to ten pages and must include:

·       A well-defined and motivated research question (ideally oriented in their ongoing doctoral work);

·       Justified selection of one or more computational research methods;

·       An overview of potential results and research impact;

·       A proof-of-concept analysis (e.g., descriptive analysis of a newly collected digital trace data, or a visualization of preliminary agent-based simulations).

On the final day of the course, students will deliver a brief oral presentation of their project and receive feedback from their peers and the instructor.

Evaluation

A Pass/Fail grade will be based on the timely submission of a 10-page course project paper and the quality of the oral presentation in the last session. A retake exam, if necessary, will be administered about a month following the ordinary exam.

Prerequisite Statistical Software

This course will use the R programming language and the RStudio IDE. Before the first class session, students should download R and RStudio here: https://posit.co/download/rstudio-desktop/. Students with no prior experience using R are strongly encouraged to complete an introductory tutorial before starting the course (e.g., sections 1-9 of R for Data Science, the first six sections of these Posit Recipes, and/or SICSS Boot Camp).

Readings

See the course plan below for assigned readings for each session. Students should complete the readings before each session and be prepared to discuss and answer questions pertaining to the reading material.

 
Workload

1 ECTS = 28 hrs

Course Plan

Session 0 (online): Introduction lecture and programming demo

In this pre-recorded, half-day, online session, students will be introduced to the course and computational social science as a field of research. Videos will discuss the breadth of computational research methods that are applied to social science research, issues like data accessibility and ethics, and highlight seminal studies in computational social science. Students will be provided with a programming demonstration and directed to extra materials for getting started with R and RStudio.

Sample readings:

-      Lazer, D., Pentland, A., Adamic, L., Aral, S., Barabási, A. L., Brewer, D., ... & Van Alstyne, M. (2009). Computational social science. Science, 323(5915), 721-723.

-      Lazer, D. M., Pentland, A., Watts, D. J., Aral, S., Athey, S., Contractor, N., ... & Wagner, C. (2020). Computational social science: Obstacles and opportunities. Science, 369(6507), 1060-1062.

Session 1: Digital trace data

This full-day, in-person session will teach students fundamentals of collecting and analyzing digital trace data. We will discuss key characteristics of digital trace data, and the opportunities and challenges they pose to social science research. Students will be guided through programming exercises that involve collecting data via APIs and web scraping, text analysis, and social network analysis.

Sample readings:

-      Tufekci, Z. (2014). Big questions for social media big data: Representativeness, validity and other methodological pitfalls. In Proceedings of the International AAAI Conference on Web and Social Media (Vol. 8, No. 1, pp. 505-514).

-      Brady, W. J., Wills, J. A., Jost, J. T., Tucker, J. A., & Van Bavel, J. J. (2017). Emotion shapes the diffusion of moralized content in social networks. Proceedings of the National Academy of Sciences, 114(28), 7313-7318.

-      Burton, J. W., Cruz, N., & Hahn, U. (2021). Reconsidering evidence of moral contagion in online social networks. Nature Human Behaviour, 5(12), 1629-1635.

Session 2: Agent-based modeling and simulation

This full-day, in-person session will introduce students to agent-based modeling and simulation. We will discuss what an agent-based model is, why some models are useful, and how to critique simulation results. Students will be guided through programming exercises in which they implement a model of opinion dynamics, run simulations, and visualize results.

Sample readings:

-      Bonabeau, E. (2002). Agent-based modeling: Methods and techniques for simulating human systems. Proceedings of the National Academy of Sciences of the United States of America, 99(suppl. 3), 7280–7287. 

-      Smaldino, P. E. (2017). Models are stupid, and we need more of them. In R. R. Vallacher, S. J. Read, & A. Nowak (Eds.), Computational Social Psychology (pp. 311–331). Routledge. 

Session 3: Integration and project workshop

In this final full-day, in-person session, we will reflect on the course material, discuss the future of computational social science, and collaboratively workshop one another’s project plans. During the project workshop, each student will deliver a brief oral presentation and receive feedback from peers and the instructor.