Social Dynamics: Understanding the spreading of products and behaviors in society

Objective:

Understanding and modeling social dynamics is key for policies aimed at large-scale behavioral change; for managerial decisions about targeting, influencer marketing, digital platform design; to anticipate unintended consequences of changes to the social systems we live in. By the end of this seminar, the students will gain both an overview of four major themes in modern social dynamics research, and the ability to implement and analyze the outcomes of computational models of social dynamics. There will be two introductory lectures on social networks, models and simulations of social dynamics, and their coding in Python. Subsequently, each student will independently study one pre-assigned research paper, attempt to replicate somee of its findings, and deliver a presentation on the paper and results.

Attending the presentations will allow the students to become familiar with four main directions in modern social dynamics research:

• Success inequalities and unpredictability. Extensive research has shown that popularity and success typically follow heavy-tailed distributions, meaning that in many social systems, few products become hits whereas many others remain unknown. The students will learn basic mechanisms underlying the observed success inequalities, and their implications for the (un)predictability of new product success.

• Simple vs. complex social contagion. Many research papers and popular books assume that products and behaviors are "infectious" and spread throughout social networks like biological viruses. The students will learn when and why this epidemic metaphor can be applied to model certain types of information spreading ("simple contagion" processes), and why it fails to describe the spreading of political views, sustainable and healthy behaviors, and costly products. For these processes, the students will be introduced to the "complex contagion" theory, and to the role that different network structures and interventions play for the success of simple and complex contagion processes.

• Seeding policies for social change. Seeding policies guide decisions about which individuals to target first to initiate the penetration process for a new product or behavior, with important applications for influencer marketing and interventions to promote sustainable behavior. The students will become familiar with network-based seeding policies and their relative effectiveness for simple vs. complex contagion processes.

• Algorithmic feedback and social dynamics. Human behavior is increasingly coupled with algorithmic behavior, with potentially unpredictable and unintended long-term consequences for digital platforms and global collective behavior. The students will learn simple models to investigate the impact of ranking or recommendation algorithms on social dynamics.

Learning outcome:

The main objective is to offer the students an overview of the main theories and methods to understand the spreading of products and behaviors in society. During the seminar, the students will:

• Be introduced to various theories of social dynamics as well as their implications for policies aimed at social change.

• Be introduced to fundamental themes in modern social dynamics research.

• Obtain an overview of the impact of social network structures, social tie characteristics, and individual-level characteristics on the success or failure of new product or behavior spreading.

• Analyze, present, and discuss diverse research papers where the spreading of products and behavior has been investigated through agent- based simulations. • Learn how to code in Python agent-based models of new product or behavior spreading on social networks.

• Develop critical thinking about the main factors to consider when designing policies for social change.

Contact:
manuel.mariani@business.uzh.ch

Type:
Seminar

Target audience:
MA students with R or Python programming knowledge and familiarity with basic probability theory. No prior knowledge of social network methods or social dynamics modeling is required. Assigned to “Wahlpflichtbereich BWL 4”.

Frequency:
TBA

AP(ECTS)-points:
3

Language:

English

Course material:
TBA

Prerequisites:

Students can choose to program in either R or Python. Therefore, we recommend the following courses:
A non-technical introduction to R (or equivalent knowledge)
OR
A non-technical introduction to Python (or equivalent knowledge).

Recommended prior knowledge:
Good programming knowledge in R or Python. Basic knowledge of probability theory.

Grading:

Presentation and Coding (70%)
In-Class Participation (30%)

Dates and Location:

Location: AND 3.46 (except for the sessions on 11.10, 01.11, 16.11, and 17.11, which take place in AND 4.53)

Further information:
www.market-research.uzh.ch

Syllabus: Social Dynamics HS23 (PDF, 219 KB)

Registration:
The number of participants is limited to 8. Application for this seminar is via the new module booking tool: https://studentservices.uzh.ch/uzh/launchpad/. In your application, please include your CV, transcript, and a short letter of motivation.

Note:
This information supports the official information in the electronic university calendar (VVZ – Vorlesungsverzeichnis). In case of doubt, the official information at the VVZ is valid.