Symposium on Opinion Shift, Social Influence, and External Manipulation
When: Wednesday 23rd April 2025, 12.30–4.30pm
Where: G.07 Informatics Forum (Central Campus), University of Edinburgh
Format: hybrid (the talks will be broadcast online, everything else is in-person only)
Registration: register here
About the symposium
Understanding how information and opinions propagate through society has become crucial in the current climate of polarisation, attempts of influence and disruption, and new communication paradigms brought by online social media. Social instability is on the rise to the point of threatening the core foundations of established democracies, and it is paramount to find mechanisms that can increase social resilience, learning, and decision making.
However, opinion dynamics is a wicked problem, as it encompasses a whole ecosystem of interacting actors. It requires looking at both the individual and collective scales, merging expertise from different disciplines, and a complexity approach that can distil the main patterns. At its core, it is important to understand the individual cognitive mechanisms of opinion formation and change, and how ideas and values align and interact. Equally important is the social aspect of opinion influence and agreement, what is expressed and how, and how specific communication patterns (by individuals or by algorithms) lead to aggregate social outcomes.
Another barrier to the study of opinion dynamics is the difficulty of acquiring data, as it requires clever approaches to mining data from databases or open servers and automating the extraction of stances from text. Furthermore, it is important that the assumptions, experiment design, and interpretation of results are embedded in the right social context of the data, which can be facilitated by frameworks from social and political science.
Why this event
The aim of this event is to bring together researchers from all career stages who study opinion formation, dynamics, and manipulation from different perspectives and disciplines. Despite some cross-disciplinary labs and initiatives already existing, there is still significant departmental fragmentation, and researchers working in this area have few chances to meet and hear about each other’s ongoing work.
The symposium welcomes diverse research approaches: quantitative and qualitative, theoretical and applied, simulation-based and data-driven. The following non-exhaustive list contains some broad areas that are relevant:
- Individual opinion change and social influence mechanisms (psychology, neuroscience, economics)
- Collective opinion change (opinion dynamics, agent-based models, echo chambers, polarisation)
- Measuring opinion changes in the real world (social media analysis, argument mining, surveys and polling)
- Interventions and external influence (policy implementation, security, political meddling and campaigning, targeted advertisement)
We also welcome attendees from other universities interested in the topic.
Agenda
| Time | |
|---|---|
| 12.30–1.00pm | Arrival & lunch (G.02) |
| 1.00–1.05pm | Introduction |
| 1.05–1.30pm | Mónica Tamariz. Social transmission: A cultural evolution theory perspective. |
| 1.30–1.55pm | Björn Ross. Revisiting the Spiral of Silence Theory: Research Directions for Computational Social Science. |
| 1.55–2.20pm | Clare Llewellyn. NRLabs: A Neuropolitical understanding of opinion and influence. |
| 2.20–2.35pm | Tea break |
| 2.35–3.00pm | James Stewart. Influence government: exploring how and why the British government uses targeted advertising platforms to shape awareness and behaviour of UK and foreign citizens. |
| 3.00–3.30pm | Panel session with the speakers |
| 3.30–4.30pm | Poster session |
Poster session
We will have a poster session, and you can express your interest in presenting a poster with your work in the registration form. You will need to bring your poster along on the day, and it is OK to reuse any recent poster you already have.
Organised by Guillermo Romero Moreno (Guillermo.RomeroMoreno@ec.ac.uk), funded by the Edinburgh Complexity.