The PolyGraphs team held the first of its two Royal Society sponsored public engagement workshops on 11 November 2022. Focussing on preliminaries, as well as group attitudes (including belief and knowledge), and the role of structural features of social networks in relation to these, the UK-based research team was joined by colleagues at the Center for Design at Northeastern University in Boston, and a range of stakeholders in the information-sharing ecosystem.

Founder and Principal of Northeastern University London, Anthony Grayling, opened the event, which also involved representatives from the tech sector, government and charitable oversight bodies, and the general public.

The team began by presenting the background to our research, including the models we employ, as well as previous results due to other researchers. Using animations,  and well as other visualizations (e.g. the one pictured left), we then engaged in structured interactions with participants to convey key aspects of our own findings, and to draw out insights from their knowledge of, and experience of, the online social networks and the media landscape.


Our workshop participants were experienced social media users.

  • 25% spent over 2 hours a day on social media.
  • 90% of our participants used more than one social media platform.

At the start of the workshop, 67% stated they did not understand Network Theory, and 33% stated they did not understand Bayes Rule. After the workshop these both fell to 0%. More specifically, we saw:

  • A mean increase to the likert scale of +1.73 for understanding of Network Theory.
  • A mean increase to the likert scale of +0.81 for understanding of Bayes Rule.

There was a pretty uniform spread of opinions on whether ‘adding more people into their social network would improve understanding’. Comparing how individuals responses changed before and after the workshop also differed – with people increasing their agreement almost exactly offsetting those who decreased agreement.

However, there was agreement that “how a community aggregates the opinions of its members makes a difference to what is known to that community”, and all participants felt at least as strongly about this statement after the workshop (mean increase to the likert scale of +0.54).

All participants said the workshop was engaging. And everyone agreed or strongly agreed that visualizations are helpful to understand features of the real world.

Introducing PolyGraphs

After the workshop, we spoke about our project with Claudio Lanza, Assistant Professor of Political Science at Northeastern University London. Multimedia Video Producer Mark Dutton recorded the exchange.

Brian discusses the Zollman Effect.

Alexandros and Brian discuss group beliefs.

We discuss the question whether we model emotion, as well as polarization.