Back to search

SAMRISK-2-Samfunnssikkerhet og risiko

Emotional Contagion (EmotiCon): Predicting and preventing the spread of misinformation, stigma, and anxiety during a pandemic

Alternative title: Emotional Contagion (EmotiCon): Predicting and preventing the spread of misinformation, stigma, and anxiety during a pandemic

Awarded: NOK 3.1 mill.

Much misinformation, stigma, and anxiety spread in the wake of the COVID-19 pandemic. The 'Emotional Contagion' (EmotiCon) project aimed to explain how and why this occurs in Norway. The main goal was to develop a multi-agent artificial intelligence model designed to predict and prevent, or at least mitigate, this sort of spreading as societies prepare for the next pandemic or the next wave of COVID-19. EmotiCon was run through the Center for Modeling Social Systems at NORCE. The data informing the development of the EmotiCon model came primarily from extensive social media analysis and a new survey of Norwegian attitudes and behaviors during the national 'dugnad' in response to the pandemic. The model development was shaped by collaboration with a team of international advisory collaborators, and subject matter experts who represent ten municipalities in Norway. The simulated agents in the computer model have cognitive architectures and weighted social network ties that affect beliefs and behaviors based on a variety of social psychological theories. They have been empirically calibrated in relation to real world data about the Norwegian population. The practical purpose of the EmotiCon project was to provide stakeholders with an empirically validated 'artificial society' that can serve as a simulation platform within which they can experiment with intervention strategies designed to mitigate the spread of anxiety, stigma, and misinformation during future pandemics or other crises. The EmotiCon team has so far had three academic articles accepted for publication, all currently in press: one on the model itself, one on the participatory process, and another on the issue of vaccination hesitancy. Three additional articles are under preparation. The EmotiCon project has also presented its work at four conferences so far. It has produced one internal policy paper for stakeholders, and is planning others. A final meeting with stakeholders is planned for March 2022.

The EmotiCon project achieved all of its overall objectives and work-package outcomes. First, it provided an analysis of social media communication about the COVID-19 pandemic on Twitter, identifying several distinctive aspects of this discourse in Norway compared to other Scandinavian as well as non-Scandinavian countries. Second, it produced an analysis of a new, longitudinal panel survey of a representative sample of the Norwegian population, identifying and comparing factors related to willingness to follow the social restrictions encouraged or required by the government. Third, based on the data above, it developed an agent-based computational model that represents a "digital twin" of Norway. Researchers and policy professionals can experiment on this "artificial" Norwegian society to discover the conditions under which - and the mechanisms by which - the spread of misinformation, stigma, and fear is likely to decrease or increase in the population.

The EmotiCon project has four interrelated work packages (WPs). WP1 organizes and administrates the grant and disseminates the project's findings. WP2 and WP3 involve the development and execution of social media analysis and an online survey of a representative Norwegian sample, respectively. WP4 will construct a multi-agent artificial intelligence model for analyzing and forecasting the spread of anxiety, stigma and misinformation in social media and offline networks in the wake of COVID-19. That computational model and its simulation experiments will be informed by the data from WP2 and WP3. All WPs will involve the close collaboration between the core research team at NORCE, their international advisory collaborators, and ten municipalities in Norway that have already agreed to participate in a user reference group (these municipalities represent about 30% of the Norwegian population). The Emoticon project will provide new computational tools for assessing and altering the dynamics of emotional and behavioral contagion during public health crises. Our team has already published computational models with the ability to simulate the effect of disease contagion threats on the attitudes and behaviors of human populations. Simulated agents have cognitive architectures and weighted social network ties that affect beliefs and behaviors based on social psychological theories such as 'terror management theory.' They have been empirically validated in relation to real world data. We will adapt these models to simulate the social contagion effects of disease contagion threats under a wide variety of parameters, including those of poorer countries. This will provide stakeholders with an empirically validated 'artificial society' that can serve as a simulation platform within which they can experiment with intervention strategies designed to mitigate the spread of anxiety, stigma, and misinformation during the COVID-19 crisis and future pandemics.

Publications from Cristin

No publications found

No publications found

No publications found

No publications found

Funding scheme:

SAMRISK-2-Samfunnssikkerhet og risiko