Senior Research at Universitat Politècnica de Catalunya
Key takeaways
A new social compass model studies how extreme opinions evolve, and how these opinions might be depolarised.
It is necessary to establish a multidimensional modelling framework that takes account of the interdependence between certain social issues.
The polar representation suggests that individuals with strong convictions are less likely to change their opinion than individuals with weak convictions.
An initial polarised state can transit to a depolarised state thanks to increased social influence.
This transition depends on the strength of initial opinions: it may be first-order (highly divergent opinions) or second-order (correlated opinions).
doctoral student at the MRC Cognition and Brain Sciences Unit at the University of Cambridge
Key takeaways
The mental health of adolescents has been deteriorating in recent years, with an increase in anxiety, depression, and suicide rates.
Most studies report an association between mental health problems and time spent on social networks, but none have identified a causal link.
To better understand the impact of social networks, we need to look at how they are used, the type of activity they engage in, and the type of content observed and shared.
Chatting with friends or family is an activity on social networks that is associated with better mental health.
Conversely, cyber-bullying or exposure to shocking content has a negative impact on teenagers' mental health.
With social networks, it is easier than before to pick up on protest signals as well as for groups to organise themselves, facilitating the demonstration.
The Yellow Vests (“Gilets Jaunes”, in French) are a perfect example of online and offline mobilisation with nearly 4 million members in Facebook groups and over 300,000 people present on the first day of mobilisation in the streets.
As a result of the MeToo movement, there has been a significant rise in sex crime complaints in the US – around +20% between 2017 and 2018 for New York City, for example.
Even if MeToo did not trigger street protests, the stock market crashes and bankruptcy of Harvey Weinstein's company are examples of its effect on the ‘real world’.
Researcher in the Physics Department at Sapienza University
Key takeaways
With the advent of online platforms, opinions are becoming polarised on many subjects.
Recommendation algorithms, which recommend specific content likely to appeal to users, are one of the main causes of this.
Using analytical and numerical techniques, researchers have simulated the evolution of user preferences according to algorithmic recommendations.
By identifying these strategies, the study could help to develop less-polarising algorithms in the future by increasing user engagement.
This is another step towards creating a more balanced and inclusive online information ecosystem.
Contributors
Germain Gauthier
Assistant Professor at Bocconi
Germain Gauthier is a professor at Bocconi University. He obtained a doctorate in economics from École Polytechnique in 2023. His research lies at the intersection of public economics, political economy and applied econometrics. In particular, he has studied the determinants and consequences of various social movements, such as the #MeToo movement.