New social media to drive positive behavior change

Collective behavior change is one of the most important and difficult challenges in public health. It is a matter of cooperation, but according to Yori Ong, a dentist and physicist at the University of Groningen, getting people to cooperate on long-term collective interests is not an easy task. In a new M20 study, Ong and colleagues look into this problem.

Collective behavior change is about changing the behavior of groups of people. Behavior patterns that belong to a group are more easily retained than adjustments in an individual's behavior. Ong explains it as follows: “As a dentist, if I urge all my patients to brush their teeth better or eat healthier, I am still addressing their individual behavior. But people are group animals: everyone influences each other and mimics each other to some extent. Nowadays, you can easily reach large numbers of people with information or education, but its effect is largely determined by the social environment a person is part of and what behavior is considered normal within it. Collective behavior change is about influencing those interactions between people that ensure that behavioral norms are maintained or changed. This is necessary to bring about sustainable behavioral change on a large scale.”

Ong initiated the project together with adjunct professor of artificial intelligence Davide Grossi. In the new M20 research, the duo, together with a yet-to-be-appointed PhD student, aims to lay the scientific foundation for an effective digital communication protocol that can be used to drive collective behavioral change.

A complex concept

Scaling up behavior change in the interest of prevention and public health has proven to be a very difficult task. According to Ong, this is because we have a limited and fragmented understanding of what exactly collective behavior is and how it works. "You can study collective behavior from the perspectives of psychology, sociology and economics, as well as biology, physics and artificial intelligence. That fragmented understanding makes it difficult to come up with working interventions. If you want to build a car that runs, you have to understand how all the parts interact to get the whole thing running. If one person understands how the engine works and someone else knows all about the fuel, you will still get nowhere without direction", explains Ong. "In this project, we are connecting the extremes. I look at the practical side of collective behavior from dental practice, but also the abstract side from physics. This in turn fits well with Davide's expertise, which deals with collaboration from a logical and mathematical perspective," says Ong.

Just because it is difficult does not mean that collective behavioral change is not possible at all. In fact, Grossi says we have already experienced it. "Just think about social media. Over the past 20 years, new communication networks have led to large-scale behavioral change, including in the political and cultural spheres. Social media is simply not optimised for encouraging cooperative behavior, but for making profits from advertising.’

New research

In their new research, the researchers aim to give the strength of digital social networks a new function: not to maximize profits from advertising, but to promote cooperative behavior with public health as its focus.

This starts with an algorithm that lets people form networks with a structure that makes it easy for collective behavior to emerge. "First of all, it is important to look at the structure of such a network. For example, you have networks in which a small number of people have many more connections than average, but you can also have networks in which everyone has about the same number of connections. That structure determines what information circulates easily and how people relate to each other," Grossi explains.

The researchers focus on the NK Boolean network model, which was created back in 1969 by physician and biologist Stuart Kauffman. With it, he described how simple interactions between genes in DNA can lead to complex cooperation. "This model can be used to describe collective behavior in different natural systems, but a communication network for humans has never been designed after it. We created such a design with different use cases at an earlier stage and now we can start investigating what such a network would do to people and why", says Ong. Once the scientific basis is well described, the researchers also want to actually develop a complete digital product that can be tested by large groups of people for different purposes.

Interdisciplinary work

The new research is an interdisciplinary project in which mathematics, computer science, physics, behavioral sciences as well as dentistry all play an important role. "As a dentist and a physicist and I am not a fan of separating different disciplines. It is easy to separate them, but in this project everything has to come together," Ong said.

According to Grossi, this interdisciplinary nature is almost a requirement these days. "Today's challenges in our society have to be solved in an interdisciplinary way. There is no discipline that has a complete view on any of these complex problems. This is true for behavioral change, but also for other complex problems."

The project is expected to begin at the end of this year when the PhD student is appointed. "We already have an excellent candidate in our sights for this project", Ong says. 

Ong explains why he is so pleased with the M20 scholarship. "I am at the start of my academic career. I have noticed that in the post-doc phase, it is very difficult to have one leg in one field and another in another. To get funding or a research position, you usually have to apply somewhere where you are judged from a narrow perspective, whereas working solutions have to be right from both fundamental science and practice. In the M20 programme, interdisciplinarity is at the very centre, so that's why it's really perfect for me."