Modelling Social Networks Reveals How Information Spreads

The way information spreads through society has changed significantly over the past decade with the advent of online social networking. But it seems we humans continue to apply the same approach that we always have to spreading information, even if we do it through Facebook, Twitter and Tumblr these days instead of leaning over the garden fence.

By Philip Garnett and John Bissell, Durham University

But the increasing complexity of our social networks means we might not be truly aware of who is influencing us and how. If asked to track a particular interest or fashion through your social network, how many different links would you have to go through to find the original source of the information?

All social networks are significant in shaping us as individuals. We are influenced by our friends (both real and virtual), colleagues, neighbours and our community so while we might be approaching communication in the same way as before, its effect on us may be different.

Modelling social networks

One model of how information spreads rapidly though social networks is the neutral model. This essentially states that a lot of human behaviour can be caused by people simply copying each other. This model works well where there is no particular advantage to one choice over the other (choice neutrality).

At a workshop for young people during the 2013 British Science Festival, we set up some basic interactive experiments to model how information spreads in small groups of students. The games were designed to show how important social networking and learning are to the spread of information.

Groups of 40 or more students and teachers were given a simple card game called “colour spread”. Three people were each given a coloured card and were given the role of the “seed”. Their task was to convince as many of their peers as possible to take up their colour, so that they would subsequently spread it further through the network. We modelled the spread of these colours through the population by entering them into a network model, with each person represented as a node in the network.

In this version of the game, choosing a colour doesn’t really put you at an advantage or disadvantage, it is simply a colour; there is no incentive or advantage of one colour over another. Although it could be said that having the colour that dominated (the winner) at the end of the experiment was enough of an incentive. People may have an innate desire to belong to a successful group, but the colour itself makes no difference.

When an incentive was provided (lolly) nearly everyone chose the same colour. Tipping Points project, Durham University

If the agents in the network are given an incentive such as a lolly, the model can change quite radically, especially if the agents are allowed to cheat. For the incentive model, the people with the colour card that spread the most would win a lolly. In this game students were allowed (by innovating behaviour) to cheat by convincing others to trade in their card for a different colour. During the incentive experiments we would either tell young people they could collaborate (cheat) so everyone could win, or see if they innovated novel ways to spread the colour on their own.

Since social copying is a key aspect of our behaviour as humans we were also curious to see how they learned socially amongst each other. For some schools this went over especially well and everyone won the prize at the end by adopting the same colour. What we learned together was that certain members of the network were excellent sales people, causing colours in the social network to spread rapidly.

Example of model where students were allowed to ‘cheat’ making it possible to trade in their colour for the same as everyone else. Tipping Points project, Durham University

Connectors matter

Malcolm Gladwell’s book The Tipping Point describes different types of people that are important in the spread of information; one such type is “connectors”. Connectors are important in social networks as they influence large numbers of people. In our experiments students who spread their colour to multiple people are examples of connectors. Connectors are found throughout history; in business, government, education and revolutionary movements. These individuals tend to have a talent for making friends at a number of different levels, increasing the breadth and depth of their social network, and therefore the spread of information.

In our experiments the connectors were essential to spreading information within social networks. In many cases the colours that would spread the most tended to have the connectors with the most social influence. In the context of history, connectors are among the people who are pioneers of social change. We need only look back at political change within global society today to find examples of connectors that played a significant role in affecting that change at the time.

Experiments of this type (perhaps on a larger scale) could help develop our understanding of how our society functions in the 21st century. The speed at which information can spread, and the fidelity of the spread of that information, is important to perhaps all aspects of society. For example, there have been a number of hoax reports of celebrity deaths on twitter that have gone viral. The viral spread of false information could have unforeseen consequences. It is important that we educate people so that they are aware of how information spreads through networks, and also make them think critically about what they read.

Funding for this activity was provided by the Leverhulme Trust as part of the Tipping Points project at Durham University.

John Bissell does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.