Why your friends may be more susceptible to social influence than you are

Regarding sensitivity to influence on social networks, “it is not only who you are – it is a question of knowing where you are in a network and to whom you are connected,” said Luca Luceri, principal scientist of the Institute of Information Sciences of the USC (ISI).

A new study by Luceri and his team at ISI reveals that the probability that someone will be influenced online does not spread evenly on a social platform. Instead, he gets cluster.

They call this the paradox of sensitivity; It is a model in which user friends are on average more influenced than users themselves. And this can help explain how behaviors, trends and ideas are accompanied – and why some internet corners are more vulnerable to influence than others.

The document, “The Sensibility Paradox in Online Social Influence”, was co-written by Luca Luceri, Jinyi Ye, Julie Jiang and Emilio Ferrara, and was presented at the International Conference of AAAI 2025 on the web and social media (ICWSM), held from June 23 to 26 in Copenhagen, Denmark. It was also published on the arxiv pre -printed server.

The paradox of influence

To explore how this dynamic works, the researchers examined two types of behavior on X / Twitter: sharing focused on influence, when people publish something after seeing it others in their network; And spontaneous sharing, when they publish without this exhibition.

They found that in cases focused on influence, people less likely to be influenced were often surrounded by others who were more likely to share what they saw. This reflects the paradox of friendship, a discovery of the science of the network which says that your friends are likely to have more friends than you.

“What makes it interesting is that sensitivity is not only a personal trait,” said Luceri, who is also an assistant research professor at the USC Viterbi School of Engineering. “This is something that you can deduce only from the influence of users of your network, which surrounds you or interacts with you.”

Where does the sensitivity are grouped?

The team also found that these more influenced users do not appear to chance on the network. Jinyi Ye, ISI research assistant and first -year doctorate. The USC computer student said that she had been struck by the way these users have remained closely.

“We could see that they were connected and influenced in the same way,” she said.

This model, known as Homophilia, was particularly strong in sharing focused on influence. People who display because they saw others were often part of united circles with similar behavior. This suggests that social influence works not only by direct exchanges between individuals, but is also shaped and limited by the structure of the network.

Predict that will share

The researchers also examined if it was possible to predict which is likely to share content according to social influence. For behavior focused on influence, they found that sensitivity measures alone were very predictive.

In many cases, knowing how the friends of a user behave enough to estimate how the user would behave. Spontaneous sharing was different. When people shared content without apparent exposure by peers, their decisions were more difficult to predict the network alone. In this case, user features, such as account metadata and personal traits, were also very informative.

This contrast strengthens a key point: certain online behaviors are more shaped by the social environment of a person, while other behaviors are summed up with individual characteristics.

What is the next step

The team now examines how sensitivity changes over time. “We want to see if people become more or less influenced depending on what is happening in their network,” said Ye. And although the results are theoretical, the authors note their potential applications of the real world.

“Understanding how and where the influence is focused on a network is a first step towards the attenuation of its harmful effects,” she continued. The cartography of the way in which online sensitivity clusters could shed light on disinformation, support public health campaigns or identify communities that can be particularly vulnerable to influence.