Click here for free access to our latest coronavirus/COVID-19 research, commentary, and news.

Support nonprofit science journalism

Science’s extensive COVID-19 coverage is free to all readers. To support our nonprofit science journalism, please make a tax-deductible gift today.

Keep in touch. A diagram showing the number and frequency of calls between British secondary students (yellow dots) and their friends and family (blue dots) before they left for university.

Keep in touch. A diagram showing the number and frequency of calls between British secondary students (yellow dots) and their friends and family (blue dots) before they left for university.

Jari Saramäki

The Only Thing Constant About Friendship May Be the Number of Your Friends

Friends come and go, but the number of close friends you have may remain surprisingly constant. That's the main result from a new study in which researchers used cell phone data from British secondary school students as they transitioned to university to track how many close social connections they maintained. The research also suggests that people have distinct social “signatures,” or patterns of intimacy with others, which they tend to maintain over time.

Regardless of how many Facebook friends a person has, most people maintain only a small number of emotionally intense relationships, says Jari Saramäki of the Aalto University School of Science in Espoo, Finland, a computational scientist who studies social networks. Studies show that these close relationships are vital to our health and well-being. But they do have a cost: time and effort that take away from work, personal time, or other relationships. So individual differences in the number of close ties likely reflect each person's ability to divvy out the finite resources of time, communication, and emotional investment that are required to sustain relationships, scientists say.

Given the importance and difficulty of creating bonds that sustain us, Saramäki and colleagues wanted to know what would happen to the social networks of students as they transitioned from secondary school to university, a period of significant flux.  They recruited 24 students ages 17 to 19, half male, half female, and gave each an 18-month contract from a major mobile phone company with 500 free monthly minutes and unlimited texts. All the students lived in the same city in the United Kingdom when the study began; by the end, 10 had left for universities in other parts of England. When given the phones, each student completed a questionnaire listing the names and phone numbers of all their unrelated friends and work and school acquaintances, as well as all their known relatives. Then, they ranked all of those individuals on a 1 to 10 scale of emotional closeness, with 10 signifying someone “with whom you have a deeply personal relationship.” 

Each student was asked to fill out the same questionnaire again, 9 months later, and a third time, after 18 months. Higher “emotional closeness” ratings on the surveys paralleled number and duration of calls throughout the study, reassuring the researchers that frequent phone calls were a reasonable indicator of intimacy. By analyzing the students’ phone invoices, the researchers were able to construct a ranking system for each contact based on the frequency and duration of calls.

Although there were high levels of turnover in the names in each individual’s network, the basic characteristics of the network itself—how many people a person called and how much time they spent on the phone with them—remained the same throughout the 18-month period. For example, a person's top three contacts typically got 40% to 50% of the person's calls, the authors report today in the Proceedings of the National Academy of Sciences. Saramäki is careful to note that the study didn’t establish why students rearranged their social lives, but one blunt interpretation is that as subjects' circumstances changed, they didn't pine for their old best friends—they simply replaced them with new ones.

Parents, siblings, and other relatives were least likely to precipitously lose rank or simply fall off the call lists, Saramäki says. “To keep rank, you need to be a relative.” But the striking stability of individual preferences for organizing—and limiting—social interactions suggests to him that each person has their own distinct preference, or perhaps capacity, for maintaining those bonds.

“It would be very nice to see what happens for people of different age groups and social situations” to see if the same pattern persists, Saramäki says. After people leave college, for example, their networks might shrink or “get more and more frozen” as time goes on, he suggests. Saramäki hopes in future research to study larger groups of people over longer periods of time, and to include other forms of communication such as social media.

The findings “do suggest a certain ‘budget constraint’ in the number of ties one might maintain, which has been suggested by studies before, but is still not universally accepted,” says David Lazer, a network scientist at Harvard University. He agrees with Saramäki that “you have to be very careful” about extrapolating from the results of the study because of its small, homogeneous sample.  Whether such a constraint exists may also depend on the type of communication being used, he says—it is not possible to spend 24 hours a day talking on the phone, for example.