Facebook Friendships and Social Influence: Guest Blog

We followed 1,700 interconnected college students on Facebook. Their profiles listed favorite bands, movies, and authors. But we wondered how their tastes might spread.
This post was published on the now-closed HuffPost Contributor platform. Contributors control their own work and posted freely to our site. If you need to flag this entry as abusive, send us an email.

Blog written by Nicholas Christakis and James Fowler *

An invisible web of intricate social relationships influences us all. Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives chronicles how human behavior, emotions, and attitudes spread in "real world" social networks examining how internet connection and contagion are rooted in evolution.

We followed 1,700 interconnected college students on Facebook. Their profiles listed favorite bands, movies, and authors. But we wondered how their tastes might spread, imagining that contagion between Facebook friends would resemble that of the real-world friends of past studies. Strangely, we found that not one of the listed bands, movies, or authors had a noticeable effect.

Past work shows that social contagion happens best between people with close relationships. Strangers can influence our thoughts and actions only when mediated through a sequence of close ties -- your friend's friend's friend or your child's friend's parent. We suspected relationships to 1,000 online "friends" might be so tenuous that their influence is less powerful than real-world connections.

So which Facebook connections are important, real-world connections?

We tracked the photos on Facebook -- because when you upload and tag someone's picture, you probably have a face-to-face relationship. We found that while students in our data generally had hundreds of Facebook friends, they had only six "picture friends," a number very similar to the "close" friends identified in previous studies of real-world social interactions. When we narrowed our analysis to "picture friends," we found that some online phenomena spread in interesting ways.

Our students' ten favorite bands were:

1.The Beatles
3.Dave Matthews Band
4.Green Day
5.Jack Johnson
7.Led Zeppelin
8.Red Hot Chili Peppers
9.Simon and Garfunkel

Only two of these bands showed signs of contagion. If a person liked The Beatles, it doubled the likelihood that a friend would like them the following year. Meanwhile, if a person liked the Killers, it increased that likelihood 11-fold. We also found that contagious appreciation of The Beatles originates in lots of different corners of the network simultaneously. But fans of the Killers were grouped into tight clusters within isolated sections of the network.

The top 10 movies were:

1.Fight Club
2.Garden State
4.Good Will Hunting
5.Lord of the Rings
6.Love Actually
7.Pulp Fiction
8.Shawshank Redemption
9.Star Wars
10.Wedding Crashers

Of these, only three movies showed evidence of person-to-person spread. People with a friend who liked Love Actually were 4 times more likely to list it as a favorite the following year. Those with a friend who liked Pulp Fiction were 5 times more likely. Those with a friend who liked Good Will Hunting were 11 times more likely.

Unlike musical tastes, movie tastes appear to be polarized. Few people had friends who listed both Pulp Fiction and Love Actually. Moreover, these movies generally affected different parts of the network. Pulp Fiction was listed by more "popular" people at the center of the network, while Love Actually was preferred by wallflowers on the periphery. Marketers often think they can create buzz by targeting influential people at the center of a network, but it turns out that we are influenced both by who we are connected to and how we are connected.

The top 10 authors were:

1.Jane Austen
2.Dan Brown
3.Fyodor Dostoevsky
4.F. Scott Fitzgerald
5.Harper Lee
6.George Orwell
7.J.K. Rowling
8.J.D. Salinger
9.William Shakespeare
10.J.R.R. Tolkien

Only two of these showed person-to-person spread. Those with a friend who listed Salinger were 4 times more likely to list Salinger as a favorite the following year, while those with a friend who listed Shakespeare were 9 times more likely.

Although there's evidence that taste in some authors can spread online, there's no specific pattern. Shakespeare's fans are as dispersed and as likely to be in the center as Salinger readers, and both groups are highly interconnected. Because book tastes are so eclectic, people usually use music and movies to define their place in a network. This is probably because music and movies are activities that are often enjoyed together by two or more people. People can discuss books, but reading is more solitary.

When we want to figure out why someone likes this musician or that movie, it's best to look at the tastes of those around them. But not just anybody can influence us. We can have 1,000 Facebook "friends," but, on the internet, influence works the same way it has always worked offline.

Those who matter most are closest to us...

Diagrams created by the authors that show how tastes in music, movies, and books cluster can be accessed on Facebook here.

* N. A. Christakis is a palliative care specialist at Harvard University.

James Fowler studied quantitative techniques of predicting American elections, and was able to achieve remarkable success and accuracy in predicting electoral outcomes.

Both authors were introduced in 2002 by Gary King, Director of the Harvard's Institute for Quantitative Social Sciences about which much more in future blogs...

Popular in the Community


What's Hot