How do we choose friends?

When asked what they are looking for in a friendship, people frequently say “loyalty”. They expect their friends to be loyal. Fair enough - who wants to be friends with a person who is likely to abandon them in times of need and go spend time with someone else?


When asked why people are friends in the first place, they commonly say that this is because they are similar. Indeed, (perceived) similarity is a predictor of friendship. This is a fancy way of saying that people who think they are similar are likely to initiate conversations with each other and eventually become friends.


This marks the end of me saying obvious things. Let me now say things that are less obvious.


There are 4 aspects of similarity that we can consider:

  • (a) similar interests and hobbies,

  • (b) similar personality,

  • (c) similar values,

  • (d) similar thoughts and world perception.


So which one of them is the most important? In other words, do you become friends with someone because you are both attending the same ECA (a), or because you are both similarly outgoing and extraverted (b), or because you both value freedom of choice and animal rights (c), or because you both view the world as a fragile place where we must preserve what we have been given (d)?


Studies* have shown that the answer depends on how long you’ve known each other. Let’s take three time points: T1 = one month of knowing each other, T2 = three months, T3 = eight months.


At T1 people become friends based on (a). After just one month in the same school / class / neighborhood we get together based on common interests or activities. We become friends with Jessica because she attends similar ECAs and because we share the same bus on our way home. Other factors - (b), (c) and (d) - do not matter. We actually don’t know much about these other factors because it takes a while to know people, to find out what their values are and what they think about afterlife and the role of women in science. We start talking, and we gradually find out.


At T2 people form friendships based on (c) - similar values. We have known each other for 3 months, and we have had a chance to hear each other’s opinions on a variety of things - for example, things they say during lessons. We think, wow, that person’s opinions are similar to mine, and we seem to find similar things in life important. Those are the people you want to spend time with. Studies show that, if we have a choice, we abandon friendships that we formed at T1 and gradually change into these new friendship patterns. Sorry Jessica, I know we attend the same ECA and share the bus on the way home, but I want to spend more time with Anabelle because we are both animal right activists, so we have stuff to talk about.


At T3 (eight months of knowing each other) people form friendships based on (d) - a similarity of world perception and other deep thoughts. We realise that we are very similar to Christina on some very deep level. She has the same kind of quiet skepticism combined with the same determination to make the world a more honest place. Christina is not an animal rights activist, and she takes a different bus route, but she is the one we now want to be friends with. Sadly, neither Jessica nor Anabelle have the same (d) as we do. We wish we noticed Christina earlier, but how could we? Our paths barely ever crossed.


But the tragedy is, it’s been 8 months, so we have already made friends. The T1-friends and the T2-friends. And these friends value loyalty. If we abandon Jessica and Anabelle and go hang out with Christina, they will both think of this as betrayal, and we won’t feel great about our behaviour either.


So is loyalty to friends such a great thing after all?


P.S. An attentive reader has noticed that (b) - similar personality - does not matter at all.

By Mr Popov

* Note: this text is based on the following psychological research study:

Duck, S.W. & Craig, G. (1978). Personality similarity and the development of friendship: A longitudinal study. British Journal of Social and Clinical Psychology, 17, 237-242.