Do we know our friends as well as we think? What the research says
Almost everyone believes they know the people closest to them well. Psychology has tested that certainty several times, with couples, friends and roommates, and the result repeats itself: we do better than chance, but far worse than we believe. And confidence grows with time spent together faster than accuracy does. This article sums up four studies, shows what they suggest for an 8-question scoreboard, like the one in our friends challenges, and ends with what can be done about it.
Forty years together, and missing more
In a study published in 2011, the researchers Benjamin Scheibehenne, Jutta Mata and Peter Todd brought together 58 couples. Thirty-eight were young, aged 19 to 32, together for two years on average. Twenty were older, aged 62 to 78, together for almost 41. Each person had to predict how much their partner would like 118 items, among foods, movies and kitchen designs, on a four-point scale.
With four possible answers, someone guessing would get 25% right. The young couples got 42.2% right. The couples of four decades got 36.5%. And both groups believed they had got about 62% right. In other words, more time together did not bring more accuracy, and the certainty stayed the same.
More time, more certainty, the same accuracy
The result was not new. In 1997, William Swann and Michael Gill had already studied dating partners and college roommates and found the same split. The length of the relationship and the degree of involvement increased people's confidence in what they said they knew about the other person, but did not consistently increase their accuracy.
The explanation they propose is simple. With time together, the picture we have of someone grows richer, with more details and more stories. Richness gives a feeling of certainty, even when the details are out of date or were never checked.
It is getting hot in here
The third study is about what we mean. In 2011, Kenneth Savitsky and colleagues asked volunteers to convey a specific meaning using an ambiguous phrase. One example from the authors: saying “it's getting hot in here” to mean “open the window”. Each phrase had four possible meanings, and the listener had to pick the right one.
Speakers expected to be understood much better by a friend or a spouse than by a stranger. In practice, friends were only slightly better than strangers. Among couples married for 14 years on average, the spouse's accuracy was statistically the same as a stranger's. The authors called this the closeness-communication bias: with people who are close, we let our guard down and assume we were understood.
That is exactly the subject of our challenge “Can your friends decode you?”, in which you say what you mean by “I don't mind”, “I'll think about it” and “5 minutes”, and your friends try to translate.
What this says about an 8-question scoreboard
Our challenges have 8 questions, and in eight of the ten each question has four options, like the scale in the couples study. That makes it possible to work out what to expect. The numbers below are our own calculation, not a measurement: they show what would happen to someone who answered without knowing you.
- By pure guessing, the average is 2 right out of 8.
- In 89% of cases, someone guessing ends up with 3 right or fewer, in the two lowest result tiers.
- Reaching 6 or more by guessing happens 1 time in 237. Getting all 8 right happens 1 time in 65,536.
- If your friends were right at the rate of the couples in the study, close to 40% per question, the average would be a little over 3 out of 8, and a result of 6 or more would come up 1 time in 20.
- In the two challenges with two options, “This or that” and “Never have I ever”, guessing gives 4 out of 8, and 6 or more comes up 14.5% of the time. In those, only a 7 or an 8 says anything.
How to read a friend's result
Putting the two together changes the reading. In a four-option challenge, 4 out of 8 is already double what chance gives, not a weak result. Six or more, by the math above, would be rare even at the rate of couples who live together, and deserves a celebration. And a 2 or a 3 does not mean a lack of interest: it is what chance gives most people who guess.
One important caveat: the questions in the studies are not ours, and our challenges have not been tested in research. The comparison is there to give the scoreboard a sense of scale, not to draw conclusions about your friendship.
Imagining does not work; asking does
If we miss so often, what helps? The most common advice is to put yourself in the other person's shoes. In 2018, Tal Eyal, Mary Steffel and Nicholas Epley put that advice to the test in 25 experiments. Asking people to imagine someone's point of view did not improve their accuracy about what that person felt, thought or preferred. In some cases, it only raised their confidence.
The last experiment showed what works. Participants who were able to talk with their partner about the subjects before answering were more accurate. The authors' conclusion is that understanding someone takes new information, not more effort of imagination. In short: ask.
How to use this with your friends
None of this is a reason to distrust the people close to you. It is a reason to trade certainty for curiosity.
- Treat a miss as news. Each question a friend got wrong is something about you they did not know yet, and now do.
- Be wary of what you have known for years. Tastes change, and the answer that held in college may not hold today.
- Ask instead of deducing. “What do you need today?” is more accurate than any guess.
- Translate your own phrases. If your “I don't mind” does not mean you don't mind, say so to the people around you.
- Do not turn a scoreboard into a grievance. It measures eight answers on one day, not the size of a friendship.
A game that ends in conversation
The friends challenges on ShareQuiz are entertainment, not a psychological assessment or a test of friendship. What they do well is create the excuse for the question the research recommends: finding out instead of assuming. “How well do your friends know you?” is the most general one, and the “Couple test” takes the same idea into life as a couple.
Sources
The dates, names and numbers in this article were checked against these pages. Links open in a new tab.
- Scheibehenne, Mata e Todd (2011): Older but not wiser: Predicting a partner’s preferences gets worse with age. Journal of Consumer Psychology, 21(2)
- Swann e Gill (1997): Confidence and accuracy in person perception: Do we know what we think we know about our relationship partners? Journal of Personality and Social Psychology, 73(4)
- Savitsky, Keysar, Epley, Carter e Swanson (2011): The closeness-communication bias: Increased egocentrism among friends versus strangers. Journal of Experimental Social Psychology, 47(1)
- Eyal, Steffel e Epley (2018): Perspective mistaking: Accurately understanding the mind of another requires getting perspective, not taking perspective. Journal of Personality and Social Psychology, 114(4)
Written for entertainment and general information. It does not replace professional advice.