Couples don't actually look more alike over time: study
Stanford University researchers find people tend to pick partners with similar facial features as their own
Do couples really start to look alike the longer they're together? While it may seem that way, researchers at Stanford University have tested the age-old theory and debunked it with modern science.
"There are two parts to our empirical findings and what we learned. First of all, people tend to select partners who look similar to themselves," study author Pin Pin Tea-makorn, a Stanford electrical engineering PhD student, told As It Happens host Carol Off.
"The second thing we learned was that those couples don't become more similar over time."
The findings, co-authored by Tea-makorn and her advisor, Michal Kosinski, were published Monday in the journal Scientific Reports.
A small study that stuck
Tea-makorn, who uses math and engineering to analyze social phenomena, says her research was inspired by another study in 1987 by the late social psychologist Robert Zajonc.
According to Zajonc, married couples living together for over 20 years started to resemble each other because of how in sync they became with one another, unconsciously imitating each others' expressions, which gradually changed their facial features.
The idea has stuck in the popular consciousness, but when Tea-makorn looked at the research more closely, she says she found very little evidence in support of the theory.
"There were only a [small] number of couples that they used … 12 unmarried couples. That is very few compared with the usual psychology standard," she said. "Even other studies that are on the lower side of participants still have at least 30 to 40 participants."
Since 1987, there have been no other studies that fully replicated Zajonc's findings, she said.
"We were curious to dig deeper, using computer vision to kind of determine which facial features are the ones that actually converge, using more data," she said.
The Stanford study looked at facial images of more than 500 couples.
"We collected photos of couples at the beginning of their marriage and at a later point in time, at least 20 years. Then we have two independent methods of measuring the face similarities," she said.
"First, we actually replicated the 1987 paper where we asked human judges to rank how similar the couples are in relation to random couples. But in our second method, instead of using human judges, we used facial recognition algorithms to determine the similarity value."
They found many couples had facial similarities from the very start of their relationships, and that didn't change over time.
The researchers concluded that people tend to look for partners who look like them.
Why pick a partner who looks like you?
"There are both biological and sociological reasons people tend to select partners who are similar to them. Not just looks but also personalities, values and socioeconomic status," Tea-makorn said.
"Part of that is we are surrounded by people who are similar to us, either geographically or socially. People also develop preference for things that they are familiar with. Since we grew up being familiar with ourselves in the mirror, or our family members, we tend to develop likability to people who look similar to us."
She referenced celebrity couples Benedict Cumberbatch and Sophie Hunter, and Gisele Bundchen and Tom Brady. Although they will not grow to look more alike over time, Tea-makorn says they chose partners with similar facial features to their own.
Biologically, Tea-makorn says it makes sense.
"People or organisms in general tend to select mates that have similar genes so that they could perpetuate or make a gene present in later generations," she said.
While the Stanford study gives a clearer picture than the 1987 study, Tea-makorn says it's also limited in terms of data. The researchers only looked at photos of white, heterosexual couples living in the United States.
She says that with a significant enough amount of data, it would be worth studying this theory with people of colour and same-sex couples.
"I know there are tons of research questions that could be extended from this initial research," she said.
Written by Mehek Mazhar. Interview produced by Sarah Jackson.
