clock menu more-arrow no yes

Study: Dogs really do look like their owners

(Michael)

There's a widely held belief that dogs resemble their owners. And if you think this idea sounds pretty far-fetched, you're not alone.

But a dig into the research into this weird topic, as Jesse Bering recently did at Slate, turns up a surprising conclusion: there's actually some truth to it.

It's certainly not a perfect resemblance, but when looking at photos of various dogs and dog owners, people are consistently able to correctly match them up at rates greater than chance would dictate. The simplest explanation for why this might be so: on the whole, pets and their owners tend to look like each other.

And although it seems pretty crazy, there are a few logical explanations for it. One is that dogs and their owners share lifestyle habits, so obese owners are more likely to have obese dogs. But another, more important one is that dogs aren't randomly assigned to people, but chosen — and on the whole, we might unconsciously prefer dogs that look like us in the first place.

What research says about the dog-owner resemblance

A number of researchers have addressed this question with essentially the same strategy: they take pictures of both dogs and their owners and ask study participants to try to match them. And repeatedly, these participants are able to match them up at rates higher than chance.

dog owner research

A sample set of random dog and owner photos, unmatched. (Nakajima et. al. 2009)

This has been true for several different groups used in various studies: ethnically diverse Americans and their dogs, ethnically homogenous Japanese people and their dogs, and even dogs and owners participating in the Venezuelan national dog show.

In the Japanese study, 373 participants were shown a set of 20 correctly matched dog-owner pairs as well as another set of 20 incorrectly matched ones, and they were able to pick the right one two-thirds of the time.

These sorts of findings don't necessarily mean that dogs resemble their owners — there could theoretically be other clues tipping off the participants. But in the photos, the researchers stripped away the background and bodies of the dogs and owners, so all the participants could see were their faces. If there's another way they'd be able to figure out who was correctly matched with which dog, it's not clear what it might be.

So why do dogs and owners resemble each other?

dog 2

(myri_bonnie)

One contributing factor is health. Research has shown that overweight people tend to have overweight dogs, because these dogs don't get walked as much. Indeed, as Sarah Yager at the Atlantic points out, researchers have measured an overall increase in obesity among all sorts of pets as human obesity rates have climbed.

A more important factor may be preference — the idea that, in general, people tend to choose dogs that look a little like them.

As part of a 1999 study, 261 women were surveyed on their preferences among four breeds of dogs: two with what are called lop ears (which hang down) and two with pricked ears (which poke upward). On the whole, women with long hair — which the researchers believed looked more like lop ears — preferred the lop ear dogs (English Springer spaniels and beagles), while women with short hair were more likely to prefer the other breeds (Siberian huskies and Basenjis).

This type of preference isn't unprecedented. Other research has shown that we tend to sit near people that look like us and choose friends who share genetic similarities with us. At an unconscious level, it seems, we have an affinity for the familiar.

But with dogs, it might go well beyond their ears matching our hair style. A recent Japanese study highlighted a particular feature that seems responsible for much of the dog-owner resemblance: the eyes.

The study was similar to the earlier photo-matching ones, except that in some cases, the researchers blocked out parts of the human and dog faces: either just the eyes, or the mouth, or everything except the eyes.

When the whole faces were shown, participants matched the correct dogs and owners 80 percent of the time, and when just the mouths were blocked, they still got it right 73 percent of the time. When only the eyes were shown, they got it right a remarkable 75 percent of the time — but when the eyes were blocked out, the accuracy rate dropped down to the levels predicted by random chance.

If it truly is our preferences responsible for the dog-owner resemblance, it seems that we're unconsciously choosing dogs with eye qualities similar to our own.


Update: The reference to the Atlantic article Why You Look Like Your Dog was added.