Despite our strong belief in dogs’ ability to sense good from bad in people, new research shows they may not actually judge human character—at least not in the way we think.
When dogs watched how humans treated other dogs, they didn’t favor the kinder person later. Even direct interactions didn’t sway their behavior. The study suggests dogs’ reputational judgments might be more nuanced—or harder to study—than we realized.
Do Dogs Judge Human Character?
Many people put a lot of faith in a dog’s instincts when it comes to judging character. If a dog immediately warms up to someone, it’s often seen as a sign that the person is kind and trustworthy. On the other hand, if a dog seems hesitant or wary, some might take it as a red flag. But whether dogs actually assess people’s behavior in this way is still not fully understood.
Research has shown that certain intelligent and highly social animals, like chimpanzees, are capable of forming opinions about humans based on personal interactions or by watching how others are treated. Because dogs have lived so closely with humans for thousands of years, scientists have been especially interested in whether dogs can do the same. However, the results from studies so far have been mixed.
One earlier study from the Wolf Science Center in Austria found that both domesticated dogs living in packs and wolves did not form reputations of individual humans, even after direct interaction or observing how those humans behaved. This lack of judgment could be due to the limited exposure those animals had with humans, which led researchers to design new experiments involving dogs with more human experience.
Testing Dogs’ Reputation Judgments Through Observation
To explore this further, scientists studied 40 pet dogs to see whether their ability to form opinions about people changes with age or development. The researchers set up a situation where each dog could observe another dog interacting with two people. One person behaved generously by feeding the other dog, while the second person did not offer any food.
After watching this exchange, the observing dog was allowed to interact with both people. The researchers closely monitored the dogs’ behavior, noting which person the dog approached first, whether the dog jumped up, and how physically close it stayed to each individual.
No Clear Preference for Kind Humans
The results revealed that dogs across all age groups did not significantly prefer the generous person who fed the dog compared to the selfish person who refused to feed the dog. Their behavior toward the two humans did not exceed chance levels following indirect eavesdropping or direct experience with them.
“It’s clear that reputation formation may be more complex than previously thought, even for animals like dogs that closely cooperate with humans,” says corresponding researcher Hoi-Lam Jim, who recently joined the faculty at Kyoto University.
Limitations in Experimental Design
This study highlights the methodological challenges in accurately capturing dogs’ understanding and evaluation of humans. Since direct reputation formation is a prerequisite for eavesdropping, the researchers were not surprised to find no evidence of reputation formation, but they did not expect the lack of evidence for it after direct interaction.
“It is possible that methodological challenges in the experimental design, particularly the use of a two-choice test, may explain our negative findings, rather than an absence of capacity,” says Jim.
Improving Future Research on Dogs’ Social Cognition
To better understand what influences dogs’ sociocognitive abilities, the research team says future research should systematically compare dogs of all ages from different populations and life experiences, expanding to include free-ranging dogs, service dogs, and police dogs.
To many of us, it seems reasonable to believe that dogs socially evaluate people, but for now we simply do not know.
Reference: “Do dogs form reputations of humans? No effect of age after indirect and direct experience in a food-giving situation” by Hoi-Lam Jim, Kadisha Belfiore, Eva B. Martinelli, Mayte Martínez, Friederike Range and Sarah Marshall-Pescini, 28 June 2025, Animal Cognition.
DOI: 10.1007/s10071-025-01967-w
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
24 Comments
This article is so full of bull .. I’m 71 and a retired engineer and I have had dogs in my life for my entire life and there is no doubt that these animals can be a good judge of character.
Your comment is auto ignored for being named Karen 🤣🤣 u r right about the dogs tho. This was a basic ass study.
I’d like to consider myself a good person. Most dogs like me but some are still scared. I’ve never been truly bitten however but came close to it. I wonder if some of the dogs interactions are from the owners being “abusive.” Where they get slapped just for doing something that the owner doesn’t like and has grown to be fearful of people in general.
Obviously the study authors have never had any k9 family members.
I like this page very informative and great topics
I’m very much liking this page seems I. Learn something new every day
This study is not targeting the way dogs think! Pretend selfish and pretend generosity is not a valid test. I really wish I could help this study with my knowledge of dog behavior and the way dogs actually think and yes sense! Contact me please! I can create a much more valid study of “dog sense in people”. 1. Testing all dogs for this study. 1a. Sensitive personality type dogs are going to display the best results. 1b. Stoic personality types are not especially if the stoic dog is extroverted/more dominant temperament. 1c. The most sensitive and meek are going to judge the most critical of humans because in a pack environment it’s these dogs who are naturally geared to detect “danger, threat, or no threat”. It’s this personality and temperament type who also can and will detect illness in other dogs and people.
2. You need to first establish bad from good by testing with people known to be bad vs people known to be kind : for example go get unkind people from a mandatory anger management center and go get people from somewhere that has unamously determined they are good or kind (psych test first both test subjects) then ask the dogs to tell them apart by reaction of the dog! Test this many times with these same individuals on different dogs then test different individuals with the same dogs-compare the results. For instance my dog can and does do medic alert and therapy work, he’s an excellent judge of a person what he looks for is a gentle temperament, i manipulated by food he will avoid a person who is in an angry mood even if that person has not yet told us they are mad. He will also tell me to leave the room with angry people even while they are still silent. He is an excellent detector of danger or no danger! ‼️ Another interesting test would be to test blind dogs against the finding of what my dog’s test reveals to determine what they see when they can’t even see! Follow me at JenKellerDogTraining on FB for more interesting dog perspectives. PS I am more of a behaviorist first trainer second.
Agreed! I had a very kind easygoing dog who accepted everyone, but once we had a workman in the house who got his hackles up big time. Totally out of character. My dog knew what was in that man’s thoughts and it wasn’t good:
dogs see auras. testing for that would be a great idea.
I agree, this test seems very unthorough. I’ve had instances with both dogs and cats that could only be explained by ESP so I’d like to see that accounted for as well in these studies that look at how and if dogs judge us.
Also, I am a kind person and love animals (non-humans) more than anything in the world. Literally. But I also tend to be an angry, irritable person. So getting angry people to participate in the test will not necessarily work. Anger only translates to unkindness if the person acts on the anger.
Nothing is worse to a thinking person than someone who declares that only cause of an outcome can exist, such as your claim that *only* “ESP” (never proven to even exist) must be the cause. Just because you misinterpret reality and logic doesn’t mean it has to follow your misinterpretations.
Correction: “only ONE cause of an outcome can exist. (Too bad this site doesn’t allow comment correction.)
Just because a person gives the dog a treat or not doesn’t judge a person ‘s character. Try using an inherently angry or even a violent person as one of the subjects. Besides, it also depends on the dog itself and its own temperment.
I am not a dog owner, infact, I am afraid of dogs. However, I had never been bitten by a pet. Infact, my sister’s dogs demand my attention and ask me to pet them. They fight with me for not giving them attention. That happened the first time. People around me should say if I am kind, but I can say that I am not abusive.
I love my dogs more than myself, I’ve gone homeless for my old girl when given the ultimatum of getting rid of her to stay where I was living at the time. Having said that simply so I’m not taken wrongly here, there’s a big difference in “not offering them food” being correlated with “being a good person”, versus “a dog seeing you hit and yell at another dog and then making the decision to stay away from you or alert other’s about you
This is also only observing modeled behaviors, not the intuition that dogs can sometimes pick up on about a sketchy person or situation. Not saying they’re fool proof, but they’re paying attention more than we are most of the time, they *want* to guard us and guide us and protect us, it’s what’s created that inseverable bond through all these millennia with them ❤️
I’ve had many dogs over the years and I trust their instincts about people. On the other hand, I’ve seen dogs that are imprinted to some people I know to be very bad news.
Dogs can feed themselves. Their judgment of your character does not rely on whether or not you feed them. The study did get one thing right: dogs’ ability to judge people is a lot more complicated than they thought it would be.
So, bottom line, the study told us nothing — that we don’t know one way or the other. A useless study, and a useless article.
I believe that dogs are actually a lot like people,in that when it comes to random humans that they may not have spent an awful lot of time around. Tendancy is often getting to know them a little bit at a time, and don’t just have a first impression sort of thing. But some dogs have a much better intuition than others which goes without saying. Having said that I believe that if a dog has a bond with his owner and meets a new person when they are all intermingling, the dog would tend to be sensitive to his person’s emotional vibes. Either laid back, or anxious… and that definitely puts most dogs in sword in a standby protective mode their person. Just some of my observations I know that there are a lot more variables that play into it, but that’s just sort of my general idea is about it.
A flaw in the study is that the researchers used 2 humans and had them “pretend” to be good or not good. Dogs treated them no differently because neither of them were *actually* “not good”. Dogs don’t judge acting. They detect energy and things under the radar of most human perception.
Dogs are dumb.
Yes, they are. But that doesn’t mean they can’t perceive gross behavioral traits like being abusive. Unfortunately, this study implies they can’t even judge that, when all they did was control the treats the dogs were given.
I think what’s much more likely is that when a dog is bonded to you it picks up on your emotions and judges based on your feelings of the person. This creates a confirmation bias of people whose dog disliked someone that they themselves disliked and assume the dog must have some innate ability to sense the value of someone’s character. But if that were true then how would we explain violent criminals and fascists who have or had dogs that love them?
This is just my opinion about dogs that love unconditionally that owner feeds them or the dog doesn’t know any better than the life they r living dogs generally r loyal ,what’s your opion ?