July 3, 2025, 12:20 pm | Read time: 5 minutes
Many have long believed that dogs possess an exceptional ability to sense who has good intentions toward them and who does not. But is this really the case? Can dogs truly judge whether a person is kind or mean—and act accordingly? This intriguing question is at the heart of a recent study that challenges dogs with a seemingly simple decision. The results are surprising to many…
As a dog owner, I often hear people say things like, “Yes, dogs know exactly who is nice and who means well” or “Dogs have an infallible sense of whom they can trust and whom they can’t.” But is that really true? A research team from the University of Veterinary Medicine Vienna wanted to find out and investigated at the Clever Dog Lab whether dogs can infer a person’s “character” from their behavior—a phenomenon known as “reputation formation.”
Can Dogs Judge Whether a Person Is Nice or Mean?
In the study published in the journal “Animal Cognition,” researchers tested whether dogs learn to distinguish between a “generous” and a “selfish” person through direct experience or merely by observing (eavesdropping). For the first time, they also examined whether the age and life experience of the animals play a role. The findings provide new insights for research but also raise methodological questions.
Many animal species live in groups and benefit from cooperation—but only if all members behave fairly. In the wild, animals closely observe who shares and who doesn’t, forming an opinion, or “reputation,” about their peers. This principle is also central to human social relationships. The research aimed to determine whether dogs, who have lived closely with us for millennia, are also capable of socially evaluating human behavior.
Previous studies yielded mixed results—some found evidence of reputation formation in dogs, others did not. Criticism often focused on the lack of realistic or methodologically sound experimental setups. The ability to “eavesdrop,” or observe interactions between others, is cognitively demanding. The current study builds on this discussion but expands the perspective to include a previously overlooked aspect: the role of age and individual life experience in dogs.
Also interesting: “An animal medium connected with my dog, and this is what he ‘said'”
What Was Investigated—and How?
The study tested 40 family dogs (20 male and 20 female) in three age groups: “young” (1–3 years), “adult” (4–7 years), and “senior” (8–12 years). The animals participated in two different experimental conditions: In the eavesdropping condition, they observed two strangers (one generous, one selfish) interacting with another dog—one person fed, the other refused food. In the direct experience condition, the dogs themselves experienced these situations with both people.
A control group saw the same gestures from the people, but without a dog, to rule out that dogs were merely reacting to movements. In several test phases, researchers analyzed which partner the dogs preferred and whether they showed more friendly behavior (such as proximity, physical contact). The tests were conducted outdoors under animal-friendly conditions.
The Study’s Results
The analysis clearly showed: None of the dogs significantly preferred the generous person over the selfish one—neither after observation nor direct experience. Age also played no role. Neither young nor old dogs chose the “friendly” person more often. They also did not spend more time with them. Only three out of 40 dogs showed a statistically noticeable preference—two for the generous person, one for the selfish person.
Overall, the distribution of decisions in the group was random. Even with twelve direct contacts with both people, the animals could not make a clear distinction. There were no differences in the analysis of eye contact, physical proximity, and other friendly behavior. A side finding: Many dogs showed a side preference (e.g., always to the left person), which could partly be explained by lighting conditions outdoors.
Significance of the Results
The results challenge the widespread assumption that dogs can form human reputations—at least under the tested conditions. Although dogs are evolutionarily closely linked to us and often appear very cooperative, the researchers found no evidence that they systematically distinguish between “nice” and “unfriendly” people. This suggests that this cognitive ability in dogs is either absent or difficult to demonstrate under experimental conditions.
Notably, even direct experience with both people—multiple rewards or ignoring—was not enough to develop clear preferences. This could mean that the test setup did not provide the right context or that dogs need more or different types of information to evaluate behavior. They may respond more strongly to negative stimuli (such as rejection), as other studies suggest, or require more repetitions over longer periods.
Considerations for the Study
The study was carefully designed to avoid previous methodological weaknesses, such as consistent control of location, clothing, positions, and uniform test execution. However, limitations remain: The number of interactions may have been too low to trigger sustainable learning processes. Additionally, both experimenters were female—studies show that dogs can more easily distinguish between people of different genders.
The use of food, which can strongly capture attention, might have overshadowed differentiated evaluations. Lastly, the dogs’ motivation may have been too low—they were not hungry, not stressed, and quickly adapted to the test environment. The outdoor setting also offered distractions. A peculiarity: Many dogs showed side preferences, possibly because one side offered more shade. Although these effects were considered in the analysis, they may have had subtle influences. Overall, it highlights how challenging it is to measure social judgment processes in animals under controlled conditions.
Paw preference of dogs depends on the owner
Cockatoos Season Their Pasta with Vegan Blueberry Yogurt
Dogs in India Prefer the Color Yellow to Food
Conclusion
The study provides no evidence that dogs—regardless of age—can socially evaluate human behavior in terms of reputation. Neither through observation nor through personal experience do they prefer the generous person over the selfish one. This underscores the complexity and susceptibility to interference of such cognitive tests. Future studies should consider longer learning phases, stronger contrasts between test subjects, and different types of dogs (such as free-roaming dogs).
For dog owners, this means: Even though dogs seem very attentive, they may not evaluate our behavior as distinctly as we sometimes believe. At least not when it comes to feeding by strangers.1