(Photo by Brooke Cagle on Unsplash+)
In a nutshell
- Humans tend to judge a dog’s emotions based on the surrounding context rather than the dog’s actual behavior, often leading to misinterpretations of their emotional state.
- People with more experience with dogs were surprisingly more likely to miss negative emotional cues, suggesting that experience doesn’t necessarily improve our ability to read dog emotions.
- To better understand your dog’s emotions, focus directly on their specific behavioral cues rather than the situation, and recognize that dogs experience emotions differently than humans do.
TEMPE, Ariz. — In the comfort of your living room, your dog tilts its head, ears perked and eyes wide. You decide, “Look how happy she is!” But are you really reading your dog’s emotional state correctly, or are you being swayed by other factors in the room—like the cheerful music playing or your own good mood?
A study conducted by researchers Holly G. Molinaro and Clive D. L. Wynne at Arizona State University suggests that humans often misinterpret dogs’ emotions based on external factors that have nothing to do with the dog itself. Their research, published in the journal Anthrozoös, shows that we’re not as good at reading our four-legged friends as we think.
“People do not look at what the dog is doing, instead they look at the situation surrounding the dog and base their emotional perception off of that,” explains Molinaro, an ASU Ph.D. student in psychology and animal welfare scientist, in a statement.
The researchers discovered that what we see as a dog’s happiness, fear, or anxiety may actually be more shaped by the surrounding environment and context than by the animal’s actual behavior. This matters for animal welfare, as misreading animals’ emotional cues could lead to misunderstanding their needs.
“Our dogs are trying to communicate with us, but we humans seem determined to look at everything except the poor pooch himself,” says Wynne, an ASU psychology professor who studies dog behavior and the human-dog bond.
Molinaro and Wynne’s study involved nearly 900 undergraduate students across two experiments. The first looked at how the presence or absence of visual context affected how participants judged a dog’s emotional state, while the second explored how perceptions changed when videos were edited to show the dog in mismatched contexts.
In the first experiment, 383 participants watched videos of a dog in both positive and negative situations. Some videos showed only the dog against a blank background while others showed the full scene. Participants rated the dog’s emotional state on scales of valence (how positive or negative) and arousal (how calm or agitated), and described the dog’s emotions in their own words.
The results were clear. When context was removed and participants could only see the dog, they couldn’t reliably tell positive from negative scenarios. More telling, when the context was manipulated in the second experiment with 485 different participants, people’s perceptions of the dog’s emotions matched what was happening in the environment rather than the dog’s actual behavior.
“In our study, when people saw a video of a dog apparently reacting to a vacuum cleaner, everyone said the dog was feeling bad and agitated,” Molinaro explained. “But when they saw a video of the dog doing the exact same thing, but this time appearing to react to seeing his leash, everyone reported that the dog was feeling happy and calm. People were not judging a dog’s emotions based on the dog’s behavior, but on the situation the dog was in.”
This tendency to attribute human characteristics to animals was widespread throughout the study, with over 90% of the responses containing anthropomorphic language. Participants often assigned complex human emotions and thoughts to the dog, such as “He’s excited about going for a walk” or “She feels betrayed by her owner.”
What makes this research interesting is how it challenges our assumptions about understanding animals. Many pet owners believe they have a special connection with their animals and can “just know” what their pet is feeling. Yet this study suggests that much of what we perceive may be influenced by our own biases and the situation around us.
In the second experiment, the researchers created videos showing a dog reacting to positive stimuli (like toys, leashes, or praise) and negative stimuli (reprimands, a vacuum cleaner, or a tape measure). They then mixed these videos to create situations where what the human was doing didn’t match what the dog was experiencing. For example, a person might be showing a vacuum cleaner (typically a negative stimulus), but the dog’s reaction was actually to seeing a leash (a positive stimulus).
The findings showed that participants consistently rated the dog’s emotional state based more on what the human was doing than on the dog’s actual behavior. This suggests that in everyday interactions with dogs, we might regularly misread their emotional states based on our expectations rather than their actual experiences.
Adding to the communication problem is our tendency to project our emotions onto dogs. Molinaro explained that while humans and dogs have shared a bond over the centuries, that doesn’t mean their emotional processing, or even emotional expressions, are the same.
“I have always found this idea that dogs and humans must have the same emotions to be very biased and without any real scientific proof to back it up, so I wanted to see if there are factors that might actually be affecting our perception of dog emotions,” Molinaro said. “If there were, if we as humans focused on other aspects not relating to the dog to deduce their emotional state, then as both scientists and pet owners, we really have to go back to the drawing board.”
Age and experience with dogs also affected perceptions. Younger participants (18-21 years) tended to rate dogs as having more positive emotions and higher arousal compared to older participants (22+ years). Additionally, those who reported being very familiar with dogs tended to see negative scenarios as less negative, which might indicate that experience with dogs doesn’t necessarily improve emotion recognition—it might actually lead to missing negative cues.
The results go beyond just misreading a pet’s happiness. If we consistently misinterpret animal emotions based on contextual factors, we might fail to notice when animals are in distress. This could have serious consequences for animal welfare, especially for animals under human care.
Consider a dog at a noisy family gathering that’s panting and wide-eyed. An owner might think this shows excitement and happiness to be part of the festivities, when the dog might actually be stressed and anxious from the overstimulation. By misreading these cues, the owner might keep the dog in a situation that causes distress.
The study challenges common theories about emotions. Traditional theories suggest that emotions are universal and can be recognized across species based on fixed behaviors. However, these findings align more with theories suggesting that emotional perception comes from multiple factors including context, culture, and past experiences.
Molinaro explained that even in studies of human perception of human emotions, it’s clear that there’s more to reading emotion than just looking at a person’s face. Culture, mood, situational context, even a previous facial expression can influence how people perceive emotions. Yet when it comes to animal emotions, no one has yet studied if those same factors affect us in the same way.
“Our research here shows that for one of those factors, the situational context, it does,” she noted.
Interestingly, when asked what cues they used to determine the dog’s emotions, participants initially said they relied on the dog’s tail and overall behavior. However, when context was reintroduced, they shifted to using the context and background instead, despite only 10% explicitly acknowledging this influence in the second experiment. This suggests that the impact of context may be somewhat unconscious or that people may not realize how much they rely on external cues.
So how can dog owners cut through these biases to better understand their pets’ emotional states?
“The first step is just to be aware that we are not that good at reading dogs’ emotions,” Molinaro said. “We need to be humbler in our understanding of our dogs. Once we can start from a basis of understanding our biases, we can begin to look at our pups in a new light.”
For dog owners, this research offers a chance to become more thoughtful observers of their pets. Instead of jumping to quick conclusions about a dog’s emotional state based on the situation, owners might benefit from learning more about dog body language and behavioral cues specific to dogs, not humans.
“Every dog’s personality, and thus her emotional expressions, are unique to that dog,” Molinaro explains. “Really pay attention to your own dog’s cues and behaviors.”
The study also raises questions about anthropomorphism in our relationships with animals. While attributing human emotions to animals can sometimes foster empathy and connection, it can also lead to misunderstandings if we project our own emotional experiences onto creatures that may experience the world differently.
“When you yell at your dog for doing something bad and she makes that guilty face, is it really because she is guilty, or is it because she is scared you are going to reprimand her more? Taking an extra second or two to focus on your dog’s behaviors, knowing that you need to overcome a bias to view the situation around the dog rather than the dog himself, can go a long way in getting a true read on your own dog’s emotional state, leading to a stronger bond between the two of you.”
Next time you look at your pet and feel certain about what they’re feeling, it might be worth stopping to think: are you reading their actual behavior, or are you being influenced by the context, your mood, or other factors unrelated to your pet? Your dog’s emotional world is likely more different from yours than you realize.
Paper Summary
Methodology
The researchers ran two separate experiments with different groups of participants. In the first experiment, they filmed a 14-year-old pointer/beagle mix dog in six different scenarios meant to create either positive emotions (treats, praise, play) or negative emotions (seeing a cat, a tape measure, or getting a reprimand). The videos were edited to create two versions of each scenario: one showing the full context including the owner and environment, and another showing only the dog against a black background. A total of 383 college students watched these videos and rated the dog’s emotional state on scales for valence and arousal. They also wrote descriptions of what they thought the dog was feeling. In the second experiment, the researchers created video edits where they mismatched contexts – for example, showing a person with a vacuum cleaner alongside footage of the dog actually reacting to seeing its leash. These videos were shown to 485 different students who completed similar rating tasks. Researchers also collected information about participants’ familiarity with dogs to see how these factors might influence perception.
Results
The results showed several key findings. In the first experiment, participants rated the dog’s emotions as more positive when context was absent, regardless of whether the scenario was positive or negative. When context was present, participants could better distinguish between positive and negative situations. For arousal ratings, videos of positive situations without context were rated as showing higher arousal than those with context, while negative situations without context were rated as showing lower arousal than those with context. The second experiment showed even more striking results: when the human in the video was doing something positive, participants consistently rated the dog’s emotional state as positive, even when the dog was actually filmed in a negative situation. Similarly, participants rated dogs in positive contexts as having lower arousal regardless of the dog’s actual behavior. Free-response analysis showed high levels of anthropomorphism across all conditions, with over 90% of responses containing anthropomorphic language. Demographic factors also influenced perceptions – younger participants (18-21) gave higher valence and arousal ratings than older participants, and those very familiar with dogs perceived negative scenarios as less negative overall.
Limitations
The researchers noted several limitations to their study. First, due to ethical constraints, the negative-emotion-inducing stimuli used were relatively mild, which may not represent how people perceive stronger negative dog emotions. The study also used only one dog of a specific breed and appearance, potentially limiting how well the findings apply to other dogs with different looks. Another limitation was the potential for multiple interpretations of the videos, as they contained various stimuli that could influence participants’ ratings. The researchers mentioned that while more controlled filming conditions could have reduced these issues, this might have made the study too artificial. The study sample was also limited to college students, so the findings might not extend to the broader population with more diverse experiences with dogs. Finally, the researchers acknowledged that the valence and arousal scales used showed a small but significant correlation with each other, which could indicate that participants weren’t perfectly distinguishing between these two emotional dimensions.
Discussion and Takeaways
The researchers emphasize that their findings challenge theories suggesting emotions can be universally recognized across species. Instead, the results align more with theories arguing that emotional perception stems from multiple factors including context and past experiences. An important takeaway is that people who reported being very familiar with dogs actually perceived negative scenarios as less negative – suggesting that experience with dogs doesn’t necessarily improve emotion recognition and might even lead to missing negative cues. This has significant implications for animal welfare, as misreading emotional cues could lead to failing to recognize when animals are distressed. The researchers suggest that we need to be cautious about anthropomorphizing animals and to develop better understanding of species-specific behavioral indicators of emotional states. Finally, they note that participants generally weren’t aware of how much context influenced their judgments, as only 10% reported using context/background cues despite evidence showing they heavily relied on these factors.
Funding and Disclosures
The researchers did not report any specific funding sources for this study in the paper. They also explicitly stated that no potential conflict of interest was reported by the authors. This suggests the research was likely conducted as part of their academic work at Arizona State University’s Psychology Department without external commercial interests.
Publication Information
This study titled “Barking Up the Wrong Tree: Human Perception of Dog Emotions Is Influenced by Extraneous Factors” was conducted by Holly G. Molinaro and Clive D. L. Wynne from the Psychology Department at Arizona State University. It was published in the journal Anthrozoös on March 10, 2025. The paper includes supplemental data that can be accessed online at the DOI link provided in the paper. The study was approved by Arizona State University’s Institutional Review Board for human subjects.
