Kerswell, Bennett et al. (2009) — Anthrozoös
Self-Reported Comprehension Ratings of Dog Behavior by Puppy Owners
Published: August 25, 2026
Most dog owners believe they understand their dog’s emotional state well. Research on what they actually use to make those assessments tells a different story — one where the signals owners report attending to are mostly the loudest, most obvious ones, while the earlier and more diagnostically valuable signals go largely unnoticed. A study of puppy owners attending socialisation classes has produced findings that have direct implications for how dog behaviour education should be approached and framed. 🐾
Researchers Keven Kerswell, P. Bennett, K. Butler, and P. Hemsworth surveyed 89 puppy owners at veterinary clinic socialisation classes in Melbourne, asking them to rate their comprehension of their dog’s emotional states across six categories — Happy, Sad, Anxious, Angry, Friendly Greeting, and Anxious Greeting — and to list the specific behaviours they used to judge each state. The combination of self-rated comprehension scores and listed behaviours allowed the researchers to compare what owners believed they could read with what they reported actually reading.
The Gap Between Confidence and Accuracy
In most emotional state categories, owners’ perceived comprehension — how well they thought they understood the emotion — was not closely related to the number of behaviours they could report using to judge that state. This is a specific form of calibration failure: owners who reported high comprehension did not report a correspondingly richer repertoire of behavioural signals than owners who reported lower comprehension. Perceived understanding and actual signal repertoire did not track together.
The behaviours that owners did report were mostly restricted to vocalisations and gross body movements — the most obvious, high-amplitude signals that dogs produce. Barking, growling, tail wagging, and extreme postural changes were what owners described. The subtler, earlier signals — the micro-changes in ear position, the slight tension in facial musculature, the shift in weight distribution, the freezing or slowing of movement that precedes more obvious stress or conflict signals — were largely absent from the reported repertoires.
This matters because the subtle signals are the most clinically valuable. A dog who is beginning to feel anxious, uncomfortable, or threatened does not immediately show those states through obvious vocalisations or extreme postural changes. The early signals are subtle, brief, and easily missed — and they are precisely the signals that, if read and responded to, prevent escalation to the more obvious and more dangerous expressions that occur later in the arousal sequence. An owner whose signal repertoire consists only of late-stage expressions is an owner who consistently misses the intervention window that early recognition provides.
What the Demographics Showed ⚠️
The self-reported comprehension scores were associated with the age and gender of the participant and with previous puppy ownership. Older participants, female participants, and those with prior puppy ownership reported higher comprehension — consistent with the intuition that accumulated experience builds perceived understanding. What the study cannot establish is whether those higher perceived comprehension scores reflect greater actual signal literacy or simply greater confidence built on accumulated exposure without necessarily greater accuracy.
Anxious Greeting was the one emotional state where the relationship between self-reported comprehension and behaviour repertoire functioned differently from the others — the anomaly the researchers note without fully explaining. Anxious Greeting may be more distinctly recognisable to owners than other anxious states because it occurs in a specific, repeated social context where owners have more opportunity to develop pattern recognition, or because the behaviours involved are more salient than those of less contextually obvious anxiety.
The finding that dog morphology was not related to self-reported comprehension is practically significant. Owners of dogs with more expressive facial anatomy — more visible eye whites, more mobile ears, longer muzzles — are not more confident in their emotional reading than owners of dogs with less expressive anatomy. This does not mean morphology is irrelevant to signal transmission — the research on brachycephalic breed signal modification is well established. But it suggests that self-confidence in reading the dog is shaped more by owner variables than by dog variables in this sample.
The Education Challenge This Research Identifies 🐕
The most practically significant contribution of this study is its identification of a specific educational challenge. Owners who report high comprehension have low motivation to seek further education about dog behaviour — if you already believe you understand your dog well, there is no felt need to learn more. The problem is that high self-reported comprehension and low actual signal literacy can coexist, and the owner experiencing that combination has no subjective awareness that their understanding is incomplete.
The researchers suggest that an approach to lowering owners’ perception of their comprehension — helping them realise what they are not seeing — may be more effective at increasing the uptake and impact of dog behaviour education than approaches that assume owners will seek out education based on a felt deficit they do not currently experience. This is a counterintuitive but well-grounded educational principle: creating productive uncertainty before offering information produces better learning outcomes than offering information to a confident audience who does not feel the need for it.
At Zoeta Dogsoul, genuine attunement to the dog’s emotional state is the foundation on which NeuroBond is built. And attunement begins with accurate signal literacy — with seeing what the dog is actually communicating in the early stages of emotional arousal before it becomes the obvious, high-amplitude expression that everyone can read. The research covered in this series on human perception of dog emotions has consistently shown that owners miss these early signals, are influenced by context and mood in ways that distort what they do read, and overestimate their own accuracy. This study confirms that pattern specifically in puppy owners attending the classes most associated with improving it. The gap between perceived and actual comprehension is where the most important dog behaviour education work needs to happen. 🐾
Source: Kerswell, K. J., Bennett, P., Butler, K., & Hemsworth, P. (2009). Self-reported comprehension ratings of dog behavior by puppy owners. Anthrozoös. Published June 1, 2009.
