The internet survey was conducted by Michele Wan and colleagues at Columbia University, New York, and was completed by 2,163 participants. There were 16 short video clips of dogs, sometimes with people, in various situations. They were shown with no sound, so people could only use visual signals. Several different dog breeds and mixes were shown in the videos so that the results would not be due to specific dogs or breeds.
After each clip, participants were asked to say what emotion the dog was displaying, which body parts showed this, how difficult it was to tell, and how accurate they thought their answer was. The emotions they could choose from were happy, fearful, angry, sad and neutral.
In fact, amongst the sixteen videos there were nine that a group of experts in canine behaviour consistently rated as showing a fearful dog (4 videos) or a happy dog (5 videos). It is people’s opinions of these videos that are reported.
Participants were also asked about their previous dog ownership and experience. They were then categorized into groups. Those with low experience had never owned a dog and had little or no experience with them (7% of the participants). Owners had owned a dog at some point in their life (68% of the participants). The professional group had worked with dogs in a professional capacity for between one and nine years (14%) and the ‘professional 10+’ group had worked with dogs for over ten years (11% of participants). Amongst the professionals, most worked in dog behaviour (e.g. training), but around 30% worked in some other capacity such as grooming.
When the dog was happy, participants had no difficulty in recognizing this, regardless of their level of experience. However, experienced participants were much more likely to recognize when a dog was fearful. These results still applied when the researchers took account of how likely people thought it was that a dog would experience happiness or fear (i.e. the experienced participants didn’t choose fear more often because they were more likely to think dogs could be fearful).
In addition, the results still held when they excluded people who had learnt about dog behaviour from books, lectures or classes – in other words, actual experience with dogs, rather than teaching about dogs, is enough to make a difference.
There were also differences in the parts of the dog that people referred to. In general, the experienced participants used more body parts in making their decision, and were more likely to pay attention to the ears. This applied to both happy and fearful dogs. For fearful dogs, all participants found aspects of the face such as eyes, ears, mouth/tongue most useful. In contrast, for happy dogs, the legs/paws and tail were most useful.
Finally, people were more confident of their decision about the happy dogs, and said they were easier to understand than the fearful dogs. Not surprisingly, the Low Experience group felt less accurate and found the questions more difficult than other participants.
These results are especially interesting since some previous studies found that experience with dogs had no effect. One advantage to this study is that the videos were consistently rated by experts as showing canine fear or happiness, so we can be confident the dogs were displaying those emotions. The results do concur with studies of human emotions, which show that differences in experience have an effect on emotion perception. It’s also the case that for humans, happiness is easier to recognize than fear. This suggests the human brain may process information about emotions in a similar way for people and other species.
An important implication is that learning how to recognize a fearful dog could help in dog bite prevention. Fearful dogs are a particular risk for biting, and yet many people were unable to recognize fear in a dog.
So, how can you tell if a dog is fearful? Signs to look for include a low tail, ears back and flat to the head, a frozen posture, crouched body position, shaking and panting.
What canine features do you use to help you recognize doggy emotions?
ReferenceWan M, Bolger N, & Champagne FA (2012). Human perception of fear in dogs varies according to experience with dogs. PloS one, 7 (12) PMID: 23284765