Animal welfare is essentially about how animals feel. I often say that if we could just ask the animals how they feel, my job would be simple. In reality, psychologists interested in animal welfare have been excited about the idea of using cognitive biases to measure mood in animals for several years. Cognitive biases are basically scenarios under which our brains don’t process things entirely accurately, but they are inaccurate in ways that are predictable based on our experiences.

For example, if you are in an anxious mood, you will be more likely to perceive harmful or scary things around you. Your brain is on the lookout for danger more so when you’re feeling anxious than when you're feeling calm. Remember when you would race down a dark hallway at night to get to the bathroom as a child, and every shadow and sound seemed to signal a monster? Your mood was anxious, and your attentional biases toward danger gave it away through your behavior. If a scientist was measuring your behavior (how fast you ran, how easily you startled), they would have some insight into your mood.

One cognitive bias that animal welfare scientists have been super psyched about is called judgment bias.  It’s the bias we refer to when we talk about some people seeing a glass as “half full” and others seeing the same glass as “half empty,” depending on their attitude. Scientists can tap into that phenomenon and ask animals what they are perceiving to deduce how they are feeling.

This approach was exciting but it had a bit of a rocky start as researchers tried to train the animals to tell us what they were perceiving without calling their attention to the question at hand. (Luckily) biases can be reduced if you are aware of them. Then there was a breakthrough. One of the first studies that figured out how to ask the dogs what they perceived without extensive training was by Mike Mendl and colleagues from the University of Bristol in the UK. I remember being excited and thinking, ok, we’re on to something we can use now!

In this task, domestic dogs would come into a room and sometimes find a tasty bowl of food on one side of the room, and other times would find an empty bowl on the other side of the room. The dogs couldn’t see how the bowls were baited. The dogs were repeatedly able to head off from a start position and learn that if a bowl was on one side, there was going to be something fantastic there (yummy food), while a bowl on the other side signaled disappointment (it was empty).  Logically, over time, the dogs learned these locations and would travel more quickly to the side that always had fantastic food, and saunter slowly to check out the side that was probably empty.

Then, the mood test. How fast did a dog approach when a single bowl was placed right in the middle of the two learned locations? This is an ambiguous location, never learned to be associated with something positive or negative. The dog’s speed revealed his mood. Move quicky as he did toward the bowl with food, the dog is feeling optimistic. Move slowly as he did toward the empty bowl, the dog is feeling pessimistic. The dogs revealed their mood to the researchers with their speed of approaching the new location.

In a more recent study, Charlotte Duranton and Alexandra Horowitz of Barnard College used this same approach to ask how dogs’ moods changed after having two weeks of practicing either “nosework” or “heelwork” with their owners at home. The authors thought that nosework, in which dogs were trained to locate hidden food, would put dogs in a better mood because it allowed them to express behavior that is natural for them, and allowed them some freedom and autonomy. The heelwork, on the other hand, allowed the dogs to spend the same amount of time with their owners and earning treats, but didn’t have the benefits of allowing the dogs to express natural behaviors and autonomy.

The authors tested the dogs before and after their two weeks of work. They found that dogs who were in the nosework group moved more quickly toward the ambiguous middle bowl than they did before their nosework began, whereas the heelwork group didn’t show the same increase in speed. Dogs who had been doing nosework appeared more “optimistic.” The dogs that practiced the skill that was natural and provided them some control were in a better mood.

Over recent years, judgment bias has been tested in monkeys, honeybees, horses, pigs, and countless other species. I find the Duranton & Horowitz study described here to be extra exciting because it starts to measure the value of allowing natural behaviors and autonomy for the animals in our care. Often the changes we measure in judgement bias tests are small, and we don’t yet know how much change in mood it takes to change behavior in these tests. However, every time we have another strong study showing an easier way to ask animals how they are feeling, we make progress toward being able to care for animals in a way that promotes great welfare.

Curious to hear your thoughts, ideas, or questions. Feel free to comment below!

Learn more about the researchers

Charlotte Duranton, Ethodog, Paris, France

Alexandra Horowitz, Barnard College, Columbia University, New York, USA

Article

Duranton, C., & Horowitz, A. (2019). Let me sniff! Nosework induces positive judgment bias in pet dogs. Applied Animal Behaviour Science, 211, 61-66.

Sources & Notes: Study demonstrating that cognitive bias revealed by dogs’ speed approaching bowls: Mendl, M., Brooks, J., Basse, C., Burman, O., Paul, E., Blackwell, E., & Casey, R. (2010). Dogs showing separation-related behaviour exhibit a ‘pessimistic’ cognitive bias. Current Biology, 20(19), R839-R840. Learn more about Mike Mendl or the Bristol Veterinary School’s research on Animal Behavior & Welfare.