Behavior and Training · Behavior Assessments · Canine Cognition

Science Says: “Nose Work is Good for Your Dog!”

We started training nose work games with our dogs about 8 years ago. One of the many neat things about this activity is observing the different search styles of each dog. For example, Chippy, our Toller, was very focused and methodical. He stopped and thoroughly sniffed each box before moving on to the next, finally stopping and indicating once all possible hiding places had been suitably examined.

Chippy’s nose work motto was: “No sniff shall be ended before its time.”

In contrast, our Golden, Cooper, lopes around the entire training center, jauntily sniffing each box as he passes by, suddenly dropping into a down (his indicator behavior) next to his selected choice. (A loud bark is often involved as well).

And then, there is Alice…… Watching Ally enjoy Nose Work is like watching a circus performance:

One of the many benefits of nose work is that dogs of any breed, age, or training level can learn and enjoy this activity. In AutumnGold’s nose work classes, we enroll adolescents to seniors, and work with herding breeds, sporting dogs, scent hounds (naturals at it) and even, this coming session, a Scottish Deerhound and an Ibizan Hound! One thing appears to be consistent – without fail, as soon as a dog learns the purpose of the game – to find stuff with her nose – it is game on!

Do dogs benefit from nose work? One often hears nose work enthusiasts expound about numerous benefits of nose work for dogs; increased confidence, enhanced enjoyment of learning, improved ability to work independently, and a strengthened bond between the dog and his owner. However, until recently, no one has actually studied pet dogs’ reactions to nose work games. Enter two of our favorite dog science researchers – Charlotte Duranton of Ethodog in France and Alexandra Horowitz of the Dog Cognition Lab in New York. In their new paper, these two scientists study one potential benefit of nose work – its influence upon dogs’ emotional states (1).

Some Background: The objective of their experiment was to examine whether or not being given opportunities  for olfactory enrichment – exploring and foraging using nose work games – influenced dogs’ emotional states in a positive way. To measure this, the researchers used a validated experimental tool called the Cognitive Bias Paradigm. In this test, animals are first trained to discriminate between two stimuli. The first is consistently associated with a positive event such as the arrival of a food reward. The second is paired with a negative event, such as an unpleasant taste or the absence of a food reward. After this training, the individual is presented with a stimulus that is ambiguous; it appears to fall somewhere in between the two conditioned stimuli. The animal’s response to this ambiguous stimulus – whether he/she approaches it and how quickly they do so – provides an estimate of whether the subject is responding positively (is experiencing an “optimistic” emotional state) or negatively (is experiencing a “pessimistic” emotional state). This test is used regularly to assess the welfare and emotional state of non-human animals that are kept in captivity, but until recently, had not been used with dogs.

The Study: A group of 20 adult dogs of varying breeds, living in homes with their owners, were recruited for the experiment. The dogs were randomly assigned to one of two groups:

  1. Experimental group (nose work training): The 10 dog/owner pairs in this group attended two weekly group classes. In the first, owners were introduced to nose work games and the dogs were trained to search for a high value treat hidden in one of several boxes. During the second class, searches increased in complexity. The owners practiced assigned nose work exercises daily with their dogs during the weeks following each class.
  2. Control group (heel work training): The dogs and owners in this group also attended two group classes. In these classes, loose lead walking was taught using a reward-based approach (high value food treats as +R). The owners practiced the assigned heel work exercises on a daily basis.

None of the enrolled dogs had experienced previous nose work or heel work training and the owners were unaware of (blinded to) the objectives of the study.

Cognitive bias test: At the start of the experiment and again after the two-week training period, each dog was tested with a cognitive bias test. The dogs were trained to expect either a high value treat in a bowl positioned 3 meters away to one side, and to expect an empty bowl placed the same distance away in the opposing direction. Training was conducted until the dog consistently anticipated either finding a yummy treat (yay!) in one bowl or nothing (sigh) in the other bowl. Following successful +/- bias conditioning, the test trial was conducted. An empty bowl was placed exactly in the middle position between the positive and negative bowl positions (i.e. an ambiguous position). The dog’s behavior with this stimulus was observed and recorded by a researcher who was blinded to the treatment groups of the dogs. They recorded latency (the time that it took for a dog to initiate approach and look inside the bowl), the dog’s behavior with the bowl, and behavior signs of stress or anxiety. Each dog acted as his/her own control by comparing results in the pre-test (before training in nose work or heeling) with those of the post-test (after training).

Results: Well, nose work had an effect, it seems:

  • Pre-training: No difference between the experimental and control groups responses to an ambiguous bowl were found. This means that there the two groups of dogs started out at the same place – they approached the ambiguous bowl in a similar period of time.
  • Post-trainng: After the two-week training period, dogs in the nose work group approached the ambiguous bowl more readily (i.e. had a shorter latency period) compared with their approach prior to nose work training. In other words, after just two weeks of learning nose work games, the dogs approached a bowl that they thought may (or may not) hold a food item for them faster than they did prior to training. In contrast, the heel work group showed no change in this behavior.
  • Interpretation: In the world of cognitive bias testing, this result – dogs deciding more quickly to approach an ambiguous bowl after learning  nose work – is interpreted as an increase in “optimism”. This means that the dogs were more likely anticipating a reward (happy emotional state) than anticipating nothing (neutral or unhappy emotional state). The fact that dogs in the control group did not alter their behavior supports the conclusion that the nose work training caused this change. Alternate (and not necessarily mutually exclusive) explanations for the results are that the dogs in the nose work group developed improved cognitive abilities from opportunities to search using their noses and/or that they had learned to enjoy foraging independently of their owner’s direction and more readily engaged in this activity post-training.

Take Away for Dog Folks: Although this was a relatively small study (20 dogs total), it is important to note that the study design and its execution were well controlled and properly blinded. The researchers also used a cognitive bias test that has been used and validated with many other species. Although those stuck in the behaviorism camp of “we can never speak about animal emotions” may argue about interpreting these results as reflecting a positive emotion in dogs, I would counter that similar constraints exist when interpreting results of studying human emotions. Yeah, yeah, I know, humans can communicate verbally and so can tell us how they feel inside. (News flash – humans are also highly capable of lying). The fact that emotional states are difficult to study in others (both human and non-human animals) does not mean that they do not exist or that they cannot be studied. You just have to be careful – just as these researchers were. 

The bottom line? Dogs who were trained to enjoy nose work, after a period of just two weeks, were more willing to investigate a stimulus of uncertain meaning. These results suggest that nose work training encouraged the dogs to work independently, to make choices on their own, and to check out something with autonomy. Perhaps they were also experiencing something akin to optimism. Personally, after watching lots of dogs learn and enjoy nose work games, I would call that emotion joy! 

Here is one more example – Senior dog Gracie and Amanda enjoy a search:

Happy Training!

Cited Study:  Duranton C, Horowitz A. Let me sniff! Nosework induces positive judgement bias in pet dogs. Applied Animal Behavior Science 2019; In press.


New! Science Dog Webinars! Click Here for Information about Upcoming Topics!


Interested in learning more about dog training and behavior? See Linda Case’s newest book – “Dog Smart: Evidence-based Training with The Science Dog“!


10 thoughts on “Science Says: “Nose Work is Good for Your Dog!”

  1. I do nosework with my dogs, and I am certain that they enjoy the activity tremendously. From personal observation, I have no doubt that nosework is a great thing to do with pet dogs and that it enriches their lives.
    However, I am not convinced that the study as done actually demonstrates what the authors state in the abstract (“We conclude that allowing dogs to spent more time using their olfaction through a regular nosework activity makes them more optimistic.”) or at the end of the article (“To conclude, the present study shows for the first time that practicing nosework increases positive judgement bias — levels of “optimism” — in pet dogs”). What you called the “bottom line” is true, but, in my reading or your post, you were more descriptive of the behavior than conclusive of the impact of nosework on the dogs’ optimism. In teaching nosework, the dogs are being taught to investigate on their own, and leaving the owner to approach the bowl would be similar to approaching nosework boxes, for which the dogs were rewarded when they found a hide. In heeling, the dogs are being rewarded for staying at their owners’ sides; leaving the owner to investigate a bowl is something the dog would expect to have no reward; teaching heeling is teaching the dog not to leave the owner to investigate or approach something else. Therefore, are dogs who were taught heeling even a valid control group? I see them as having been taught a behavior that would reduce the chance they would approach the bowl, not as a neutral comparison group. Wouldn’t one have expected the nosework group, rewarded as a result of investigation, to be more likely to approach the test bowl than would be dogs rewarded for staying at their owners’ sides, or even dogs with no training/rewards at all, without invoking optimism? Are the conclusions of the study valid from the design of the study itself?

    Liked by 1 person

    1. That was my reaction, too — the nosework behaviour taught closely resembles the conditions in the bias test. I wonder why this aspect of the experiment was not considered; it would be great to hear from those who ran the study about how they think about this, and what other ways there may be of measuring optimism/bias 🙂

      Liked by 1 person

    2. Hi LS – Thanks for your comment. You make good points (and in fact, I agree with you). I cannot speak for the researchers of course (or unfortunately, share the entire paper yet because it has not been published). However, in the paper, the researchers do describe the background, use and interpretation of cognitive bias tests in detail. They make the point that an individual’s emotional state (positive/optimistic vs. negative/pessimistic) influences judgement and reaction to novel or ambiguous stimuli. Hence the cognitive bias paradigm test and its use to assess welfare of captive animals. In this particular study however, as you aptly note, it is quite possible that the type of training (nose work) is confounded by the form of the cognitive bias test (investigating a container). Your point about heeling +R staying close and not investigating is also a good one – since the dogs would be expected to maintain a positive emotional state (optimistic!) for staying close rather than investigating. So, perhaps a more neutral control would have been more apt, I agree. My personal point of view is that, while I agree that there is unfortunate confounding in this experiment (for example, if they were testing for the emotional effects of agility training or flyball, then the bias test would be [I suspect] different enough to not confound the training), that the behavioral benefits are still important (as I think you agree) and may also be accompanied by simultaneous emotional benefits (even if we cannot conclude that with confidence). Thanks for your thoughtful input! Best, Linda

      Like

  2. My dogs’ absolute favourite rainy day activity is a foraging game – Hunt the Treat. Three rounds, with half a dozen tiny treats hidden around the sitting room, settles them down for a couple of hours. The formal scent class we tried was less successful, I think because it went too fast. It was a one day class – and Sophy (papillon) in particular was a bit off colour that day. She prefers to set the rules for games herself, and, having indicated the position of the treat box at the far end of the hall with a flick of her head, sent me to collect it while she went to blag better treats from her friend the instructor!

    But she is brilliant at tracking anyone she knows well enough to associate their name with their scent, and never expects a solid reward – it is enough that she has done something the poor, helpless human cannot, and received genuine and heartfelt thanks. If asked she will track my sisters mislaid in parks and vast garden centres, or find the cats when they are shut in one of the 15 local garages (or tell me which bed they have been sleeping under for the past 8 hours), and she is Always Right. I found out she could do it quite accidentally one day, asking her jokingly to find my sisters in a maze-like garden – she sniffed, dropped her nose to the path, and set off with determination, and two minutes later she had found them. I am now a bit wary of doing any more classes, for fear of spoiling such a useful self taught skill!

    Liked by 1 person

Have a comment? Feel welcome to participate!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s