Clicker-plus-food and food-only are equally good dog training methods

Scientists test the use of a clicker-plus-food versus the use of food only in a positive reinforcement tricks training course for novice dogs, and find both work equally well.

Photo: Corey Terrill / Shutterstock

The study, by Lynna Feng et al (La Trobe University), used a randomized design in which dogs were assigned to one of three groups: a group that was taught with a clicker and food rewards (clicker training), a group that was taught with just food rewards, and a group that was wait-listed for the course to act as a control.

Dog trainers have a range of beliefs about the value of a clicker in dog training. Some trainers say the benefits include more efficient training, more fun, and a better human-animal bond. On the other hand, some trainers say the clicker is awkward for novice dog trainers to learn to use and that clicker training leads to more excitable and impulsive dogs.

An earlier study that compared the clicker-plus-food to use of a verbal marker-plus-food to food-only dog training found no differences in terms of the dogs’ ability to generalize to two new tasks after the training session.

The current study builds on prior research by testing pet dogs that took part in a six-week trick training course. All of the tricks were taught with positive reinforcement, using a combination of capturing, luring, and shaping.

The results show no specific benefits or disadvantages to using a clicker in dog training in terms of the dog’s impulsivity and problem-solving skills or the relationship between owner and dog. Essentially, both methods work, and people in both groups found the training fun and also challenging. There was a specific benefit to the clicker-plus-food over food-only for just one of the tricks taught, nose-targeting an object.

I asked Lynna Feng, lead author of the study, what dog owners should know about these results. She told me in an email,

“I think there are a few main points. First, if you find that the clicker training is too difficult, for you or your dog or the two of you together, and you're just looking for a well behaved pet, toss the clicker and just use food. 

Second, primarily interesting for those who teach puppy classes or general manners, even when first starting out with clicker training, the extra steps don't seem to discourage most people from training and having fun with their dogs.  

Finally, we found initial evidence that clickers helped owners feel that the training was less challenging for one of the more complex tricks.  

Although we saw little functional difference between clicker training and just using food rewards in our study training a novice population of dogs and owners, this may not hold true for individuals engaging in training at a competitive or working level.”

There were 15 dogs in each group. The trick training took place at the dog’s own home or at another location suggested by the owner where the dog would feel comfortable. The first session included an introduction to reward-based training. After that, the trainer began to teach the dog a trick, before teaching the owner how to teach it, and setting homework.

Every subsequent session began with a review of the previous week’s trick, followed by the introduction of a new trick. The tricks were nose targeting a hand, nose targeting an object, spin, chin target on the ground, play dead, and stay on a mat. The course lasted 6 weeks, which is a typical length for a dog training class.

Before and after the training, all three groups (including the control) took part in a battery of tests at the university. After the second test, the dogs who had been wait-listed were given the 6-week course so they did not miss out on learning the tricks. After that, all three groups of dogs came back to the university for a third and final set of tests.

The tests included measures of the dog-owner relationship, the dog’s impulsivity, the owner’s training technique and teamwork between human and dog. As well, questionnaires completed by owners after every training session assessed how they found the experience.

At the start of the tests, the owners completed a version of the Monash Dog Owner Relationship Scale and a dog impulsivity questionnaire, then did a reaction time test, while the dog had time to wander round and get used to the room.  Then the owner helped the dog do an obstacle course that included a stay on a platform, weaving between cones, jumping through a hoop, and walking along a plank. Finally, the dog did two problem-solving tasks that measure impulsivity.

The tests were conducted and assessed by experimenters who were blind to the condition the dog was in.

In the lab and in all training sessions, food was used as a reward. Owners were asked to have food rewards the dog would like, but the scientists were also prepared with a mix of treats including pieces of chicken, cheese, and hot dog, to ensure the dog would get a food reward they liked.

Perhaps surprisingly, there were no differences between any of the three groups on changes in the tests between the first and second visits to the lab. There were improvements in the dog’s abilities to do the tricks that were taught, and this was maintained (but not improved) six weeks after the course ended. There were no improvements in other tasks as compared to the control group (e.g. shake paw, loose leash walking), suggesting owners were not applying the skills they learned to other areas of the dog’s life.

There were no differences between the clicker-plus-food and the food-only group in terms of the relationship between owner and dog, the owner’s technical training abilities, how well the dog-owner team did on the obstacle course, or measures of the dog’s impulsivity or results of the problem-solving task.

People in both training groups rated the training as fun, and as challenging. As well, there were no differences in the amount of practice people did. In fact, about a third of the participants in the clicker-plus-food and food-only groups practiced 5 times in between sessions. This shows both groups of participants had a strong commitment to the training. However, they did not tend to do other kinds of training in between sessions.

These results confirm that reward-based training in which food is the reward is fun for the owner, and owners also perceive it as fun for their dog. For the tasks used in this study, clicker training works - and so does positive reinforcement training without a clicker.

This study adds to the literature on clicker training. Earlier research has found that positive reinforcement training is beneficial for older dogs, that even shy shelter cats can learn tricks, and that clicker training chickens increases people’s perceptions of their intelligence.

This is a well-designed and thorough study on a topic that will be of interest to many dog trainers and owners. Although the results may not apply to training at higher levels (e.g. competition), they suggest that for novice dogs (and trainers), a clicker does not have the advantages or disadvantages that are sometimes claimed. Essentially, whether you use it or not is your choice.

Further research can investigate the use of a clicker versus verbal marker for those kinds of training tasks where a marker is expected to be particularly beneficial.

You can follow the first author of the research, Lynna Feng, on Twitter.

Do you like to use a clicker in dog training?


Reference
Feng, L. C., Hodgens, N. H., Woodhead, J. K., Howell, T. J., & Bennett, P. C. (2018). Is clicker training (Clicker+ food) better than food-only training for novice companion dogs and their owners? Applied Animal Behaviour Science.

Companion Animal Psychology is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com and affiliated sites.
Companion Animal Psychology is also a participant in the Etsy Affiliate Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Etsy.com.

Why Don't More People Use Positive Reinforcement to Train Dogs?

Everyone who has a dog needs to teach them how to behave. But why do many people still use methods that have risks for animal welfare?

Photo: Corey Terrill/Shutterstock

A new paper by myself (Zazie Todd) looks at the barriers to the adoption of humane dog training methods by ordinary people. To understand how people make decisions about dog training, we need to understand people’s attitudes to different methods and what influences them, as well as people’s knowledge and technical ability in using those methods.

Humane dog training methods use positive reinforcement and negative punishment. They are also known as reward-based methods, positive reinforcement, or force free methods, and they basically involve giving or withholding rewards contingent on the dog’s behaviour.

There are many reasons to use humane methods rather than aversive ones (which, technically speaking, are positive punishment and negative reinforcement). The use of reward-based dog training methods is associated with better welfare, and there are some indications it may even produce better results (see the dog training research resources page if you want to delve into the literature).

Some behaviour problems are due to fear or stress, but aversive methods do not resolve this (and may even make it worse). Some problems occur because the dog does not have appropriate ways to engage in normal behaviour (e.g. chew toys). And training with positive reinforcement can be a good way to provide cognitive enrichment, which is important because good animal welfare includes positive experiences.

But studies show most ordinary dog owners use a mix of positive reinforcement and positive punishment to train dogs – so-called ‘balanced’ dog training. From an animal welfare perspective, it’s important to understand why many people continue to use aversive methods at least some of the time – and how we can bring about change.

Many different factors will affect people’s attitudes towards training methods and the actual methods they use. The paper considers these factors, and in some cases it draws on the literature related to children and parents’ use of corporal punishment, which has been more extensively studied.

Many people (including some dog trainers) still use the idea of dominance to train dogs. Unfortunately, this frames the dog-owner relationship in antagonistic terms, and so may encourage people to use aversive methods.

There is no regulation of dog trainers, and no requirement for education. This means some people who hire a dog trainer may get out-dated advice. Some trainers may not be clear about the methods they use on their website, which may make it difficult for people to find a reward-based trainer.

Amongst dog trainers who do use reward-based methods, there are a few points of disagreement. One relates to the use of no-reward markers. This is a word or phrase (e.g. “Too bad!” or “Oops”) that signals to the dog the behaviour they did was not the one requested, and hence they didn’t earn a reward. Some dog trainers use them, and some don’t. For most people, errorful learning with prompt feedback is more successful than error-free learning (the exceptions include those with amnesia). But we simply don’t have good data on this for dogs.

Another point of disagreement relates to the use of negative punishment. Negative punishment means withholding a reward so the frequency of a behaviour goes down. One example is withholding rewards when a dog does not perform the right behaviour (which is inevitable some of the time). Another example is the use of ‘time out’. Evidence-based parenting programs teach ‘time out’ as a non-aversive way to improve children’s behaviour. We know there are some common mistakes parents make, and it seems likely dog owners make some of the same mistakes with their dog (for example, using many warning cues instead of just one before implementing the time out).

Photo: Anucha Pongpatimeth/Shutterstock

The legality or otherwise of certain methods (such as electronic shock collars, which are banned in a number of countries) will affect perceptions of whether or not it is okay to use them.

The different positions taken by professional bodies, including veterinary associations, dog training associations, animal behaviour associations, groups that train working dogs, rescues and shelters, may also affect people’s perceptions of social norms about dog training. Some organizations have clear position statements on dog training methods. But when some organizations include aversive methods as a ‘last resort’ it may give people the mistaken impression they are sometimes necessary, or that there is not a scientific consensus on the best methods to use.

There’s also an issue of competency, since technical expertise etc. may affect the success of attempts to train with positive reinforcement. We don’t know how dog trainers or owners make decisions when they think positive reinforcement isn’t working; that is, whether they take advice or find someone with more expertise, or whether they decide to use positive punishment instead.

Veterinarians also have an important role to play in referring dog owners to trainers and animal behaviourists. Advice for veterinarians stresses the importance of positive methods, but again some organizations allow for some methods to be used as a ‘last resort’.

Of course, many factors relating to dog owners themselves will also influence their choice of training methods. These include their technical skills and the reinforcement they use (which will affect their success rate), their knowledge of dog training methods, methods they have seen promoted on TV and elsewhere, people’s ability to read their dog’s body language (e.g. to recognize if the dog is fearful), and personality characteristics.

Unfortunately the quality of information in dog training books is highly variable and the same likely applies to other sources of information such as TV and the internet.

All of this shows that encouraging more people to use humane dog-training methods is a complex issue. But a model from social psychology known as the reasoned action approach (and its predecessor the theory of planned behaviour) has been quite successful in predicting people’s intentions and behaviours in a wide range of topics – including parents’ attitudes to and use of corporal punishment. This would be a good fit for investigating what influences people’s choice of dog training methods.

I would love to see more research on the best ways to encourage people to use reward-based training methods, and how best to teach them. As an update, this 2019 study shows the importance of dog trainers building people's confidence in using positive reinforcement.

What do you think would encourage more people to use humane dog training methods?

Love Companion Animal Psychology? Did you know you can support me on ko-fi? Ko-fi is like a tip jar that lets you buy a coffee for creators whose work you love.


Reference
Todd, Z. (2018). Barriers to the Adoption of Humane Dog Training Methods. Journal of Veterinary Behavior: Clinical Applications and Research.  25C(28-34).

As an Amazon Associate I earn from qualifying purchases. As an Etsy affiliate, I earn from qualifying Etsy purchases.

The Guinea Pigs' Perspective and Humane Dog Training

A couple of news items: new post at Psychology Today, new article in press at the Journal of Veterinary Behavior.


I have a new blog post at Psychology Today called Animal-Assisted Therapy: The Guinea Pigs' Perspective.

It's about a new study by Gut et al that looked at the behavioural response of guinea pigs in animal-assisted therapy sessions with and without the option of retreat. There are implications not just for animal-assisted therapy but also for anyone with pet guinea pigs.

Photo: 12071/Pixabay

In case you missed it, in February I had a post on Choosing Dogs that can Breathe.

I also have a paper in press at the Journal of Veterinary Behavior: Barriers to the adoption of humane dog training methods. Temporary free access is available at this link.

Photo: Ksenia Raykova/Shutterstock

To gesture or not to gesture in dog training?

Are visual cues more effective than verbal cues in dog training?

Guest post by Sienna Taylor, MSc (Hartpury University Centre).

Photo: Dorottya Mathe (Shutterstock)

A new study by Anna Scandurra (University of Naples) et al. investigates whether visual cues as opposed to verbal cues are more effective when dogs are trained to fetch an item under four conditions: using only hand cues, using only verbal cues, using both hand and verbal cues and using contradictory hand and verbal cues.

It turns out that dogs responded better to visual hand gestures than verbal cues although speed of response was quicker when both hand and verbal cues were used together.

Many pet owners teach their dogs to respond to both visual and verbal cues, for example, an owner might ask their dog to lie on the floor by simply using the verbal command “Lie Down” or alternatively using a hand gesture such as pointing or perhaps a combination of both!  Whilst dogs do use vocalisations to communicate (such as attracting attention, with vocalisations usually being context specific) (Serpell, 2017), they communicate largely through the use of discrete body postures (Landsberg et al. 2013), both intra-specifically (dog-dog) and inter-specifically (human-dog).

Dogs are adept at responding to our gaze or if we nod (Kaminski and Nitzschner, 2013) or point towards a particular object (Lakatos et al. 2012).  Sometimes we find that when we ask a dog to verbally “Lie Down” the response is a blank look or an altogether different response!  Yet if we use a hand gesture such as point to the floor or use a combination of both verbal and visual cues the dog instantly lies down.  This more immediate response to a visual cue, even when in combination with a verbal cue, has often puzzled owners and begs the question are visual cues more effective in dog training than verbal cues or should we be using both?

The study by Scandurra et al. (2018) set out to test whether 13 pet dogs responded better to their owners using either visual or verbal cues alone (unimodal) or both visual and verbal cues (bimodal) which took into account both the dogs acute visual and auditory capabilities.

Dogs were trained to ensure they responded equally well to both verbal cues and visual hand gestures and were asked to perform a pre-test fetch task. Objects included a piece of wood, a plastic bottle and a pencil case.

Twenty four trials took place in the pre-test phase, eight trials used verbal cues only (spoken in Italian, with the voice command directing the dog to retrieve one of two items), followed by eight trials using hand gestures only (where the owner directed the dog by pointing to one of two items). A further eight trials used both verbal and visual cues (the owner directed the dog to one of the two items through the use of both verbal and visual cues at the same time).

Nine dogs met the requirements of the pre-test phase and were selected to take part in the final eight trials where a combination of both cues were used. However, this time the cues contradicted one another, for example when asking the dog to retrieve, the owner pointed at one object but named another.

Dogs were found to respond equally well to both verbal and visual cues when used on their own although, when both verbal and visual cues were given together, dogs were found to respond significantly more quickly to the task.  When dogs were given contradictory information, 78% of dogs (7 out of 9 dogs) chose the hand gesture. The remaining two dogs performed at a chance level and randomly chose to retrieve the verbally indicated or the object visually pointed at equally often.  What’s interesting is that none of the dogs preferred the verbal cue over the hand gesture.  This leads us to further question the importance of verbal cues to dogs.

How we use verbal communication (e.g. quality of spoken word) and also level of eye contact has been found to impact level of responsiveness in the dog. Fukuzawa et al. (2005) found that when a dog was asked to sit with the command played through a tape recorder, there was a significant decline in performance. It also took the dog longer to learn the command in the absence of lip or facial movements.  Similarly, when the person obscured their eyes by wearing sunglasses and the command was played through the tape, the dog’s responsiveness to the command also reduced.  However, when sunglasses were worn and a spoken verbal command was given no reduction in responsiveness was evident.  The authors concluded that eye contact is important to dogs but not in all contexts.  Fukuzawa et al. (2005) also found that effectiveness of command was reduced when a person’s back was turned.  This implies that body postures appear to be important to the dog in understanding signals as part of human-dog communication but may be context dependent.

Next time you use a cue, if the dog does not respond it is worth following up with a hand gesture to see if you get a better response!

About Sienna Taylor:

Sienna Taylor FdSc, BSc (Hons), MSc, FHEA, is a Lecturer in Animal Behaviour and Welfare at Hartpury University Centre, Gloucestershire.  Her research interests include human-animal interactions and the use of olfactory enrichment in companion animals.  Sienna enjoys training her two year old Labrador Bailey and they are currently working towards their Grade 3 Gundog Test.

You can follow Sienna Taylor on Twitter: @Taylor5Sienna.





References
Fukuzawa, M., Mills, D.S. and Cooper, J.J. (2005) More than just a word: non-semantic command variables affect obedience in the domestic dog (Canis familiaris). Applied Animal Behaviour Science. 91(1), pp.129-141.
Kaminski, J. and Nitzschner, M. (2013) Do dogs get the point? A review of dog-human communication ability. Learning and Motivation. 44, pp. 294-302. https://doi.org/10.1016/j.lmot.2013.05.001
Lakatos, G., Gácsi, M., Topál, J. and Miklósi, Á. (2012) Comprehension and utilization of pointing gestures and gazing in dog-human communication in relatively complex situations. Animal Cognition. 15, pp. 201-213.
Landsberg, G.M., Hunthausen, W.L. and Ackerman, L.J. (2012) Behavior Problems of the Dog and Cat, 3e.Oxford: Elsevier Health Sciences.
Scandurra, A., Alterisio, A., Aria, M., Vernese, R. and D’Aniello, B. (2018) Should I fetch one or the other? A study on dogs on the object choice in the bimodal contrasting paradigm. Animal Cognition. pp. 1-8.
Serpell, J. (2016) The Domestic Dog: Its Evolution, Behavior and Interactions with People. Cambridge: Cambridge University Press.

Companion Animal Psychology is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com and affiliated sites. Companion Animal Psychology is also a participant in the Etsy Affiliate Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Etsy.com.