© Can Stock Photo / feelphotoart
By Diane Garrod
In some circles, using treats in training has gotten something of a bad rap. But I have to ask myself, why is there even hesitation about this? Using food in positive reinforcement for results IS science. It is a proven, highly productive technique that involves pairing good things with triggers, or reinforcing behavior so it is repeated, or giving a “paycheck” for a job well done, a skill learned.
A reinforcer is something a dog considers it is worth working for to get. Actually knowing and finding out what an individual dog finds reinforcing takes time and experimentation. Dog guardians may assume that what they already know their dog likes, or whatever is on hand, will be reinforcing enough to change behavior. But this may not be the case if the dog doesn’t consider the food to be of sufficiently high value.
The American Veterinary Society for Animal Behavior (AVSAB) (n.d.) states that it “endorses training methods which allow animals to work for things (e.g., food, play, affection) that motivate them rather than techniques that focus on using fear or pain to punish them for undesirable behaviors.” AVSAB advises dog owners to “[l]ook for a trainer who uses primarily or only reward-based training with treats, toys, and play. Avoid any trainer who advocates methods of physical force that can harm your pet such as hanging dogs by their collars or hitting them with their hands, feet, or leashes.” (AVSAB, n.d.).
In my experience, however, common concerns and criticisms professional dog trainers may hear from their clients about using treats in training include statements such as: “My dog only listens to me when I have food.” (This is one I hear the most); “I don’t want to carry treats forever;” “I think my dog should obey me without food;” “I don’t want my dog to get fat so I don’t use treats to train her;” or “My dog doesn’t like treats.”
Eating Is Calming
The ingesting of food is a natural process for dogs and other animals, vital to survival and fitness. And the process of eating creates saliva, which is calming to the dog. Saliva lubricates a dog’s mouth, can help prevent tooth decay and gum disease, and starts the process of breaking down the food eaten for digestion. Trainers and owners should be cognizant of the fact that the stress of learning, or being in an environment a dog doesn’t feel safe in can lead to indigestion. Or the dog may refuse food completely because they are too stressed to eat. In addition, some dogs are allergic to certain types of food, so attention to those types of details is important.
According to White et al. (2016): “Complex relationships commonly exist between owners and their companion animals, particularly around feeding behavior with an owner’s affection or love for their animal most pronounced through the provision of food. The results of the study show that treat giving is commonplace in feeding regimes and that treats are embedded in the feeding behavior of many dog owners.”
The bottom line is that dogs eat. They are motivated by food. It is natural, it is reinforcing, and it changes behavior. Pairing good food with what a dog perceives as their trigger, changes the way they see that trigger (i.e. counterconditioning).
Domestication of Dogs
Dogs may have become domesticated because our ancestors had more meat than they could eat. During the ice age, hunter-gatherers may have shared any surplus with wolves, with the friendlier, less fearful animals staying closer to human settlements. One theory suggests that humans domesticated dogs to help them with hunting, while another scenario has wolves scavenging human waste dumps and becoming accustomed to being people.
Lahtinen et al. (2021) suggest that the key may have been a surfeit of meat with hunter-gatherers possibly taking in orphaned wolf pups – perhaps viewing them a bit like pets – and fed them on spare lean meat. While they may not have had any long-term goal in mind, the tamed wolves would have later proved to be useful hunting partners as well as useful watchdogs, thereby reinforcing their continued domestication. “They must have been very attractive for hunter-gatherers to keep,” state Lahtinen et al. (2021).
Dogs are scavengers and spend a good portion of their lives seeking food. They counter surf, take an interest in garbage, forage around outdoors, and are capable of chasing and killing prey. Food has always been an incentive, a motivation, and a survival tool with a happy ending. It makes complete sense that food should be the tool of choice in teaching; it should be welcomed, not made out to be the bad guy.
Of course, there are reasons for the above concerns of well-meaning dog guardians. They may be passing on information they’ve heard from family, friends, or well-meaning strangers, none of whom are dog training professionals. Advice should, however, be taken by those who are educated in science, who are skilled in implementing the training techniques, and who possess the knowledge of how and why these methods work best, and not by those who have no credentials.
Using Food in Dog Training
Let’s examine some of the concerns a little more closely, starting with, “My dog only listens to me when I have food.”
This almost always has to do with technique. Perhaps the guardian misunderstood the training technique, or the trainer wasn’t specific enough about how to disburse rewards, body position, or hand movement.
Dogs are masters at pulling together motion, objects, and how they all work together. If the cue you give includes rustling in your bait bag or standing by a jar or bowl filled with treats, then the dog learns to pay attention because the consequence will be that they get food. This can encourage the dog to beg for food.
In my experience, dog owners often hold a treat in their hand and then move it up to their chest, causing the dog to focus on, you guessed it, the food.
Another common occurrence is that people give the cue while they rustle their hands in a treat pouch or bait bag. So the rustling, the body position, and the raising of the hand with the treat in it all become part of the cue.
As a result, it may seem as though the “dog only listens to me when I have food,” even though that is not actually the case. To make matters worse, when a cue, for example, “sit,” is given or repeated, the dog doesn’t do the cue until the other cue is added, i.e. reaching in the pocket, treat pouch, or bait bag for the food, which has now become a lure.
The response to the cue will now only be given with the completion of the cue chain rather than for the joy of performing the skill and then being reinforced for doing so.
Mechanics of Training
In such instances, the learning was faulty, not the food. The importance of clear communication refers not just to the dog, but to all learners.
When teaching a skill, the hands should not digging be around or making a noise where the food disburses, and the body should be still. The trainer’s hands should be at their side, holding five to 10 treats in a fist. They give the cue, then pause and wait.
When the cued behavior is completed, the trainer gives a marker (clicker, or marker word like ‘Good!’ or ‘Yes!’) for the action of completing the cue, and the reward is given – without digging through pockets, bags from the store, bowls, a treat pouch, or a bait bag.
If the trainer does have their hand in the bait bag, it shouldn’t move until the cue is completed as asked. By making the cue completion, the secondary reinforcer that releases the primary reinforcer and the implementation of the cue becomes stronger.
Removing any unnecessary motion and body positions eliminates potential issues with begging. If a dog only pays attention when food is around, it means they have learned to beg and have been reinforced for the behavior by hand movement, bag noise, and/or body position.
Earning a Paycheck
Another criticism trainers may hear from their clients is, “I don’t want to carry treats forever.” This often has a lot to do with peer pressure, comments from others, and a client who might be embarrassed by the fact they have to use treats. But it has nothing to do with the science, or the reasons food is motivating and behavior changing.
In fact, the disbursement of food for a job well done carries similarities to our own lives, e.g. a paycheck (so we can eat), or a good meal as nourishment and a reward. Stopping a paycheck wouldn’t be an option, nor would not receiving food for a job well done or as a celebration.
Indeed, I like to look at rewarding for desired behavior as a celebration. A celebration of what the dog is doing right – instead of the behavior we don’t want – is well worth rewarding. As doing a behavior becomes stronger, then the disbursement of treats varies and/or adds different rewards as a paycheck.
High-Value Rewards
What about clients who say, “My dog doesn’t like treats?” This criticism often revolves around the owner not taking time to find the reinforcers that the dog loves, but instead using treats they personally like, or think the dog likes.
It’s a fact of life that dogs eat, so using rewards as part of their food allocation can help, as can varying the reward for renewed interest.
Some dogs have very clear tastes, and some dogs are just too stressed to take a food reward (e.g. fear because of another dog or person coming into view, sounds, motions, or even when focused on prey). In these cases, keeping stress levels normalized is key and again, proper use of technique in highly distracting situations.
Stress Affects Learning
Another common criticism, “I think my dog should obey without food” shows a basic lack of understanding of how to change behavior – and possibly also a reason to use aversive techniques. It is an opinion, and is not based on fact or results.
Food rewards make trainers and owners worth listening to; dogs learn that all good things come from us and so, eventually, will work for us willingly and happily – even without food.
Stress at acute and chronic levels affects how dogs learn. Often, they can’t learn when they are fearful, anxious, hyper-focused on something else, or overstimulated by something else in the environment. Because they are in a situation they cannot handle, it can look like they don’t like treats, even though this is not the case.
The next criticism, “I don’t want my dog to get fat, so I don’t use treats to train her,” has a point in that overfeeding can cause a dog to become overweight, which affects overall health. Dogs are overweight at a high rate in the U.S. According to Williams and Downing (n.d.): “Approximately 25-30% of the general canine population is obese, with 40-45% of dogs aged 5-11 years old weighing in higher than normal.”
But training with food is not necessarily the reason. Owners may give large commercial rewards that are too big, or they overfeed or free feed, they do not understand nutrition and diets appropriate for individual dogs, or perhaps do not provide the physical exercise required to work muscles and bones and keep fat levels manageable.
Not using treats as reinforcement will not take away the fact that a dog becomes overweight.
Better Learning with Food Rewards
The benefits of using food rewards are clear; studies (e.g. Blackwell et al. (2008); Hiby et al. (2004); McGowan et al. (2014); Rooney and Cowan (2011) – see Resources) show repeatedly that dogs learn better with rewards, dogs learn a new task better, faster, using rewards.
Food rewards should be sized properly and incrementally disbursed, ideally 10 trials at a time, no more than 30 trials per session. Owners can adjust the dog’s meals to the number of rewards disbursed, or use meals as a teaching tool.
The types of treats should be varied as dogs may get bored with the same rewards over and over, and this can also lead to a lack of food motivation. It is important to identify what the dog really loves and is willing to work for.
Reinforcement can progress from continuous (in early and new learning) to varying when reinforcement is offered, which strengthens the behavior we want, as learning advances. A reward might be given every other time, or third time, or fifth time, i.e. varying the “when.”
Behaviors that are strong in performance no longer need continuous reinforcement and can be increased in duration before the reward is given. Alternatively, we can introduce a behavior chain before the reward is given, and/or use different types of reinforcement, such as praise, petting, functional rewards (sniffing for example), toys, or games (catch, tug etc.).
Generalizing Behavior
Dogs do not always generalize well, so each new environment means going a little bit back to whatever level the dog can handle in terms of performing a specific behavior.
For example, we can go back to continuous reinforcement and then advance from there. Practicing behaviors in varying environments is important to be sure that training is generalized to other areas and contexts. We need to watch stress levels closely by reading the dog’s body language and making sure reinforcers are available in varying types to address varying levels of stimuli/triggers/distractions. If the dog is too stressed, we just try again another day.
In summary, food rewards are used because they work!
Food is a meaningful, natural motivator. What is high value varies with each dog, but some examples are liver, roast beef, warm hot dogs, salmon, lamb lung, chicken, mozzarella sticks, whipped cream, or other meat.
Never hesitate, or scrimp, in using food as a reward, but embrace it as reinforcement system to strengthen desired behavior.
References
American Veterinary Society for Animal Behavior. (n.d.). How to Choose a Trainer
Lahtinen, M., Clinnick, D., Mannermaa, K., Salonen, J.S., & Viranta, S. (2021). Excess protein enabled dog domestication during severe Ice Age winters. Scientific Reports
White, G.A., Ward, L., Pink, C., Craigon, J., & Millar, K. M. (2016). Who’s been a good dog?” – Owner perceptions and motivations for treat giving. Preventive Veterinary Medicine 132:14-19
Williams, K., & Downing, R. (n.d.). Obesity in Dogs