Ask the Trainer: Follow-up: “Should I use treats in my dog training?”

Should you use treats to train your dog? We received so many questions regarding a previous post (“Should I use treats in obedience training?” – June 26, 2012), it was clear that “using treats in your dog training” is an important topic. We therefore decided to give the question of treats-in-training some additional attention in our series of posts about how to train your dog.

While we expected questions and comments (even disagreement!) from pet-owners, trainers themselves strongly endorsed the principles described. Interestingly, almost all the questions also came from dog trainers. As a result, we selected one question from a trainer who best exemplified the comments from others who wrote.

Using Treats in Your Dog Training:  Question/Comment
Hi — I was appalled to read the story of the elephant [see Blog April 10, 2012]. I’m a trainer and this story made me think about my own training (mostly Rottweilers and other large dogs). I’ve tried using only treats to get or change behaviors but I need some help. People tell me that when they bring the dog home, he or she reverts to the same problem behavior. Any advice you could give would be appreciated.
Harold B. – 3 weeks ago

From Pat Brown-John:
Pat Brown-John is a five-time Canadian Schutzhund champion, with national and international titles spanning a training career of more than 30 years.  Here is her response:

Dear Harold,

We as trainers have a responsibility to the owners of the dog we take in for training.  They fully expect us to be able to train their beloved dogs to be well-mannered members of the family they can be proud of at all times.

We do the job of dog training because we love to be around dogs, especially well behaved ones and we have a gift of being able to communicate to the dog what is expected of him in his day-to- day life.

When a dog has been given full rein at home to do whatever he pleases, until no one can take it any more, food is not going to change anything.  When a puppy misbehaves, the mother dog doesn’t run and find him a piece of food, rather she gives a quick correction and they move on.

So is it with our training.  First we show the dog what we want of him.  He complies and is given lots of praise (“good boy” and a ruffle of the fur, maybe we even jump back and play) then the command again.  Work sessions should be short which will leave the dog wanting more.

After the dog shows he understands the command you are asking, bring in a distraction, perhaps a person walking past, or repeat the exercise in a different area.  If, this time, the dog refuses the command, give a correction that gets his attention but does not put him into a fight or flight mode. Only use as much correction as you need for that particular dog.  When he complies, break and play or give quiet praise, whatever the situation requires.

I do use food in some situations.  When a dog is in the “down,” I’ll leave the room. When I come back, I’ll give pets and quiet praise or I’ll drop a piece of food.  The dog will not see my hand bringing the food because I don’t want the dog expecting that every hand has food in it.  A lot of my families have children so a dog wanting to take food out of their hands won’t work.

A basic truism: If the dog truly doesn’t understand that he must do the exercises asked of him, then when he goes home he will not do any of the exercises asked of him.

When the dog goes home he will try to take back what he had before he left.  It’s our job as trainers to work with the family and the dog together to ensure they are all on the same page.  When the dog knows that the family knows the same rules he has been taught, he will follow those rules.  I take time with my families to ensure they know how to make a proper correction– not drag the dog into position.  A week later, I work with the family again to make sure that all the training is working for the family.  After that I am just a phone call away.

If you are getting referrals from your families, you’re on the right track.

Gook luck Harold and good training.

Pat Brown-John
All About Obedience

For an in-depth perspective on this subject from another professional trainer, click on: Balanced Training.

We'd love to get your comments and feedback on this topic and post. To add your comments, scroll to the bottom of the page and "Leave a Reply" (type your thoughts into the text box).

, ,

5 Responses to Ask the Trainer: Follow-up: “Should I use treats in my dog training?”

  1. Norman W Wilson, PhD July 25, 2012 at 4:41 pm #

    Do we give children a “treat” every time they do as we ask? I think not. Then, why give an animal a treat because it behaves well, does what we ask? Isn’t the best treat in the world love and affection?

    • Laurie July 25, 2012 at 8:45 pm #

      This is a good point.
      I think the answer to this would be, “it depends on the dog (or the person)”. My nephew will do anything for a hug. My son preferred a reward of Legos. My dog Angie wanted food. My current dog want an ear scratch. I think the receiver, not the giver, decides if something is a reward.

  2. Laurie July 25, 2012 at 6:22 pm #

    Great post.
    It pains me when I see clients who think that because behavior X has been trained, the reinforcements can stop.
    I trained my dog not to take food from the table when I first got her as a rescue. In the past few months I’ve been lazy about leaving things down too long and allowing her to get it. Bad trainer, not bad dog. I changed my behavior, praised her for ignoring food on the table, stopped being an idiot and leaving food down for hours when I wasn’t watching and suddenly I have a good dog again (har-har)
    Years ago I had a Service Dog and went to a demo. I had left my dog Angie in a down while I went to do something. When I came back after 45 seconds there was a woman who was sprinkling liver flavored biscotti all over my dog and announcing to people near-by “Service dogs never eat food off the floor” Angie held her down and scooted to get every morsel. I didn’t correct my dog, but I did correct the woman.
    Best advice I read in your post ” Only use as much correction as you need for that particular dog.” As a trainer I like to use all my tools. Food? Sure. Corrections? When appropriate. Avoid a bad situation if my dog isn’t ready for it? You bet. Be a good trainer and yo u’ll have a good dog.

  3. Dianna Young July 26, 2012 at 3:20 pm #

    Finding Your Dog’s Currency
    How do you motivate a dog to learn? You “pay” him with something he wants. That thing that he wants is his “currency.”
    A tidbit of food is a type of currency that many dogs recognize and accept, and a lot of trainers use a bit of cheese or a piece of dried liver to motivate their pupils. But many other kinds of currencies are available as well. For some dogs, an effective currency might be an opportunity to play with a tennis ball or a tug toy, or a chance to go for a brief swim. For others, an effective currency might be petting and praise. Sometimes, a dog’s reward can be the simple pleasure he gets out of performing his skill. Consider, for example, a member of a retrieving breed who just loves to bring back a thrown object or a downed bird, or the assertive member of a breed noted for police work who gets to bite the padded arm of a person during protection training.
    While many dogs recognize and accept food as a currency, we believe that a trainer often can find better motivators. Yes, you can teach a dog to sit by bribing it with cheese. The method does work. But keep this in mind: The quality of the effort you get from a dog often depends on the type of motivator that drives him to perform the work. You get better performance from an animal that desires to please – and to receive approval from – its leader. These desires spring directly from a canine’s instinctive need to belong to a pack. While you certainly can get some results by bribing a dog with cheese, that technique does not connect with any of his inborn psychological needs.
    How can you best determine the optimum method to use with your dog? You can begin by considering the natural tendencies of the particular type of dog with which you are dealing. Is your dog a type of hound? Is it an aggressive type of canine that might be used for personal protection? Is it a herding dog? A retriever? If it’s a mixed breed, what do you think might be the combination of heritages from which it springs?
    Dogs are highly individual creatures, just like people, so it is foolish to paint with too broad a brush. We often find exceptions to the general rule. Still, an assessment of your dog’s breed (or its primary breeds in the case of a mixed-breed animal) is a good place to start in trying to determine effective currency, and you need to be willing to be flexible in your approach. While I’m not a devoted fan of food incentives, for example, some types of dogs – such as hounds — sometimes do comparatively well with them, so food might be a good place to start in trying to find a currency that will work for your beagle.
    A hound, by the way, is a dog – usually with long, floppy ears – that makes its living tracking other animals by means of a scent trail that animal has laid down on the ground (as opposed to scent that is pulled from the air). Hounds have many tens of thousands more scent receptors in their heads than do most other types of canines, and the most important thing in a hound’s world tends to be its olfactory system. Hounds tend to be less concerned than others about obedience tasks and more concerned about what they smell or taste. This does not make them “bad” dogs. They simply are bred to operate with different priorities than, let us say, herding dogs.
    Canines with a tendency toward high prey drive (the desire to chase things that move), such as a Doberman Pinscher or a Rottweiler, for example, often will work hard for an opportunity to chase a tennis ball.
    Working and herding dogs, such as Australian Herding Dogs or German Shepherd Dogs, tend to have a tremendous willingness to please their owner or handler. A show of approval from that person often is a very adequate and effective currency.
    Your Labrador Retriever might work for the same kind of reward, or he might work even better for an opportunity to chase a stick thrown into the water.
    How then do these things affect us as owners and trainers? These factors mean we should be realistic about our goals for a particular dog, whether we acquired it to be a working animal or a family pet. Some people join groups that practice obedience with their dogs competitively, for example. Others join similar groups that compete in canine agility. Still others become part of organizations that compete in their dogs’ skills at scent detection and discrimination. Depending on what kind of dog you have, and the selective breeding that went into developing it, obedience might not be the best competitive endeavor for your pet. Your basset hound might not excel there, for example, or in the agility ring, either. But he is built really well for scent detection and discrimination, and that might be a fascinating activity for you and your basset to get involved in together.
    The point is to be realistic about your goals. Work with your canine and not against it. Play to its strengths. Because the average basset hound, for example, does not excel in the obedience ring does not mean that it cannot excel in some other activity for which it has been properly wired through hundreds of generations of selective breeding.
    Having considered all of these factors, then, how does a trainer determine the best currency to use with a particular dog? He does it by observation, followed by trial and error. Does your dog trip over his own feet to get to you to (1) get to that cheese that you have in your pocket or to (2) get to that tennis ball you have in your hand or to (3) just get to be with you and bask in your presence? You determine this by observation and interaction. And you create simple tests involving all of the possible currencies, and then observe your dog’s reactions.
    Having determined what currency to use, determining how best to use it also is a process of trial and error. At the beginning of training, we want to make an association for the dog between the behavior we desire and the reward we intend to bestow. So, during this phase we will make our reward more readily available so he can see the connection between it and the behavior we desire.
    Once the initial connection has been established, however, a trainer must determine how much of the currency is appropriate, how much is too much, and how much is not enough. Obviously, we must use enough of the currency to give our dog the satisfaction he is trying to earn with his performance. However, the more one receives of something the less one tends to value it, and if a trainer is too liberal with a dog’s currency, the currency becomes less effective. After the connection is made, you probably will not want to use the currency more often than once after every third to fifth successful repetition.
    How you disperse your currency must be individualized for the particular dog. I could pull 10 dogs out of our kennel from among those who are here for training, and would find a different amount of currency optimum in each case. Let’s say, for example, that each of the 10 is motivated by praise. I probably would give each of them different amounts of praise, a different duration of praise and a different type of praise, all depending on a particular dog’s personality. With one, I might want to scratch him behind the ear or pat his side while I tell him what a wonderful job he has done. With another dog, I might just have to say, “Ah, good job!” with no physical praise attached.
    Dogs are as individual as people. Some are very touchy-feely, others less so. Let’s say I’m working with a happy-go-lucky Golden Retriever which is very affectionate. Maybe he is a dog that appreciates a lot of physical praising and touching. Another dog might prefer just a simple tickle behind his ear, or only a verbal “Attaboy.”
    The amount of currency you pay, whether it’s verbal or physical, must be individualized to the particular dog. You determine whether you’re doing it correctly by the results you are getting. As a trainer, you have to be flexible. It’s like driving a car. As you drive down the road, sometimes you need to apply a little more gas, and then a little less. When you start down a hill, you might take your foot off the accelerator and coast for a while. If you start to pick up too much speed you might even touch the brake. Working with a dog is much the same. At times you may want to provide a little more reward, at other times a little less. You have to pay attention, read your dog, and read the results to determine what is correct.
    When your dog is performing your demands on cue, without any lag, and is showing a desire to perform for the reward, you know you are doing it right.
    Whatever type of reward you administer, its timing is of utmost importance. Regardless of the type of currency, you have approximately three critical seconds within which to mark the canine behavior you want to reward (or to discourage). About three seconds, maximum, is all the opportunity you have to bring the dog’s attention to the behavior. Taking less time is better. Anything in excess of three seconds, and the window of opportunity is closing.
    Does this mean that you interrupt training in order to reward the dog for a job well done? You interrupt training, yes, but you do not interrupt a particular task. Let us say, for example, that your dog is learning to sit on command, and that your dog’s currency is an opportunity to retrieve a thrown tennis ball. You should toss the ball for him upon his completion of several successful repetitions of carrying out the command, and within the critical three seconds of having completed the final repetition.
    Early in his training on a new task, you will reward him after every successful repetition. As he begins to excel at the task, you will request more repetitions from him before offering the reward. This goes back to a currency’s value. If you offer too much of it, the currency loses value.
    Every professional trainer has his or her preferred currency, which he will use as his first choice if it happens to be effective with a particular dog. Different trainers favor different methods, and most of the methods work to one degree or another. At Camano Island Kennels, the method we prefer is to develop a dog’s natural willingness to want to please its pack leader. A bribery-type incentive, such as food, would be our last choice. Bribery is not a natural way that dogs choose to motivate each other. Have you ever seen a mother dog teach her puppies things through acts of bribery? We believe developing a dog’s willingness to please results in a stronger, more reliable working attitude because it taps into the dog’s instinctive need to belong to a pack, either canine or human. We believe it develops a better, more reliable work ethic. Most canines are inclined by instinct to perform tasks that their leader, canine or human, sets out for them.
    Another method worthy of mention, because it is used by some trainers, is force. However, we do not favor it. Certainly, we could browbeat a dog into working for us, but it probably would not inspire the attitude in the dog that we think results in the best performance. We prefer that our animals enjoy the work they do, and believe that this results in their doing a better job.
    While we consider our approval the most desirable and effective currency to use most of the time, it may not work in every case. No single currency does. In that event, a trainer must be flexible and resourceful. If you are working, for example, with a very self-assured and independent canine that doesn’t particularly care about pleasing, you will have to find a reward that is more effective with that individual. Perhaps it will be the thrown tennis ball or the brief swim. Perhaps, in a pinch, it even may be food. Remember, however, that no single currency works 100 percent of the time with 100 percent of dogs. Not even food.
    We keep in mind that most dogs need affection more than they need anything else – certainly more than they need a bit of cheese – and that nearly all dogs long for acceptance by the pack. Because they have a need for acceptance and approval, it usually turns out to be one of their driving forces in wanting to perform for us.

    Dianna Young is a certified, professional dog trainer and canine behaviorist who operates Camano Island Kennels Dog Boarding, Grooming and Training Facility and Stella Ruffington’s Doggy Playcare in West Seattle. She can be reached at (360) 387-DOGS or at (206)932-RUFF

  4. Dr. Patricia Bloom July 26, 2012 at 10:01 pm #

    Thanks to all who commented on this important post. Laurie — I loved the vision of Angie scooting along the floor in a “down” position gobbling up all the goodies. A dog after my own heart!

    A special thanks to Dianna Young for your thoughtful and insightful remarks regarding the subject of giving treats in training. I particularly appreciated your comment: “Have you ever seen a mother dog teach her puppies things through acts of bribery?” That really summed up the whole subject for me!

    Readers: While the remarks from Dianna Young may be difficult to read (that darn tiny print, which I can do nothing about) it is really worth your time to do a cut and paste, enlarge the font and read everything she says. It’s instructive and exceedingly valuable.

    Dianna Young is one of our top trainers in the west coast and, as you can tell from reading her remarks, has an enviable reputation among dog trainers, which is justly deserved.

Leave a Reply