I recently found myself watching an "old-style" dog trainer working with a beginner's dog obedience class. I use the phrase "old-style" because this instructor's recommended training techniques that brought to mind the dog training methods recommended by people like Blanche Saunders in her 1946 book, Training You to Train Your Dog. These earlier training techniques mostly involved using your hands or the leash to manipulate the dog into a position and then verbally praising the dog with something like "Good boy!" This current instructor did not allow the use of food or petting as a reward maintaining that "Your words of praise should be enough. Dogs don't require codling and they will respect you more and behave more obediently if you don't pamper them."
More modern trainers accept the scientific data which says that reward training, usually with food rewards, is a more effective method of teaching a dog obedience (click here or here for examples). Even instructors who are not particularly happy with the use of food during training usually allow petting the dog along with vocal praise, so it seemed to me that reliance on the human voice as the only form of reward was not the best training method.
Perhaps it was because of this recent experience that I began to poke around the scientific literature and found a research report that was recently accepted for publication in the journal Behavioural Processes*. The researchers were Erica Feuerbacher of the University of Florida and Clive Wynne at Arizona State University. Although this study did not look directly at learning it did look at the reward value of verbal praise and petting. A number of different variables were measured but the main thrust of the research is easily described. In the first experiment 2 chairs were set up roughly 3 feet apart. One person sat in each chair (either the dog's owner or an unfamiliar person) and the dog was brought into the room and led to each person in turn. One of the people verbally praised the dog using a high pitched happy tone and saying things like "What a good doggie!" The other person petted the dog in a friendly manner but said nothing. Then the dog was led to the other side of the room and released. It is well established that dogs seek out things which they find rewarding and will stay close to people who make them feel good, so the measure that the scientists used was how long the dog stayed near each person. Once the dog was in a marked zone near a person that person continued to verbally praise and talk to the dog or to pet them for as long as the dog was close. The main result was that the dogs stayed near the person who did the petting, whether that person was their owner or an unfamiliar person.
In a follow-up experiment the researchers used only one person in a chair. The most important measure for our purposes is in the condition where a person used verbal praise compared with a condition where the person simply sat in the chair and made no attempt to interact with the dog at all. Here the investigators found that the verbal praise did virtually nothing at all. Ignoring the dog or verbally praising the dog were equally ineffective, and the dogs were not attracted and did not stay close to the person in either condition.
Now this recent experiment dealt with only the reward value of verbal praise and petting but did not specifically address the issue of whether verbal praise by itself was enough to sustain obedience training. For that answer we have to go all the way back to a study that was published in 1967 in the journal Psychological Reports**. This investigation was sponsored by the United States Army and carried out by Roger Mc Intire and Thomas Colley of the University of Maryland. It involved training dogs to respond to basic obedience commands (sit, down, come, heel, and stay). The training was remarkably similar to that advocated by the old-style trainer that I described earlier. For the first eight days the dogs received only verbal praise as a reward in the form of a happily spoken "Good dog!" If the dog did not respond to the command within 15 seconds the dog was physically manipulated into the position (or the leash used to tug him in for the "come" command) and then verbally praised. The measure of the effectiveness of training was how quickly the dog responded. The verbal praise was not very effective since on average it took 14 seconds for dogs to respond (remember that after 15 seconds the trainer enforced the command). For the next 17 days petting was added to the verbal praise. Petting involved five seconds of stroking the dog's head or ears, or occasionally chucking the dog under the chin. This led to a very rapid improvement in performance with the response time dropping down to a group average of only 5 seconds. Next the researchers decided to see whether the verbal praise by itself would be enough to maintain obedience performance once the tasks were well established. For the next 10 days the petting was eliminated and only the "Good dog!" praise was used. At the end of this period the performance had deteriorated to a level almost equal to the first eight days of training. For the last 10 days of the experiment the petting was reinstated to accompany the praise, and it took only a few days for the performance of the dogs to improve. With the petting back in the mix the responses to the obedience commands rapidly sped up and reached to their earlier rapid response level.
The conclusion that can be drawn from this research is that verbal praise alone is not very rewarding and certainly is not enough to establish or even to maintain performance of learned commands in dogs. So if an instructor chooses to ignore the data which says that food rewards are most effective in training, at least petting should be used since that kind of touching seems to be quite rewarding for a dog and seems to be enough to effectively sustain obedience training.
Stanley Coren is the author of many books including: The Wisdom of Dogs; Do Dogs Dream? Born to Bark; The Modern Dog; Why Do Dogs Have Wet Noses? The Pawprints of History; How Dogs Think; How To Speak Dog; Why We Love the Dogs We Do; What Do Dogs Know? The Intelligence of Dogs; Why Does My Dog Act That Way? Understanding Dogs for Dummies; Sleep Thieves; The Left-hander Syndrome
Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission
* Feuerbacher, E.N. & Wynne, C.D.L. (2014). Shut up and pet me! Domestic dogs (Canis lupus familiaris) prefer petting to vocal praise in concurrent and single-alternative choice procedures. Behavioural Processes. http://dx.doi.org/10.1016/j.beproc.2014.08.019
** Mc Intire, R.W. & Colley, T.A. (1967). Social reinforcement in the dog. Psychological Reports, 20, 843-846