When it comes to training dogs there appears to be a consensus which says the more frequently you train your dog the faster he learns and the better his performance. As a result of this belief some people who are preparing their dogs for obedience competitions can be seen training their animals for an hour or more every day. Contrast this to a typical pet owner, who might spend an hour training their dog on one of the weekend days, and for some this is augmented by an hour a week spent in a dog training class. Are the competitive dog trainers doing the right thing while the typical pet owner is impeding their dogs learning through lack of time spent training them? Up until relatively recently there was no scientific data which addressed this issue, however a pair of studies published in the journal Applied Animal Behavior Science
by researchers from the University of Copenhagen seems to answer this question, and the results seem to be contrary to what many trainers might expect.
It seems surprising to me, given the number of people who own dogs as pets, that so little scientific research has been devoted to the best practices for training dogs. Perhaps that is changing a little since just a while ago there was a preliminary study which looked at the best ways to reward a dog during training (click here). However, with the addition of these two studies we now have data which may tell us how often we should train our dogs each week, and how long each training session should be.
The first study by Iben Meyer and Jan Ladewig compared the performance of dogs trained one session per week with those who were trained almost daily for five sessions per week. The dogs tested were beagles, and the task involved the dog going to the end of the table and placing his paw on a mouse pad. Training was done using food rewards and a clicker was used to signal a correct response. The task was taught using successive approximations, and was broken into four levels to allow scoring. Each training session consisted of 15 repetitions (repetitions are technically called "trials" by psychological researchers). The investigators then monitored how many training sessions it took for the dogs to learn to perform the task at the highest level. Many people might be surprised to find that it took fewer training sessions to teach the dog the task when he was trained only one session once a week. The dogs trained weekly took an average of 6.6 sessions to master the task while the dogs trained five times a week took an average of 9 sessions. Thus the daily training routine involved nearly 50% more teaching time expended by the human handler for the dog to learn the task. Of course the flip side of this is that the dogs receiving daily sessions knew the task after just under two weeks of training while the dogs in the weekly sessions took 6 to 7 weeks to master it.
The second study was by Copenhagen researchers Helle Demanta, Jan Ladewigb, Thorsten J.S. Balsbya and Torben Dabelsteena. They also used beagles but this piece of research was more elaborate. It compared the performance of dogs trained one or two times a week with those trained five times a week, and also looked at the effects of the length of each training session. The task used was a bit more complicated, requiring a dog to go to a basket some distance away and to sit and stay in it while the trainer walked a distance away and then finally returned to the dog. The task was broken into 12 stages of performance for the purposes of training and scoring. As in the previous study training involved a clicker and food rewards. Each training session involved 6 trials.
This time the dogs were divided into four teaching conditions. One received a single training session (that is 6 trials) once or twice each week, while another received three back to back training sessions (18 trials)once or twice each week. A third group received one training session (6 trials) five times a week while the fourth group received three training sessions (18 trials) five times a week. Each dog received a total of 18 training sessions (for a total of 108 trials) and it is only how those sessions are distributed which is varied. It all means that the amount of training and handler time spent with each animal was the same.
Since the task was more complicated and the number of training trials per session was smaller than in the previous experiment, it is not surprising that not all of the dogs managed to attain perfect performance. This allowed each dog to be scored based upon the highest level (1 to 12) of accomplishment that it managed to achieve. On average the dogs receiving one training session once or twice each week demonstrated the highest level of performance, averaging around level 11, which is close to perfect which would be level 12. The dogs receiving 3 training sessions in a row once or twice each week lag significantly behind averaging around level 8. The dogs receiving one training session 5 times a week only averaged level 7 at the end of their instruction, while the dogs receiving the most intensive training, 3 back-to-back sessions a day for 5 days a week, had the worst overall performance managing only an average of level 5.
The important thing to emphasize here is that the amount of trainer time spent teaching each of these dogs was exactly the same. Therefore, on the basis of actual effort and time expended by the human instructor, the dogs trained only a couple of short sessions each week seem to be the best bargain, since at the end of their lessons those dogs are performing more than twice as proficiently as the dog subjected to long, almost daily training sessions.
Of course, as in many other areas of life there are trade-offs. In calendar time the dogs with the daily long sessions will complete their course in under 4 weeks, while those with the short sessions twice a week will take at least 9 weeks. However the dogs with the spaced out practice will perform twice as well as those who had their training crammed together into daily sessions. Remember that the actual amount of time that the dog will have spent being taught by a human will be the same in both cases. At the very least these data show that the casual pet trainer who trains his dogs only one or two short sessions each week can end up with a well trained dog despite the small amount of time spent teaching their pet each week—but of course, some training is still required!
Stanley Coren is the author of many books including: Born to Bark; Do Dogs Dream? 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, reposted without permission
Data taken from:
Demant, Helle; Ladewig, Jan; Balsby, Thorsten J.S.; Dabelsteen, Torben (2011). The effect of frequency and duration of training sessions on acquisition and long-term memory in dogs. Applied Animal Behaviour Science, 133(3-4), 228-234.
Meyer, Iben; Ladewig, Jan (2008). The relationship between number of training sessions per week and learning in dogs. Applied Animal Behaviour Science, 111(3-4), 311-320.