all dogs train

Image components licensed from clipart.comSource: Image components licensed from clipart.com

I recently was a witness to a rather heated argument between two dog trainers about the relative effectiveness of "clicker training" compared to "marker training". After much noise and hand waving it appeared that neither could provide convincing evidence and the issue was left unresolved. Fortunately, there is now scientific data which does answer the question and should settle their debate.

For those of you who are not all that familiar with clicker training it really is quite simple. It depends upon a little plastic toy that makes a clicking sound. The dog learns that whenever he hears that sound a treat is coming immediately after. So the dog sets about trying to produce behaviors which cause the trainer to sound that click knowing that it will be followed by a reward. In so doing the dog learns which behavior is wanted. Thus the training sequence is quite simple: get the behavior; mark that behavior (with a click); reward the behavior. The more that the behavior is repeated and rewarded the stronger that behavior becomes. To understand what is going on psychologically during clicker training you should read this.

Marker training is exactly the same as clicker training, the only difference is that instead of using the clicker as a reward signal you use some other signal. It might be a voice signal, such as "Yes" or "Good dog". When the basic concept behind this form of training was introduced to the general public by psychologist B. F. Skinner in a 1951 Scientific American article, he chose a clicking sound simply because it appeared to be more precise and was easily heard. However according to Skinner himself the marker doesn't even have to be a sound but could be some sort of visual signal such as a light flash or a hand movement.

Clicker training began to grow in popularity among dog trainers following the publication of Karen Pryor's book "Don't Shoot the Dog" in 1984. Her workshops and videotapes emphasized the use of the clicker and after a while the presumption among dog trainers grew into a belief that the only form of truly effective marker training is when a click is used as a signal. Numerous dog training websites suggest that the use of a clicker is the "only" efficient way to train dogs for any kind of complex task. Meanwhile a number of other dog trainers continue to side with psychologists who say that there is nothing special about the clicking sound but any kind of signal that can be precisely delivered when the dog produced the wanted behavior will work as well as a training aid.

Recently a team of researchers headed by Cinzia Chiandetti of the University of Trieste decided to see if clicker training was the most effective way to teach dogs. They used a sample of 51 pet dogs, each of which was to be trained on a novel task and then also tested to see how well they generalized their learning (that is how they applied their learning to new situations which were new but somewhat similar). A group of 17 dogs were trained using a clicker and 17 were trained using a verbal reward marker, namely the word "Bravo". When these investigators designed the study they thought that a third group of 17 dogs would be trained using only a reward and no marker, but they later recognized that the dogs probably were responding to a visual signal, namely the bowing of the trainer when she moved forward and bent over to deliver the reward to the dog. Regardless of the marker used all of the dogs were highly motivated because the reward was either a piece of sausage or cheese, depending on their preference.

Training used a method called "shaping" or "successive approximations" in which the dog is rewarded for behaviors which get closer and closer to what they finally have to learn to do. The task of they were learning was to open a plastic breadbox by pushing the handle up with their nose or muzzle. The dogs had up to three training sessions a day until they were able to open the breadbox eight out of 10 times in a row.

The dogs were then retested a week later to see how much of the original learning they had retained. At this time there were also tested on two other tasks which could be solved using the same motion of their muzzle and head. The simple test used a different colored breadbox that had the back removed (in other words very similar to the box used during training). The second test was somewhat more complex because the test apparatus was a different size and shape and made out of wood instead of plastic.


Related posts: