Traditional Training Vs. Operant Conditioning

This forum is all about training and behavior. Everything from potty training to working titles!

Postby TheRedQueen » April 26th, 2010, 8:31 am

Traditional Training vs. Operant Training:
What IS the difference?
by Brenda Aloff

http://www.behavior.org/animals/index.c ... avior.org/
animals/animals_aloff.cfm

Traditional Training vs. Operant Training:
What IS the difference?
Brenda Aloff

The first thought that came to mind when I looked at this topic was: Too Big, Too Much! How would I ever quantify and qualify the myriad differences? Which brings us to:
Difference #1. Traditional Training presents information in big blocks. Operant training presents information broken down into tiny approximations.

In Traditional Training the dog is expected to perform at a Terminal Response level, and anything less than this is "corrected" in some manner. In operant training, because behavior is built in tiny steps, whoever is getting trained has many opportunities for reinforcement. This has the long-term effect of making the behavior very strong; and also the tiny steps are much easier to assimilate. In this interchange the trainer's efforts are reinforced also. Success begets Success.

Difference #2. Operant Training is a mode of actually communicating with the animals and communicating with understanding rather than with the knee-jerk reflex of correction.

That “knee-jerk” correction may not be true focus of traditional training. However, it certainly seems to have become the main tool in the practice of the traditional trainers I watch, as it was in my own mind when I was being taught traditional raining.

With operant training, instead of relying primarily upon collar pops, or the lack thereof, to provide information, one works hard to establish a line of communication–a communication loop, to be exact. The behavior of one member of the loop directs the other member toward a behavior. Thoughtfulness prevails here, intimating also that calmness and confidence, a good atmosphere for thinking, will also prevail.

Establishing that line of communication via a primary reinforcer is just the beginning, the first approximation, of the level of communication that will come. I get excited just thinking about it! From the moment I offer a treat to the dog, we are embarking on a journey together, that of establishing a common language, a mode of communication between us.

Difference #3. Once this initial line of communication is established one uses the primary reinforcer to create the Heart & Soul of operant training: a conditioned reinforcer.

Now the magic can really begin, because with the addition of a CR, I can bring our relationship to a focus on information and not just on the giving and taking of food. As far as I can ascertain there is no real parallel to a CR in traditional training.

Difference #4. To raise the sophistication of the exchange even more, one can add an extinction cue, conditioned aversive, or “no reward marker.”

The terms are synonymous, so it's just a matter of individual preference. This leads to an even more sophisticated exchange of information. If the learner offers a behavior, and gets no click, the learner continues to seek reinforcement by offering a different behavior. The animal seems to ask, “It's not that? Well then, how about this? Or this?” Again, I see no apparent parallel in traditional training.

Difference #5. In traditional training all undesired behavior is punished (so the dog won't “learn how to do it wrong”).

In operant training, experimentation is encouraged. The smothering effect that punishment has on behavior often haunts you later in the form of unpredictable avoidance responses. Therefore, with punishment one may “train in” many behaviors that will require fixing or retraining later. Also, since the animals are not allowed any experimentation to discover What Works vs. What Doesn't Work, they usually wait to experiment until a circumstance arises where punishment or the punisher is not operative. (When you are using traditional methods to train a dog for competition, the principle opportunity for this experimentation is during competition in the obedience ring itself!)

With operant training, my dogs and my students. dogs do all their experimentation during the acquisition phase. If the dog experiments later, the undesired response is easily extinguished because of the established reinforcement history already in place for correct responses.

Difference #6. Correction-based training leads to distractability

Anyone who has done traditional dog training is aware of the devastating results of correction, particularly a poorly timed correction or a cumulative series of corrections. For the soft or very sensitive animal there is nothing left to work with; the dog simply shuts down or stops responding altogether. For the more resilient dog, however, your own behavior becomes the best reason to ignore you and look around at what the environment has to offer.

Recently I worked with a typical correction-trained dog. I took the leash and the dog couldn't wait to figure out some way to by-pass me. I was just In The Way Of All The Fun. Before I could even train this dog I had to back-track and teach the dog the value of the conditioned reinforcer, which all by itself will prevent many problem behavior problems from developing. Contrast this with my operant trained dogs, even the cross-over dogs [dogs who were traditionally trained first, and then converted to operant training: Ed.] The operant-trained dogs spend all their time thinking about what reinforceable behavior to offer me next. Instead of focusing on the problem of “How do I get around her to win the prize and reinforce myself?“ operant-trained dogs spend their time solving the problem of “How do I align myself with her to win the prize?”

Difference #7. Operant Training works on all breeds of dogs and all varieties of temperaments, as well as all other species with a spinal cord.

I originally began searching for alternative training methods when I began to train my first smooth fox terrorist [sic] for the competition obedience ring. Traditional methods quickly proved to be a dead-end with my independent, predatory, dog-aggressive Smoothie, (and all of those behaviors were quickly escalating). Operant training netted me amazing results. If, with operant training, exotic animal trainers can elicit voluntary blood taking with a Rhinoceros, surely we can obtain the comparatively mundane Sit and Down with our domestic dogs.

Difference #8. Operant Training allows you to train behaviors that you may not be able to train traditionally.

If you want, for instance, the scent discrimination exercise performed reliably and with a specific topography, that is, eagerly and happily, operant training can elicit results that may not be obtained with Traditional Training.

Difference #9. The trainer's mental attitude.

The emotional baggage that our own words carry is so vast. Everyone knows the danger of labeling a child. If a child is labeled “difficult” by adults, other adults and educators treat him or her differently. If a child is labeled “stupid” by classmates, children who hear this before they even know that child will treat that child differently upon meeting them. Basically, just a label, all by itself, can maintain a correction cycle. This also happens with adults: a co-worker is labeled “lazy,” and whether true or not, now that person will have to prove to co-workers that she is not lazy. For humans, the spoken and written word has far-reaching effects on the mental attitude or state and has a determining factor in the end result of the behavior that that person exhibits.

When I participated in traditional training I was constantly looking for what the dog was doing wrong, so that I could administer a correction to that wrong behavior. The operant trainer looks for what is right, and administers a reinforcer. The effect on the TRAINER is as profound as the effect on the dog.

Traditional vs. operant training classes for the public
Here's an example. I was teaching an operant class at my local Kennel Club. The class before mine was a traditional training class. I would watch the participants training their dogs and then leaving the building. Most of the participants were not smiling. Their body postures were tense (human and dog). For the most part, the dogs were focusing their attention outward, at the other dogs and everywhere but at their handler. The people did not converse or interact with one another very much; they just got their coats and left the building.

My class would enter. After the first night they were no longer inhibited by being strangers, so they entered the building chatting and laughing. They would come in and sit down with their dogs. The dogs would give their attention over to the handler relatively quickly and settle down without noise, offering Sit and Down behaviors for reinforcement opportunities. By the third night of class I commonly saw my students offering help to each other, perhaps holding a dog while the person took off his or her coat, or sharing training experiences: This is what worked for Max. The major difference here was the mental attitude of the handlers. The people who were working with operant training had more tolerance for their dogs. They were expecting the dogs to display incorrect behaviors, because that is NORMAL for dogs, so the frustration level for the handler was lower. Also, because we WANT the dog to offer us behavior so we can give them feedback, the fact that the dogs were sometimes offering behavior that hadn't been asked for was not upsetting to the handlers.

With the above scenario I do not mean to imply that traditional training is all wrong and operant training is heavenly. And of course, the people in the class before mine did interact some; my observation is just that the interactions were minimal in comparison to the class who was working with operant training.

Technique vs. instructor skill
Also the key difference here was not, in my mind, the instructors. The instructor before me had more experience than I, and was an excellent trainer. The difference was in the technique. If you are in the mode of looking for wrong, that is where you are and what you are doing. Criticism spawns more of the same, and it is difficult to pull out of that mind set. I remember myself, going home from traditional training classes with a cloud hanging over my head. I was frustrated with my Fox Terrier, who was doing poorly with the training; and with myself for being too ignorant to understand why. My husband would notice as soon as I got home, and ask me why I wasn't having any fun.

This is the aspect I love best about operant training. Instead of looking for the Wrong, I can just observe behavior and give the dog feedback: either “Yes I want to see that behavior again” or, “Not quite what I was looking for, try something else.” Suddenly the focus is not: “What can I find that is going wrong,” but “What can I find that is going right?”

This does wonderful things for ME. I remain relaxed and my frustration level is lower. If I do get frustrated I find that it is easier for me to stop training, return to the drawing board and try again tomorrow. I might not get the behavior today, but I am confident with my technique and I know I will get it tomorrow or maybe the next day. Since I am in the state of forgiving my animal, and being non-judgmental toward them, I am kinder to myself and can make errors of my own without passing judgment on myself. This was something that I did not expect from merely a training technique, this change in myself that I enjoy so much.

Difference #10. The operant trainer's view of life in general

One of my students and I were discussing this topic recently. She noted that she generalized this behavior of concentrating on correct behaviors instead of blaming and finding fault. from her dog training into how she looks at life in many other aspects. Her comment was: "I am no longer afraid to be wrong, or embarrassed by it. I figure if I just keep working on it, I will get it right. By looking to the things I get right, I don't dwell on what I got wrong, and I get a lot more done!”

So the end result, one I certainly didn't anticipate, was the same phenomenon my friend noticed. I feel better about my interactions with my pets, my friends, my family, and my students, and I have more confidence in myself.

Brenda Aloff is a pet behavior consultant, obedience instructor, and operator of a boarding kennel and Training Center Heaven On Arf in Midland, Michigan. She is a member of the Association of Pet Dog Trainers.

"I don't have any idea if my dogs respect me or not, but they're greedy and I have their stuff." -- Patty Ruzzo

"Dogs don't want to control people. They want to control their own lives." --John Bradshaw
User avatar
TheRedQueen
I thought I lost my Wiener... but then I found him.
 
Posts: 7184
Location: Maryland

Postby maberi » April 26th, 2010, 8:37 am

Looks like an interesting article. I don't have time to read it right now but will on my lunch break.

My one question after skimming the first couple of paragraphs is, isn't traditional training actually operant training (conditioning)? Aren't P+ and R- components of what we term operant conditioning?
Look beyond what your own eyes see
User avatar
maberi
I Save My Empty Calories For The Bottle
 
Posts: 2781
Location: rochester, ny

Postby maberi » April 26th, 2010, 11:38 am

I think a lot of what I would agree or disagree on in this article would be dependent on what her definition of traditional training is.

I agree with the general idea she is trying to convey, but her usage of some of the terminology is quite confusing and I honestly don't see that many trainers out there these days that are to the extreme that she is describing. Most of them use a combination of the principals of operant conditioning, although some are a bit heavy on the P+ and R-, and those are generally the trainers I disagree with the most.

I'm not a big proponent of training a behavior using P+ or R- and cringe seeing a dog trying to figure out what the trainer wants in these types of situations. Another pet peeve that drives me crazy is when trainers are proofing a behavior using P+. I've seen it time and time again when they exclaim the dog knows the behavior but they are proofing the behavior in a context the dog has never seen before and quite honestly the dog DOES NOT know the behavior in that context yet. A dog that knows sit in a quiet distraction sterile environment, does not know sit in a room full of other dogs running around and playing. Just like I know how to drive a car on the expressway but I would crash and burn if you put me in a race car and told me to go 200 mph around a race track.
Look beyond what your own eyes see
User avatar
maberi
I Save My Empty Calories For The Bottle
 
Posts: 2781
Location: rochester, ny


Return to Training & Behavior

Who is online

Users browsing this forum: Bing [Bot]

cron