http://www.nytimes.com/2011/03/15/health/15brody-animals.html
If only they shared more information about how these therapy animals are trained! Everything is very hush hush, especially with service dog training – which I believe needs to be much more efficient.
Trailer for a DVD on operant conditioning and its applications at chicken camp and zoos in Sweden. Starring Bob Bailey, my mentor’s mentor. It’s one of my dreams to attend chicken camp while Bob Bailey is still teaching it!!
now that you know what the true definition of reinforcement + punishment are, it’s important to clarify the terms positive and negative. These terms are also ones that already have common misconceptions in our everyday language. Positive = a good thing, and negative = a bad thing. However, this is not applicable to ABA.
Positive denotes the presentation of a stimulus – something that had not been there before the behavior.
Therefore, positive reinforcement occurs when after a behavior, something that was not there before is presented, and this causes the frequency of the behavior to increase in the future.
Example: Your dog is standing in front of you. You put some treats in your hand, and lift your hand up over the dog’s nose. This will cause your dog to bend his head as far back as possible until it will cause him to sit down. You then give him the treats. You repeat this one more time and the same thing happens. You remove the treats from your hand and just lift your empty hand, and your dog sits. Now every time you lift your hand in front of your dog, your dog sits.
In this case, the treats that are given to the dog after he sits, is the new stimulus that is presented (positive) and since the sitting behavior increases, it is reinforcement.
Negative denotes that something that was already in the environment is taken away.
Therefore, negative reinforcement occurs when after a behavior, something that was present in the environment is removed, and this causes the frequency of the behavior to increase in the future. This is usually associated with aversive stimuli.
Example: You get a mosquito bite on your leg, and it itches. A lot. You know you shouldn’t, but you scratch it. When you scratch it, the itching goes away and it feels good. So in the future, you scratch your mosquito bite more often to get rid of the itching.
In this scenario, the itch is the stimulus that is taken away (negative) by the behavior, and since scratching increases in the future, it is reinforcement.
Next post: Positive & Negative Punishment
When most people are faced with the terms reinforcement and punishment, they say that reinforcement is “good” and punishment is “bad”. In our language, this is true – reinforcement has a positive connotation, while punishment has a very negative one.
The Oxford English Dictionary provides the following “psychological” definition of reinforcement:
The action or an act of establishing or strengthening of a particular response to a stimulus or a learnt behaviour through the repetition of a rewarding or unpleasant event contingent on the response.
YES! I was so excited to see that ABA-accurate definition in there.
Unfortunately, though, the OED’s “psychological” definition for punishment is not as good as the one for reinforcement:
The infliction of an unpleasant stimulus, as pain, deprivation, etc., on an organism as a method of behavioural control, so that unwanted forms of behaviour are suppressed and the desired behaviour pattern is established.
That’s more of an everyday definition that we use.
In order to understand the basics of ABA, it is essential that you know that reinforcement means increasing the future probability of a behavior contingent on a response. punishment means decreasing the future probability of a behavior contingent on a response.
Note how the only difference is either increasing or decreasing the behavior. There is no “good”, “bad”, “right”, or “wrong” associated with either of the terms in ABA.
Let’s say you have a plant, and you water it every other day. The plant stays healthy and green (what you want). As a result, you continue to water your plant every other day. Therefore, your behavior of watering the plant every other day is reinforced by the plant’s apparent health. But, again, reinforcement is not always associated with favorable situations.
Let’s say you have a son, and you are shopping at Toys `R Us for a present for the child’s friend. Your child sees a brand new Bakugan toy and asks you to buy it for him. You say “no”. Your son then throws himself on the floor, cries, and yells “I WANT THE TOY!!!!!” repeatedly. You then groan “fine~” and buy it for him. On future shopping trips, you find that if you say “no”, your child automatically throws a tantrum. Voila! Your child’s tantrum behavior is now reinforced by you giving him what he wants.
On the flip side, you water your plant every other day. The plant starts to wilt and becomes yellow. You think that your watering schedule is affecting the plant’s health, so you stop watering the plant every other day. Therefore, your behavior of watering the plant every other day is punished by the plant’s rotting. Similar to the reinforcement examples, punishment does not always have to be associated with adverse situations.
Now that your son throws a tantrum every time he wants something in a store, you are almost at the point where you don’t want to take him anywhere. So you decide to put your foot down and you just won’t take anymore of this! The next time you are in a store, your son throws a tantrum over a candy bar. You ignore his tantrumming and take him straight out of the store. After two more events like this, you notice that your son’s tantrums are diminishing. Therefore, you have punished his behavior of tantrumming. Congratulations! See, just because you punished the behavior, doesn’t mean you were a horrible parent or inflicted pain or physical trauma to your son.
Whew! I know this is a lot, but I hope you guys are still following along!
Next post: Positive + Negative Reinforcement/Punishment
I think the hardest thing to accept/learn in ABA is the basic terminology. There is so much information, but I’ll start with what I think makes up the foundation of the science.
There are two general classes of behaviors and learning:
1. respondent (or reflexive) behaviors / respondent conditioning
2. operant behaviors / operant conditioning
Respondent behaviors are those that are innate – those that do not require any learning. A good example is the eye blink reflex. We have a reflex of blinking our eyes when something comes too close to our eyes or comes close suddenly – like a poof of air.
Respondent behaviors can be conditioned. As in, a new stimulus that was not originally a respondent behavior can become one. Ivan Pavlov is probably the most well known researcher to demonstrate this. Pavlov observed that dogs in his lab would salivate when the door to the lab was opened. The sight of the door opening seemed to have become a “signal” to the dogs that they were going to be fed. In order to test his hypothesis, Pavlov conducted his famous experiment during which he played a metronome right before he fed the dogs. After a couple of trials, he found that just playing the metronome alone would trigger salivation. The metronome, which initially was a neutral stimulus, was now a conditioned stimulus. And in the presence of that conditioned stimulus, salivation occurred; it formed a conditioned reflex. Something that was unrelated now triggered an innate behavior to occur.
Operant behaviors are those whose future frequencies are determined by their consequences in the past. Unlike respondent behaviors, which are triggered by a stimulus before the behavior, operant behaviors are controlled by what happened after the behavior in the past. The example I used in my first post about turning on the light switch is an operant behavior (we turn on the light switch more often when it is dark because in the past, turning on the light switch has made the room brighter).
Operant behaviors can also be conditioned, or learned. We can use the principles of operant conditioning to teach an organism to do more or less of a behavior. In order to do this, though, we have to understand reinforcement and punishment – two terms that have very different definitions in ABA as compared to our everyday vernacular.
Next post: the foundation: reinforcement and punishment
PS. I haven’t gotten around to editing the video of teaching midas how to do “paw” (and it was much harder with him than expected), but I will post a video or two as soon as i get around to it!
I’ll be visiting Midas this weekend and will be teaching him some basic tricks.
I’m going to try my best to take videos and post them here for your entertainment!
ABA, or Applied Behavior Analysis, is the science of behavior modification. Although it is under the field of Psychology, it is very different from a lot of subfields within Psychology that we are familiar with – disciplines such as cognitive science, psychotherapy, or Freudism.
ABA deals with behaviors that one can see. It is also based on the assumption that individuals behave. Behavior is a result of the interaction between an individual and the environment. For example, when we come home in the evening, we turn on the light. The house is dark, so we flip the switch. Turning on the light is a behavior, and it was done because we wanted to change the environment we are in. Furthermore, in the past, we have flipped the light switch because by doing so, the room becomes brighter. We are reinforced by the brightness in the room. Therefore, in the future, the likelihood that we will flip the switch when the room gets dark increases. This is called reinforcement history.
At this point, you’re probably like, “whaa?” and wondering where I’m going with this. As a student of ABA, I have learned a lot about the theory behind the practice and want to share that theory with you. I believe that ABA is a universal science – it is just as applicable to me as it is applicable to the squirrel right outside your window. Animals learn the same way we do – from experience. I want to apply ABA to dog behavior and want to show everybody how effective ABA can be when it comes to modifying our pet’s behaviors.
I hope you come back soon to learn more about what dogs plus ABA can equal!
with midas & regatta
the pink dog 