Register Here

  © 2008
  University of Idaho
  All rights reserved.
  Psychology Dept.
  University of Idaho
  Design - P&D  CTI

In essence, the failure of a response to produce a reinforcer causes a response to decrease until the organism no longer emits a response.  That is generally the technical definition of extinction,   Overall, you have some kind of response that has been followed by a reinforcer and you are now responding at some high steady rate.  Then you stop giving the reinforcer.  As a result, over time, the behavior decreases.  Now there’s a related aspect within reinforcement and extinction.  That concept is in slide three and called an extinction burst.  An extinction burst basically is this.  When you begin to stop delivering the reinforcer, the behavior first increases before it goes to zero.  So as we see in slide four, the organism has made a gradual increase in some response, and it now is at some high steady rate for a period of time.  Then we stop the particular reinforcer.  As a result of that stopping, the organism makes a major burst of activity and then it drops off and goes down to zero.

This is also shown in slide five using a cumulative recorder.  Here the organism basically has been doing a particular set of responding for a period of time.  Then extinction begins resulting in a major burst of activity before the extinction process begins to decrease the behavior

So how does that relate to you?  What is the importance of that?  Well let’s go back and look at slide four again.  Let’s say you are in love with somebody and you are doing all sorts of behaviors to be in love.  You’re sending this person flowers and you’re taking them to dinner and on and on.  Then they stop reinforcing you with attention or caring, they stop calling you, etc.  What happens to your behavior?  Well, you have a major burst of activity, that is you send more flowers, you start calling more often, you give them more things, you shower them with attention, on and on.  When none of that works, you basically begin to decrease the particular behavior that you were doing.  So, you have a major burst of activity, then, responding drops off.

A similar aspect occurs oftentimes with kids having behavioral disorders.  A kid with a behavioral disorder is usually being reinforced for some particular thing.  Let’s say it’s being reinforced for hitting or some other kind of thing and you decide, well, the reason this kid is hitting is because it’s getting lots of attention.  So, you decide to extinguish the behavior.  So, what initially do you see when you stop extinguishing?  When you start the extinction process, the kid starts hitting and acting out even more than it was before, then the behavior will drop off. 

The problem is that if you reinforce during that burst of activity, what have you done?  Well, you have increased the rate of behavior for some new activity.  Consequently, it will become even harder to extinguish.  This is the classic example of head banging behavior that you oftentimes see in institutional settings.  A little kid is doing some acting out behavior at home.  As a result, the parents decide they are not going to pay any attention.  So the kid starts to scratch and bite itself.  The parent makes them stop doing that by giving them lots and lots of attention.  So when they decide they are going to extinguish that behavior what happens? The kid starts hitting their head on the coffee table and the parent then gives it lots of attention.  Now it wants to make the kid stop, and what we have to do is put them into some kind of isolated setting with something padded where they can hit their head on the wall or on the table or whatever, where, but not hurt themselves.

A final example relates to most people: Crying babies.  A little baby doesn’t have good communication skills, and so what it does is cry.  It wants some love from mom and dad.  So, what does it do?  Mom and dad put it to bed and it starts to cry.  Mom and dad then give it some reinforcement.  They pick it up and give it some love.  So, what does the little kid learn?  The little kid learns that “Hey, I want mom and dad’s attention before I go to bed and so I’m going to cry.  So what does crying do?  Crying goes up and it’s now at a high steady rate.  So, now any time mom and dad leaves the kid, even for a brief second, the kid starts to wail its head off.  Mom and dad then decide to put it on extinction.  So, what do they do?  They put it in the crib at night when it’s going to bed and say “We’re going to hang in there; we’re going to hang in there.”  So, what does the kid start to do?  The kid starts to cry and it cries for awhile.  Then it starts to cry louder and louder, screaming, yelling and on and on.  But instead of mom and dad hanging in there, what do they do?  They go in to the kid and pick it up.  What does the kid learn now?  The kid learns that, “Hey, when I yell and scream and whine and cry really loud, I am going to get picked up.”  So the behavior has a major burst of activity.

 On the other hand if mom and dad had hung in there and let the kid cry and scream for a little while, crying would have decreased and stopped.  Ultimately, mom and dad would be able to put the kid in the crib and the kid would stop screaming and yelling.  So extinction bursts are extremely, extremely important.  And if you reinforce a behavior during that particular time, it just changes their behavior to a new level, which causes lots of problems. 

Now there’s a wide variety of factors that will influence extinction and those are shown starting in slide six.  The classic example relates to the size of the reinforcer.  In essence the larger the size of the reinforcer, the more reinforcers you get, and the longer it takes to extinguish a behavior. 

Number two: the delay of the reinforcer. Occasionally the reinforcer is delayed.  As a result, the response will take longer to extinguish.  So the little kid is now crying and screaming and yelling.  Mom and dad do something and as a result of that, it takes longer to decrease the behavior.  So, what are you going in and see the kid?  What does the kid learn?  The kid learns that sometimes instead of only crying for 10 minutes before mom and dad show up, sometimes I have to cry 40 minutes.  So, the kid learns it sometimes takes a long time before mom and dad will come in.

Finally, partial reinforcement schedules.  Intermittent reinforcement schedules, take longer to extinguish.  This is very similar to the delay that we talked about in the previous paragraphs.  So, there’s a wide variety of variables that influence extinction. 

Now let’s talk about the other way to decrease a behavior.  That is through some kind of punishment.  When we talk about punishment, as we see in slide eight, it’s often a very misused procedure, and often used in ways where behavioral changes won’t occur.  The classic example is the criminal justice system.  However, when used with a reinforcing procedure you get much better results, especially in educational setting.

 Now there are two major papers that are the classic papers that everyone should read if you’re go on in psychology.  These two papers are listed in slide nine and the first paper is Punishment by Azrin and Holtz and the second paper is Punishment by Campbell and Church.  Azrin and Holtz are more applied; Campbell and Church are more theoretical.  However, both groups of theorists and both papers are the classic papers on punishment and they still hold today.

So let’s talk about the Azrin and Holtz aspects of punishment first.  Basically Azrin and Holtz define punishment as a consequence of a behavior that reduces the future probability of a behavior.  And it’s the reduction of a future probability of a specific response as a result of the immediate delivery of a stimulus for that response, that is, the response followed by something bad, you reduce the behavior.  Now the stimulus for Azrin and Holtz is called the punishing stimulus and the process of delivering those punishing stimuli is called punishment.

Now there are several important aspects of this definition for Azrin and Holtz and these are defined in slide 12.  The definition of punishment, says Azrin and Holtz, is not in some kind of terms a subjective feeling or states of being.  It’s not just the punishing stimulus because it makes you unhappy, it has to be something very specific.  In addition, a specific event must be produced by a specific response in order for it to be considered to be a punishing stimulus.  Basically decreases in responding aren’t sufficient reasons to classify a procedure as a punishing stimulus, as we saw with the process of extinction.

Now Azrin and Holtz primarily looked at some aspects of punishing stimuli and these begin to be listed in slide 13.  There are about five or six of them.  I’m just going to list them and walk through them.  The first thing that Azrin and Holtz required for an ideal punishing stimulus is precise physical specification.  That is, you must be able to measure the ideal punishing stimulus extremely accurately.

Number two, the ideal punishing stimulus must have constancy in terms of the actual contact that it makes with a subject.  That is, a stimulus measured at its source maybe different from it’s particular contact on the organism.

Number three, as we see in slide 14, the organism must not be able to perform a behavior to reduce or escape from the aversive stimulus.  That is, it can’t perform some unauthorized escape behavior.  So for example, if you are using electric shock, one of the key tricks for a rat, is to roll over on its back and with its hair underneath it protecting it from the shocking stimulus.  Also, if the shock is relatively low, the animal will actually not receive any shock.

The next one.  There must be few skeletal reactions to the stimulus.  That is, if you give some kind of stimulus, it can’t paralyze the organism.  So if you give some kind of punisher that shuts the animal down, it doesn’t learn anything.

And finally, the punishing stimulus must have the capability of being varied over a wide range of values.  Thus, you should have a variety of different aspects or different values of the particular punishing stimulus.

Now for Azrin and Holtz, the classic punishing stimulus that met all those requirements was electric shock.  However, there are many types of punishing stimuli.  So let’s show a few of them in slide 15.  The first type is an air blast.  Animals and humans do not like blasts of air, especially in the face.  Skinner also used a second type of punishing stimulus called the bar slap.  It was a device that was like a mousetrap.  So when the organism padded or pushed a bar in an operant chamber, it would get smacked on the paw.  Another type of punishing stimulus is noise.  Noise can be extremely effective, especially loud noise.  Provided the hearing system is not destroyed.  The classic example might be an air horn.

Electric shock can be an extremely effective punishing stimulus as you might know.  Finally, smells and water mists are other types of punishing stimuli which can be extremely effective.  These are used often used in group homes and other types of settings where electric shock cannot be used.

Now there’s a wide way that we can deliver punishing stimuli.  The classic way for doing most of the experiments is to use, as we see in slide 16, an operant chamber.  We’ve discussed an operant chamber before and shown you what it is.

Now, let’s talk about punishment.  Discussion of punishment types goes back to the time of Skinner, but Azrin and Holtz talk about this as well.  There are two types of punishment. The first type, as we see in slide 17, is positive punishment.  This is defined as when a response is followed by a stimulus and the behavior decreases, the procedure is called positive punishment and the stimulus is called the punisher.  In general, when some kind of response is followed by a zap, ala electric shock, smack on the butt, whatever it may be, when the response is followed by a zap and the behavior goes down, that’s called positive punishment.

Now there are some major differences between negative reinforcement and positive punishment and these are shown on slide 18.  In positive punishment, the response is followed by something that’s aversive and in essence the behavior goes down.  So we’re adding something and the behavior decreases.  In negative reinforcement, the response is followed by the removal of the aversive stimulus and the behavior goes up.  So in negative reinforcement, you are removing something, ala negative, and the behavior increases, ala reinforcement.  So remove an aversive stimulus and the behavior goes up.

Now there are two types of negative punishment.  So let’s talk about each of these in specific.  And the first type of these is shown on slide 20, and that is what is called response cost.  In response cost, the response are followed by the removal of the goodie and the behavior goes down.  It’s not like positive punishment where you give something aversive and the behavior goes down.  The classic example is the kid yelling in the house and the parent takes away it’s TV time.

Now the second type of response cost procedure relates to what is called time out.  Technically it is called time out from positive reinforcement.  As we see in slide 21, in time out, the response is followed by the removal of the reinforcing stimulus and the behavior goes down.  Generally, in a time out situation, the organism is placed in a neutral situation where there are no reinforcers or stimuli available.  The classic example goes with Dennis the Menace.  Dennis the Menace does something bad and he gets put in the corner.  Or a kid yelling in the house and the parents take the kid and place it in the corner or put it in their bedroom.

 Now let’s provide an example that relates to time out.  This is shown on slide 22.  The key about time out situations is that sometimes neutral situations have reinforcers.  So you have to make sure that the neutral situation has no reinforcers.  The classic example goes something like this.  The kid is yelling in the house.  Mom says go to your room.  Now what’s in the room? Toys!  So instead of the kid being negatively punished, the kid is actually being positively reinforced.  An alternative might be sitting the kid on the bed and having him stay on the bed quietly.  Usually there’s nothing on the bed.  Now if there are stuffed animals on the bed, then you might want to find a different place.  For example, when I put my child in time out, it has to sit on a step.  There’s nothing on the steps.  So, there’s not very much for it to do. 

This procedure can be quite effective.  The key is to not have any reinforcers available where the time out procedure is being used.

In the next section, we’re going to start talking about some variables that are related to punishment.  So until that time have a great day.