[gib65]

Have there ever been any studies that show a distinction between conditioning a behavior and conditioning the rate of CHANGE of a behavior? For example, let's say I had a rat that I wanted to press a leaver every 10 minutes or so, and I only reinforced it once for every leaver press so long as it's been ten minutes since the last reinforced leaver press. And let's say there was another rat that, for every leaver press, I wanted it to do so sooner than the time between its last two leaver presses, thereby increasing its leaver pressing frequency over time. Have there ever been any experiments to show that it is possible to control which one of these two kinds of conditioning one would want to apply?

Gib

[gib65]

Have there ever been any studies that show a distinction between conditioning a behavior and conditioning the rate of CHANGE of a behavior? For example, let's say I had a rat that I wanted to press a leaver every 10 minutes or so, and I only reinforced it once for every leaver press so long as it's been ten minutes since the last reinforced leaver press. And let's say there was another rat that, for every leaver press, I wanted it to do so sooner than the time between its last two leaver presses, thereby increasing its leaver pressing frequency over time. Have there ever been any experiments to show that it is possible to control which one of these two kinds of conditioning one would want to apply?

Gib

[Kelton]

So to be clear....the animal will, after his first responses, begin to gradually respond at an acelerated pace?
It this is what you're saying, I think that it would be very challenging to condition an animal to do such a thing in with only the lever and no other signaling device. It could be impossible, but it's an interesting question.

[badchad]

hmmm, good question. Your first example, e.g. conditioning a rat to press a lever only after a specified time I can tell you it is done often. Our laboratory does these experiments and I'm a contributing author on the paper. We trained a group of rats so that they press a lever every 16 seconds for a food reward. If they press too soon, the time resets. Thus, after some time, we are able to say that the rat can reliably judge the passage of 17 seconds. Then we gave the rat LSD and watched the shift. LSD is known to alter human perception of time, and it did the same for the rat.

Your second example is more complex. COuld we train the rat to press sooner and sooner each time? hmmm, I'm not sure. If we kept the intervals exactly the same each time, I think it's possible. To my knowledge however, it hasn't been done before.

[coquina]

Quote

COuld we train the rat to press sooner and sooner each time? hmmm, I'm not sure. If we kept the intervals exactly the same each time, I think it's possible. To my knowledge however, it hasn't been done before.

Suppose you shocked the rat if it waited too long and fed it if it didn't? Not particularly pleasant for the rat, I admit. How many times does a rat have to have a negative reinforcement forfore it learns anything from it?

[Kelton]

Reinforcement of Accelerating Rates of Behavior: That would probably be the title of such a procedure.....I can almost remember such a thing from a reading, but just not sure. I just can't see it really happening, though.

Coquina, negative reinforcement is a good thing. Everyone wants to be negatively reinforced because it means that an unwanted thing has been removed.

[Glider]

True. Just so people are aware of the terms:

Positive reinforcement = Application of a good thing (reward)
Negative reinforcement = Removal of a bad thing (reward)
Punishment = Application of a bad thing (er...punishment)

[Kelton]

In casual conversation, people often use "negative reinforcement" interchangeably with "punishment". I didn't mean to be disengenuous as I DO know what Coquina meant in her usage of the term.

There are some important problems with the use of punishment as a behavior change procedure and behavior analyst are very often hired to teach other types of practictioners about alternatives to punishment. That is, to teach others how to change behavior with the use of reinforcement rather than aversive methods. Ironically, the common reputation of the behaviorist is that of "the punisher!".

Punishment is clearly necessary in certain types of circumstances, but it is use FAR more than it should be and it's less likely to be used by a behaviorist than other forms of people in the field.

This is something of a change of subject isn't it?

This is my first ever avatar. Man, it was a bit of a trick getting it down to the acceptable size!

Kelton