Operant Conditioning Terminology

Jean Cote

Administrator
Staff member
Do you have any sources for this?

I might have to change some of my lessons according to this, I believe I might have mixed terminology in the Reinforcement lesson. :)
 

bipa

New Member
Yeah, it can be quite confusing at times. That’s why I tried to use really simple and hopefully straightforward terms in my table.

The table you found isn’t talking about what is being added, but rather what the result is after something has been added or taken away. A subtle distinction, but it changes the table.

. | decreases likelihood of behavior | increases likelihood of behavior presented | positive punishment | positive reinforcement taken away | negative punishment | negative reinforcement

Here's my table again, this time with the results of the actions added. I've used my original examples. Maybe that will help clear up the confusion between the two perspectives. You'll notice that the two tables do match up, but are organised differently.


. |Add...|Take Away...
Something nice|Positive reinforcement (giving treat will increase likelihood of behaviour)|Negative punishment (stop play with puppy will decrease likelihood of behaviour – nipping)
Something unpleasant |Positive punishment (jerking leash will decrease likelihood of behaviour)|Negative reinforcement (stopping the shock when dog has done task will increase likelihood of behaviour)

I took my table from

by Mark Plonsky, Ph.D.

I liked the way he presented it differently from the standard. It just seemed to make more sense to me his way. Guess I should have linked to his article, sorry about that.


For more sources, take a look at some of these articles:
 

l_l_a

New Member
Good post Bipa, thank you!

I wish whoever first defined the operant conditioning terms would have used different words to make it less confusing since the scientific meaning of "positive" and 'negative" is different from our everyday conversational meaning!
 

bipa

New Member
Some folks gets confused about positive punishment (adding something unpleasant when the dog has done something unwanted) and negative reinforcement (something unpleasant to the dog is already happening, and is taken away at the moment the dog does something that is wanted).

Let's take the example of leash corrections. Some people use lots of light leash corrections when teaching their dog to heel. Shock collars operate on the same principle (only difference is that people's arms don't get tired from giving constant corrections.) The dog will learn to stay in the proper position in order to avoid having his neck jerked around or being shocked. That is actually negative reinforcement. It is also called avoidance, or escape.

If the dog is walking nicely, sees a cat and lunges, at which point the leash is jerked or the dog is given a shock, then that is an example of positive punishment. Nothing unpleasant was happening to the dog until he lunged and only then was the leash yanked or shock given.

The problem with negative reinforcement is that in effect the dog is being made to feel uncomfortable BEFORE he has even had the opportunity to do anything at all. If the dog is still at the beginning stages of training and is uncertain how to perform, then theoretically this unpleasantness could last a long time, or even forever. Being made to feel discomfort for no reason that the dog can see or understand, can lead to learned helplessness. That is the major argument against the use of this type of training.

So for an easy way to spot the difference, folks have to ask themselves what was the dog experiencing BEFORE the dog did the action, and before the command was given. Then it becomes very clear if the leash corrections or shocks are positive punishment, or negative reinforcement.

Hope this helps!
 
Top