Operant Conditioning



Operant Conditioning is a form of learning (one of many) which states that responses from an organism are controlled by the consequences of responding. Classical conditioning is distinguished from operant conditioning in that the former involves reflexive responses such as salivating and blinking. Operant conditioning in contrast is assumed to involve voluntary responses. However what has become evident is that both forms can exist jointly in some aspects of behavior.

Terminology


Reinforcement occurs when the consequence of a response increases the chance of that response occurring again. Reinforcement is seperated into two distinct types;
Positive reinforcement involves a response being rewarded by adding something positive to the situation following the response. Feeding a dog a biscuit for sitting on command
Negative Reinforcement involves removing an aversive condition following the response. Removing the choker collar from the dog after he sits to a command.
Punishment is defined as something that has the effect of decreasing the response from occurring again. Punishment is also divided into two distinct types;
Punishment involves creating an aversive condition to follow the response. Giving the dog a kick in the stomach after it sits on command.
Negative punishment involves terminating a pleasant condition following the response. Removing the dogs food it it sits on command.

Skinner distinguishes between different sources of reinforcement. Primary reinforcers are events that occur without learning taking place and include food, water and sex. cited in LeFrancois,G.F (1994). Secondary reinforcers are those that become reinforcers because they are linked with primary reinforcers. The sound of the fridge door opening becomes a secondary reinforcer to a cat that associates the sound with being fed. A generalised reinforcer is a secondary reinforcer which is paired with more than one primary reinforcer. Money is a generalised reinforcer as it can buy food, water and sex.


Reinforcment can be manipulated according to different schedules.
Continuous reinforcment involves reinforcing every desired response
Intermittent reinforcement involves reinforcing only some of the responses. When using intermittent schedules of reinforcement the experimenter can further choose two more options.
Interval schedules involve reinforcing the responses intermittently and according to time. Every 5 minutes a response is reinforced.
Ratio Schedules involve reinforcing a proportion of responses. Every 5th response will be reinforced.
Schedules can be further divided into random or fixed. Where random schedules are unpredictable and fixed schedules are unpredictable.
The effects of different schedules leads to different effects on the acquisition and extinction of the response. Initial learning is most effective if continuous methods of response are used. However continuous schedules lead to faster extinction of the response when the reinforcer is removed. The most effective schedule found is an initial continuous schedule followed by a variable ratio schedule.

One of the most useful things to develop from Skinners theory is the idea of shaping as a method of training animals. Shaping involves reinforcing every behavior that brings the animal closer to the desired response.


Interesting Links
How to train with a clicker
Dogs
More about Skinner
Back to Page 1

References

LeFrancois,G.R.1994. Theories of Human Learning (3rd ed)Pacific Grove, Ca: Brooks Publishing Inc.