Si quieres ver este blog en versión española, pincha aquí.

Dec 12, 2012

Operant Conditioning

Developed by B.F. Skinner, it is a learning method (understood as "behavior modification based on experience") supported on the relationship between a given behavior and its consequence, so that the subject can voluntarily decide what behavior develop (operate) based on the result that it will report him.
This approach based on the consequence we obtain exists in all animals, including humans, and is governed by the following premise:

- One consequence perceived as positive after a particular behaviour is given, will cause that behavior to recur more often or  with more intensity (increase);

- One consequence perceived as negative after a given behaviour, will cause that behaviour to recur with less frequency or with less intensity, even extinct (decrease).

Thus, the fact that will determine which behaviors are repeated more or less often, will be the consequence caused by those behaviors, often repeated if the result is pleasant and disappearing if the result is unpleasant. Applying this theory is easy to conclude that, if we want to change a particular behaviour we must change its consequence (later on I will explain that behavior is also influenced by its antecedents, which can be modified too, in what analysts call “the ABC of behavior (Antecedent – Behaviour – Consequence).

These consequences that influence behavior can be classified as follows, being essential to clearly understand this scheme in order to evaluate our actions and thus understand what strengthens or weakens certain behaviors, desired or not:


Increases behavior
Decreases behavior

Add stimulus

Positive Reinforcement (+R)

Positive Punishment (+P)

Removes stimulus

Negative Reinforcement (-R)

Negative Punishment (-P)


As we see, what causes an increase in the behavior (in frequency and / or intensity) is called "reinforcement", being it "positive" when it requires the addition (+R) of something, and "negative" if it involves the subtraction (-R) of something; equally, anything that causes a decrease in the behavior (in frequency and / or intensity) is called "punishment", being it "positive" when it requires the addition (+P) of something, and "negative" if it involves the removal (-P) of something. Please note that the word "negative" in this context has no connotation or qualify anything at all, just meaning "to remove, to subtract”, just as the word "punishment" has no negative connotation nor its popular meaning, but rather it simply means "to reduce or to extinguish" a behavior.

Let´s see some examples of each one:

- Positive reinforcement (+R): (I offer something you like). If the parrot is silent, I will reward him with a peanut. Thus, the desired behavior (the animal being silent) will tend to be repeated because the parrot obtains a reward (peanut) that satisfies him. Similarly, if the animal screams and we reward him (even if we are unaware of it), he will tend to scream more often or loudly. For example, if your parrot, as most of them do, loves the hustle and bustle and you approach yelling at her to reproach her own screaming, you are actually reinforcing her behavior so she will scream more often.

- Negative reinforcement (-R): (I remove something you do not like) My neighbour comes home and quickly approaches my parrot´s cage repeating "nice bird, nice parrot, hello, tell me something, parrot, hello, nice" ... however, if the terrified bird remains silent my neighbour soon becomes bored and gets away. Next time my neighbour enters home, the parrot will be very quiet to cause the neighbour disappears, ie, the behaviour of being silent is reinforced by the disappearance of the neighbour (which is what the animal is looking for).

- Negative punishment (-P): (I take away something you like) Who has not been found in the following scenario? We leave our apartment, close the door behind us and call the elevator. We get inside and press the "0". We descend and exit the elevator on the ground floor, bumping head on with one neighbour. We kindly say "good afternoon", but he does not respond (even though it is clear that he has heard us). Nothing. Silence. We leave and think "such a rude guy… next time I won´t greet him unless he does it first." And, indeed, next time we see him out of the elevator we don´t say a word, waiting for him to greet us. This is a typical case of negative punishment because our behavior (greeting the neighbour) will tend to decrease because of the removal of the stimulus (the greeting from the neighbour) that motivated us to behave that way.

- Positive punishment (+P): (I give you something you do not like).The parrot screams ad nauseam and his desperate owner hits the cage irritated, startling the animal. The screams tend, a priori, to diminish due to the violent response of the owner, but as we will deeply explain in other entry, this “solution” is not such, and it will only entail future problems.

Thus, we see that we can influence our pet´s behavior modifying the consequence it obtains as a result of its own actions, enhancing those desired responses and trying to minimize the unwanted ones… and here is where the job of the sensible trainer begins.

No comments:

Post a Comment