Click conditioning / target question (Psychology theory)

[cham87]

http://www.youtube.com/watch?v=8C5fTBLM ... e=youtu.be

Please watch the video it is only 1 minute long, and is easier for me to ask in video format than through the medium of text.

Thank you!

[Michael]

Clicker/target training is based on operant conditioning (Skinner). Pavlov is classical conditioning. In the case of the dog, the bell makes the dog salivate. This is a classically conditioned response that is outside the dogs control. It's not necessarily making the dog hungry, it's just making it salivate. Therefore, even if some sort of classical conditioning occurs in conjunction with using a clicker, it is not particularly important or relevant to operant training.

However, there may be some benefit to "warming up." When you begin a training session, some times doing a few easier known behaviors to get the bird interested/hungry in training makes the tougher ones go better. But I think this works more like having an appetizer and getting your system going and irrelevant to pavlov.

A more realistic example of classical conditioning might be a cockatiel's crest rising at the sight or sound of a clicker. Even if no reward is administered.

[Andromeda]

Just watched your video (nice shirt, by the way).

I'll try to answer your question but when Michael comes along he might have a better answer. Edited to add: I was still writing this while Michael responded so see his post for the short and sweet version.

The experiment with Pavlov's dog was classical conditioning. Clicker training is operant conditioning.

However, you are right in that teaching the animal to associate a click with a treat in and of itself is classical conditioning.

In classical conditioning you are working with involuntary responses, so, using Pavlov's dogs as an example, the involuntary response was the salivation (an automatic, natural response to seeing food). The "neutral stimulus" was the bell (which later became the conditioned stimulus), and when paired with the food (the unconditional stimulus) enough times the bell became a "conditional stimulus" at which point the bell itself caused the involuntary response (salivation).

The reason why clicker training is operant (meaning "voluntary") conditioning is because you are not working with involuntary responses. A bird does not automatically have a natural reaction to a targeting stick or trick prop, so the point of operant conditioning is to modify behavior using a type of reinforcement. There are four different types of reinforcement and the one used in clicker training is "positive reinforcement" meaning that you reward desired behavior and thereby increase the frequency of that behavior.

Once the clicker has been established as a conditional stimulus it's used as a "bridge" (or "secondary reinforcer"). A bridge is used to mark the exact moment the bird does something desirable and also bridges the gap in time between when the bird does the desired behavior and when it receives a treat.

Once the bird associates the click with a treat you must always follow the click with a treat or you risk the clicker becoming less effective. There is an exception to this and that is when you are using a ratio schedule of reinforcement (you do not reward every time but rather on a variable or fixed schedule) as opposed to continuous reinforcement (rewarding every time).

If you are just starting out with clicker training you're best starting off with rewarding every single time for new tricks, and once the bird has a trick down-pat you can phase out the clicker and choose to reward on a variable schedule.

If you still are wondering about how to get your bird to cooperate when it's being less than cooperative I can post about that in a follow-up but hopefully this answers your core question: You don't want to use the clicker if you're not going to treat because what you have taught the bird through classical conditioning is "the click marks desirable behavior and also means you have a treat incoming." Since the bird is conditioned to know that click means treat that's what you should deliver every time. If you do not pair the clicker with a treat the clicker will lose its meaning because unlike Pavlov's dogs you are not using the clicker to elicit an involuntary response (in Pavlov's case salivation).

[Michael]

Very well put. I'm not on top of all the classical conditioning phraseology because for all intents and purposes it is unimportant to parrot ownership EXCEPT that you want to avoid suddenly classically conditioning phobias.

[Andromeda]

Very well put. I'm not on top of all the classical conditioning phraseology because for all intents and purposes it is unimportant to parrot ownership EXCEPT that you want to avoid suddenly classically conditioning phobias.

He said he busted out some books and did some reading about Pavlov which is why I thought I'd get into the specifics of everything since he seemed curious.

But yes, classical conditioning phraseology is unimportant to parrot ownership.

[cham87]

Thank you.