Ask Lemmy

Clicker training is just paired stimulus to provide an easier and cheaper mode of reinforcement, training a behavior is separate

like you do the clicker training by associating a noise (or whatever) consistently with a positive stimulus. what stimuli you use to prepare depends on the learner as different stimuli have different potency depending on the learners preferences. eg you can say a blanket “I’ll use food” for your dog and for some dogs this is fine. mine certainly seems to be food indiscriminate with no serious preference and very few refused items. But even with that I still need to take care not to use the non preferred (he spits out lettuce and celery. otherwise literally anything gets him salivating)

but then to change a behavior you’re still relying on operant conditioning which would be something like upon exhibiting the desired behavior provide access to reinforcement consistently and then fade it out as the behavior strengthens. Operant conditioning is much more complex than this of course but this is a pretty standard jumping off point.

That said there are pros and cons to clicker training humans. This is something that is practiced and even has some evidence behind it. The clicker solves a lot of potential reinforcement issues: it’s far easier to deliver immediately (which matters a lot). But I worry about the potency loss translating a strong reinforcer to a clicker. If you pair it with a food you really love or something it will potentially be effective but never as effective as the food itself.

This is still potentially worthwhile as food reinforcers are often problematic (increasing caloric intake, often food reinforcers for people aren’t healthy options, promoting unhealthy eating habits) and reducing it to a click eliminates those issues. But if the behavior you’re trying to create is particularly difficult or aversive the reduction may mean the potency is no longer high enough to motivate.

Often this can be countered by making the behavior less complex and working up to it (eg instead of learning a complex task in its entirety breaking it down into more manageable chunks). In practice this may look like just initiating the task at first and providing reinforcement, then as comfort increases raising the bar for reinforcement. Eg I need to keep my room clean but I hate cleaning so to start out I provide reinforcement for just picking up one item/small area. But then when I do that consistently I raise the bar and now I have to pick up 2 areas. Etc. or you could approach as a tolerance thing, I start by cleaning for only 3 minutes and reinforcing, then 5, 7, 10, etc. numbers are arbitrary and depend on the learner. Approach depends on the learner too, the toleration approach makes more sense for most people but if you do a bad job cleaning and need to develop the skill of cleaning thoroughly the first can make more sense. Then reinforcement is not time based but quality based, Eg did you clean the area sufficiently even if it took you 8 minutes. Drawbacks and positives for every approach

And of course there’s the issue of delivering your own reinforcement. If you control access what’s to stop you from just taking the thing even though the behavior wasn’t exhibited. These strategies typically work better with external control of r+, but some people do have the self discipline to do it alone.

There’s a LOT more to conditioning and reinforcement but I’m getting bored of this lmao. Also you may notice I didn’t describe anything about punishment. That is intentional because it is generally at a much higher risk of creating adverse effects and some studies suggest it is not nearly as effective as reinforcement based strategies wrt general population (and some specialized populations)