4

My parents have two cats, who are great cats, but have learned some slightly annoying habits since I left for college. My parents don't maintain reinforcement of what I taught the cats.

I will have the cats for 2-3 weeks, and plan to use that time to teach them better habits for my parents. Unfortunately, I know that whatever I teach the cats will not be consistently enforced once I return them to my parents.

The habit I'm most interested in addressing now is Silver's tendency to start meowing excessively, and get right underfoot when he thinks someone is about to get out the wet cat food (once a week). I want to make him less of a nuisance when the food is being prepared.

His trained "good" behavior will be rewarded by only giving him his wet cat food if he behaves. My parents will not enforce this rule; he will get wet cat food regardless of his behavior. I plan to convince him that food is a reward for his "good" behavior, so he will maintain the "good" behavior longer.

I can handle the training; my question is, what exact training is most likely to stick, considering the teachings will not be properly enforced?

Another example is if I teach him to sit on a vocal command until he is fed, my father won't use the command until Silver has annoyed him, at which point Silver may have already figured out food is incoming and he doesn't need to listen to get his food. Should I train Silver by putting the can away when he starts begging for it, to try to teach him that begging means he doesn't get a reward; will that stick with him longer then the idea that sitting gets a reward?

In a nutshell, how do I get the most idiot-proof training that will stick to the cats without anyone consistently reinforcing the habits? I plan to continue reinforcements on my weekend visits.

  • There is no way a cat will be able to understand that because he meowed too much earlier in the week, he gets no treats today. – Oldcat Jun 3 '15 at 17:09
3

That's an interesting question about reinforcement schedules. First I'll try to define some important terms of positive reinforcement training:

  • Positive reinforcement: something that the subject is given and that increases the probability that the reinforced behavior will be repeated in the future
  • Reward / reinforcer: something that is given after the subject performed the behavior to reinforce that particular behavior
  • Continuous reinforcement: you reinforce every time the behavior is performed
  • Variable reinforcement: you reinforce randomly or following a defined variable schedule. The subject can't really predict if the reinforcer will be given for that repetition
  • Extinction: the disappearance over a period of time of a response when the behavior is not reinforced

So your question is: What reinforcement schedule provides the best resistance against extinction?

A variable reinforcement schedule is what makes behaviors the most resistant to extinction.

When you train a new behavior you first start with continuous reinforcement until you are quite confident that he understands what's going on. Then you start proofing the behavior in new situations, with distractions, etc. and you switch to a variable reinforcement schedule.

It actually makes the behavior stronger, that's a bit counter-intuitive, as you might expect that less reinforcement leads to a weaker response, but it's the opposite. As the cat/dog doesn't when/if he will receive a treat he's kind of "making every repetition count", performing in hope that he'll get one.


That's about the reinforcement. I wanted to add a few points:

  • A reward is supposed to be absent during the time the behavior is performed and appears after. Here you'll be preparing the food while you're asking him to sit. It is not a very clear setup.
  • A variable reinforcement schedule works well if the cat can't predict when/if he'll get a reward. That means that it shouldn't be "someone = no reward", "someone else = reward", that's predictable
  • What you describe is not really about giving a reinforcer or not. It is about reinforcing one behavior or another one, which is totally confusing for the cat. When your parents will prepare the food, they will actually reward whatever behavior the cat is doing, most likely jumping around and trying to get to the food as quickly as possible.

Here is an alternative setup that you could consider:

  • train a correct "sit" independently: no when you'll give wet food, do a few sessions a day with tiny food treats that are easily accessible. As you are happy with the response, switch to a variable reinforcement schedule. Proof the behavior in many circumstances (in the kitchen, with someone else around, etc.). Ask your parents to participate.

  • Eliminating unwanted behavior requires: 1/ a good rule structure that everyone enforces 2/ an alternative behavior to which you can redirect. Note that when I say "rule structure" and "enforce", I don't mean punishment and "being the leader" or whatever. What I mean is: IF the cat knows another behavior, redirect to that behavior, be very consistent (the whole family should do it). The difference with many trainers is that I emphasise that the "rule structure" should be trained. At the beginning you'll have to be patient, ask for a sit, reward, etc. Then it will eventually become automatic. At the beginning reward him when he sits (with a treat, not with the wet food).

The important part is that you'll have to do some active parts of the training, and when it will be done sufficiently you and your parents will go for passive training: the cat know how to sit, is not always rewarded for that, understands the rule structure, etc. So the only thing your parents have to do is to wait until he sits.

  • It's worth noting that the power of variable reinforcement is not limited to pets, but rather is present in all animals, including humans. Variable reinforcement is an extremely powerful tool for influencing human behavior, as is well known to the developers of addicting browser games and the like. Powerful enough that some are starting to wonder if there isn't some line where it becomes too abusive to use it to sell your gimmick, and if we don't need some kind of limit on it. – KRyan Jul 13 '18 at 14:55
0

The approach I've taken with my kids is that treats are not offered every day or at predictable times, so they have no reason to think pestering me for them might work. On the other hand, when I'm offering a treat I give them a very clear signal (a loud click of the can tab or banging the dishes together, repeated) so they know what to expect.

Breaking a habit once acquired can be difficult. It may be easier to train positively for a behavior that's incompatible with the one you want to discourage. I have broken cats of pestering habits, but that required a willingness to deliberately not give them what they were asking for until they stopped that behavior -- see my comment elsewhere about dealing with cats scratching at bedroom door by using getting earplugs for a few weeks rather than giving them anything that might be taken as attention or encouragement.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.