3 ms·

Training Humans

4 points by mad44 18y ago

RiderOfGiraffes 18y ago
One of the things that's rarely mentioned in these training attempts: Start by rewarding consistently, then become inconsistent, only rewarding, say, one time in 3. Or 4. Why? If you consistently reward every instance of good behavior, then when you stop, the good behavior will tail off remarkably quickly. You are caught in the trap of having to reward every good instance. On the other hand, suppose the trainee has got used to being rewarded at random for good behavior. Then you stop entirely. The first time the trainee says "Yes? Oh, no. Oh well, maybe next time." The next time the trainee says "Yes? Oh, no. Oh well, maybe next time." And on it goes. Randomly reinforced good behavior persists for a very long time after the rewards stop. Very occasional rewards makes it even longer.
- mad44 18y ago
  Nice trick. Thank you for sharing.
- chris11 18y ago
  Actually that is not quite the most effective way to do it. You want to start out rewarding randomly, and keep rewarding randomly. Basically you want to set up a lottery where on average, every x times the person is rewarded once. But you should keep the chance of getting rewarded constant. http://en.wikipedia.org/wiki/Reinforcement#Simple_schedules http://en.wikipedia.org/wiki/Reinforcement#Simple_schedules This is also the most effective way to combat extinction. Near extinction, there is usually an extinction burst. And to keep the person or animal motivated, you want to up the rewards when they are about to quit. Just think of someone addicted to drugs or alcohol. When they are trying to quit, they get cravings, and can binge on the substance. http://en.wikipedia.org/wiki/Extinction_(psychology) http://en.wikipedia.org/wiki/Extinction_(psychology)
  - RiderOfGiraffes 18y ago
    Thank you - nice to see some of the technical details of this subject. My post was from memory, and I'm pleased it was as accurate as it was.
  - eru 18y ago
    But you may still profit from increasing x over time. Especially if your subject is learning how to perform what you want better and better. (Compare http://en.wikipedia.org/wiki/Shaping_(psychology) http://en.wikipedia.org/wiki/Shaping_(psychology)) If you want to mess with your probands mind, at some point switch silently to a variable time schedule. ("Variable Time (VT) provides reinforcement at an average variable time since last reinforcement, regardless of whether the subject has responded or not." from the Wikipedia article) By the way, for example computer games usually combine several reinforcement techniques.

4 points by mad44 18y ago

One of the things that's rarely mentioned in these training attempts: Start by rewarding consistently, then become inconsistent, only rewarding, say, one time in 3. Or 4. Why? If you consistently reward every instance of good behavior, then when you stop, the good behavior will tail off remarkably quickly. You are caught in the trap of having to reward every good instance. On the other hand, suppose the trainee has got used to being rewarded at random for good behavior. Then you stop entirely. The first time the trainee says "Yes? Oh, no. Oh well, maybe next time." The next time the trainee says "Yes? Oh, no. Oh well, maybe next time." And on it goes. Randomly reinforced good behavior persists for a very long time after the rewards stop. Very occasional rewards makes it even longer.

mad44 18y ago
Nice trick. Thank you for sharing.
chris11 18y ago
Actually that is not quite the most effective way to do it. You want to start out rewarding randomly, and keep rewarding randomly. Basically you want to set up a lottery where on average, every x times the person is rewarded once. But you should keep the chance of getting rewarded constant. http://en.wikipedia.org/wiki/Reinforcement#Simple_schedules http://en.wikipedia.org/wiki/Reinforcement#Simple_schedules This is also the most effective way to combat extinction. Near extinction, there is usually an extinction burst. And to keep the person or animal motivated, you want to up the rewards when they are about to quit. Just think of someone addicted to drugs or alcohol. When they are trying to quit, they get cravings, and can binge on the substance. http://en.wikipedia.org/wiki/Extinction_(psychology) http://en.wikipedia.org/wiki/Extinction_(psychology)
- RiderOfGiraffes 18y ago
  Thank you - nice to see some of the technical details of this subject. My post was from memory, and I'm pleased it was as accurate as it was.
- eru 18y ago
  But you may still profit from increasing x over time. Especially if your subject is learning how to perform what you want better and better. (Compare http://en.wikipedia.org/wiki/Shaping_(psychology) http://en.wikipedia.org/wiki/Shaping_(psychology)) If you want to mess with your probands mind, at some point switch silently to a variable time schedule. ("Variable Time (VT) provides reinforcement at an average variable time since last reinforcement, regardless of whether the subject has responded or not." from the Wikipedia article) By the way, for example computer games usually combine several reinforcement techniques.