Re: What is behavior modification? Are there different methods of it?

Dear Kyle,

Behavior modification is the psychologist's term for 'training'. It is
accomplished through rewards and punishments. So when you hear about a
mouse's behavior being modified, it means that someone wants the mouse to
carry out some kind of behavior when cued and he/she does so by rewarding
the mouse (with food pellettes usually) every time the mouse displays the
desired behavior and/or punishing the mouse every time it displays the
opposite behavior.

There are several types of behavior modification. Rewarding and punishing
are two kinds. Another kind is direct vs. shaping. Shaping is used when the
desired behavior is not likely to occur naturally or by chance. You see,
before one can reward or punish a mouse (or any other creature), it must, by
chance, display the behavior. But when that is unlikely to happen, one can
"shape" the animal's behavior by starting with some behavior that is likely
to happen, and then from there gradually reward or punish the next closest
approximation until the behavior resembles the desired one enough. A third
classification is intermittent modification which means rewarding or
punishing only part of the time. There are four kinds: fixed ratio, fixed
interval, random ratio and random interval. Ratio modification means
rewarding or punishing every so many number of times the animal displays the
desired behavior whereas interval means every so many intervals of time
(like every 5 minutes). Fixed means a consistent number like every 5 times
the animal displays the behavior or every 5 minutes whereas random means
that the rewards/punishments come randomly like after 10 displays of
behavior then after 2 displays or after 4 minutes then after 15 minutes. You
might be surprised to hear that intermittent modifications works better than
constant (can you guess why?)

To answer your last two questions, mice can learn fairly quickly. With the
right consistency and a strong enough intensity, 2 or 3 times can be all it
takes. But of course, it all depends on how strong a reward/punishment you
use, what the optimum level of intermittent modification is for the animal,
and also whether the desired behavior comes naturally or not to the animal.
All else being equal, rewards usually work better than punishments. The main
reason for this is that when you punish an animal for an undesired behavior,
all this teaches the animal is what NOT to do. It may take some time before
it learns what the desired behavior is, and even then if it can stumble upon
a way of escaping punishment without having to execute the desired behavior,
it may attempt it.

Hope that answers your question.
Gibran