Shaping a Rat to Press a Lever Left alone, rats will rarely press a lever. If they do, it is often accidental. They may brush their tails across it, or bump into it. In order for the law of effect to work--for behavior to be strengthened by satisfying consequences--the behavior must occur at least once so that the satisfying consequence may be produced. Shaping or reinforcing successive approximations of the desired response is used to teach the rat to press the lever. The experimenter waits for for some movement toward the lever to occur and then delivers food. This makes movements toward the lever more likely. The experimenter now waits until the rat moves closer to the lever before presenting food. Each food delivery requires a response which is closer to a lever press than the one before it. Within a short time the lever press response can be firmly established. Verplank's Verbal Conditioning Study College students were unaware that Verplank was conducting
the following experiment. The experimenter carried on what appeared to
be an informal conversation with the subject but actually behaved according
to a plan. The experimenter determined in advance to reinforce statements
of opinion such as sentences beginning with "I believe," "It seems to me,",
etc. The reinforcement consisted of the experimenter saying "You're right,"
or "I agree" after each statement of opinion. Extinction was carried on
in another portion of the experiment--silence following a statement of
opinion. During the period when verbal reinforcement was given, statements
of opinion increased markedly. During extinction, they decreased. The experimenter
controlled verbal behavior in this situation in much the same way that
an experimenter controls rats in a Skinner box. In some studies of this
kind, subjects may begin to realize that the experimenter is actually manipulating
their verbal behavior. However, there is evidence that in some cases they
do not (Rosenfeld and Baer, 1969)
|