Study 6: Shaping and Operant Conditioning
1 March 1999
Abstract
Each rat will be placed in one of the chambers and reinforced
by a pellet of food for successive approximations to the response
of pressing a lever.
The intention of this study is to provide an opportunity to shape an
operant behavior.
Introduction
The systematic study of operant behavior began with
Thorndike's classic experiments on problem solving (Thorndike, 1898). Thorndike, of course, found that "satisfiers" could
increase the frequency with which a certain behavior was
performed. The procedure was also developed by other investigators, most notably
by B. F. Skinner (Skinner, 1959). Subsequently, learning theorists noted that Thorndike's
procedure involved an explicit response contingency whereas
Pavlov's did not. Accordingly, many proposed a theoretical distinction
between Pavlov's "classical" procedure and Thorndike's
"instrumental" or "operant" paradigm.
In many respects, the relationship between these two
paradigms remains at issue even today. The fact that two procedures differ does not necessarily
imply that the underlying learning processes are different. Separating the effects of a "classical" process from an "operant" process is a difficult theoretical and empirical
problem.
For the most part, we will not be concerned with this
issue. However, we will be studying many aspects of operant
conditioning in the next few laboratory exercises.
Consequently, we will need to train our animals to
perform an operant response. This week's study is designed to let you "shape" your rat
to perform a simple operant.
Methods
Subjects:
Our Sprague-Dawley rats will serve as subjects.
Apparatus:
We will be using the six custom-constructed chambers
to test our animals. Each of these chambers will be fitted with a response
lever that the animal can press.
In addition, we will use the electric clocks to measure
time intervals during training. Reinforcers will consist of chocolate sprinkles delivered
to the food cup in each box after a tap on the chamber wall.
Procedure:
We should be able to run our lab in one shift, for a
maximum of 90 minutes. During that time we will use the method of "successive approximation" to shape the rat to press the lever for the chocolate
sprinkles.
Start by making sure your rat is magazine trained. That is, the tap on the chamber wall should be a
strongly conditioned stimulus for approaching the food cup.
Spend a few minutes reinstating the tap-food
contingency and verify that your rat will approach the food cup from anywhere in the chamber when he hears the tap.
Successive approximation means that the animal is
reinforced for performing some component of the response. The criterion for reinforcement is changed
systematically until the final level of performance is attained. Consequently, you and your partner should plan a
sequence of criteria (sort of like a curriculum for rats) that will
lead to a correct lever press.
For example, your plan might be something like:
- turn towards lever.
- move in direction of lever.
- approach to within 10 cm of lever.
- approach to within 5 cm of lever.
- raise left paw to lever.
- touch lever.
- press lever down.
Start by reinforcing the rat every time it performs the
first response in the sequence. Once it seems to perform that response reliably, move to
the next level and only reinforce those responses. Continue this way until you reach the last step. Every two minutes, note down the approximation that
you are currently working on.
When (or if...) the rat acquires the lever pressing
response, let him press it for continuous reinforcement for 5
minutes.
After the session, weigh and feed your rat.
Results
This exercise is not really an "experiment"; hence we do
not have any hard data to analyze.
However, I would like to see the sequence of
approximations and your time log.
In addition, I would like to see a "Discussion" section for
this and all subsequent lab reports. In the discussion, describe the rat's final performance and
evaluate the effectiveness of your own shaping behavior.
References
Skinner, B.F. (1959) A case history in scientific method. In: S. Kosh (ed.) Psychology: A study of a science. Volume 2 New York: McGraw Hill, pp. 359-379.
Thorndike, E.L. (1898) Animal intelligence: An experiment study of the associative processes in animals. Psychological Review, Monograph Supplement 2, 8.
Page Created: 28 February 1999 Last updated: 28 February 1999
Site maintained by: Michael R. Snyder <msnyder@psych.ualberta.ca>
|
