Study 8: Concurrent Schedules of Reinforcement
15 March 1999
Abstract
Rats will be given a choice between two variable interval
schedules of reinforcement.
An interchangeover analysis will be used to determine whether the
amount of time
spent on one side is different than the amount spent on the other.
Introduction
Schedules of reinforcement have become standard tools for
the study of choice.
One paradigm, in particular, has dominated operant
research on choice; that paradigm is the use of concurrent
schedules of reinforcement.
In a concurrent procedure, an animal can choose between
two (or more) independent schedules of reinforcement.
Typically, each schedule is associated with a different
response.
Of interest is the amount and pattern of responding on each
schedule.
A good review is provided by de Villiers (1977).
Interestingly, however, there are few studies involving rats
as subjects.
Norman and McSweeney (1978) reported one such
study; it provides a good overview of how a concurrent schedule
can be established.
In our study, we will examine our rats' performance on a
concurrent variable-interval variable-interval schedule.
That is, our rats will be able to choose between two
schedules, each of which delivers reinforcement on a variable
interval schedule.
One of these will be a VI-100 sec schedule and the other will
be a VI-30 sec schedule.
For this week, we will concentrate on the pattern of
switches, or changeovers, between the different responses.
Next week, we'll examine how choice depends upon the
relative rate of reinforcement earned from each schedule.
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 two response
levers that the animal can press.
In addition, we will use 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:
Our lab will be 60 minutes
long.
During that time we will reinforce the rats according to a
VI-30 sec schedule on the right lever of each chamber, and a VI-
100 sec schedule on the left lever.
The interreinforcement times for each schedule are
given in Tables 1 and 2 (to be passed out in class). As in last week's study,reinforcement is delivered for the
first response after the time interval has elapsed since
reinforcement.
In this case, however, we have two responses, each with its
own schedule of reinforcement.
This means that a time interval could elapse on, say, the
left response, while the rat is pressing the bar on the right.
Your team will have to keep track of both schedules, and the
times at which a reinforcement "hold" is in effect.
In addition, to analyze changeovers, you should record the
time that an animal starts responding on a particular lever and
when he switches to the other.
Probably the best way to do this is to have each member of a
team watch one response lever.
He or she can keep track of behavior and reinforcement
availability on that lever.
As was the case last week, the digital timers will be
available to save eyestrain.
However, since only one timer is available, you won't be able
to reset the time after each reinforcement.
You'll have to note down the time that a reinforcement
occurred, add the next interval to that time, and then reinforce
the animal when he responds at the new time.
For example, suppose you're in charge of the right
response.
If the timer reads 154 seconds when a rat is reinforced for a
right response, and the next interval is 17 seconds, then
reinforce the rat for the next right response after the timer reads
171 seconds.
Remember also to record the time that the animal switches
from the left response to the right, and the time that he switches
from the right to the left.
After the session, weigh and feed your rat.
Results
We will be looking at changeover performance for this lab.
In point of fact, although the distribution of changeover
times is important to many theoretical analyses of operant
behavior, there are few studies that have looked at the pattern of
changeovers.
Heyman's study with pigeons (Heyman 1979) is one of these; he
found that changeovers were random in timewhich suggests
that changeover times should be distributed according to an
exponential distribution.
To my knowledge, there are no studies that have examined
this question with respect to operant behavior in rats.
As a simple graphical test of this hypothesis, plot the
frequency histogram of left and right changeover times.
Because there are training factors involved with learning
the concurrent schedule, you may want to plot changeovers
from the latter part of the session.
Decide what part of the session you want to plot, and
comment on what you think is a good basis for that choice.
References
de Villiers, P.A. (1977) Choice in concurrent schedules and quantitative formulations of the law of effect. In: W.K. Honig and J.E.R. Staddon (eds.) Operant behavior. New York: Prentice Hall, pp. 233-287.
Heyman, G.M. (1979) A markov model description of changeover probabilities on concurrent variable-interval schedules. Journal of the Experimental Analysis of Behavior, 31: 41-51.
McSweeney, W.D. (1978) Matching, contrast, and equalizing in the concurrent lever-press responding of rats. Journal of the Experimental Analysis of Behavior, 29: 453-462.
Page Created: 13 March 1999 Last updated: 13 March 1999
Site maintained by: Michael R. Snyder <msnyder@psych.ualberta.ca>
|