Study 3: Pavlovian Conditioning

January 27, 1997

Synopsis

In this study, we will examine Pavlovian conditioning using the anticipatory activity paradigm. Each rat will receive two signals: one which signals the delivery of food and one that does not. Pavlovian conditioning will be assessed in terms of the rat's tendency to show anticipatory behaviours to the signal for food.

Background

Although Pavlovian conditioning is one of the oldest learning paradigms to be investigated, there still is some dispute about the best way to study it. On the one hand, there are investigators who feel that Pavlovian conditioning is best studied using basic reflexive systems similar to those employed by Pavlov (Schreurs, 1989). Others argue that Pavlovian conditioning is a general process of association that can be studied with any behaviour that shows the consequence of an associative process (Rescorla, 1988).

Our third study is based on the second of these viewpoints. It is patterned after an early experiment by (Sheffield & Campbell, 1954), and a later one by (Holland, 1977). In this preparation, food serves as the unconditioned stimulus (US). The conditioned response is the animal's activity during a signal for food delivery. In their experiment, Sheffield and Campbell found that conditioned activity increased substantially as a result of pairing the conditioned stimulus (CS) with the delivery of food.

Holland's contribution was to show that the activity measures exhibited a pronounced interaction with the modality of the CS. In an earlier experiment, Holland had obtained good activity conditioning when tone was the CS, but poor conditioning when light was the CS. In his 1977 experiment, Holland demonstrated that this pattern was due to the fact that the two types of stimuli controlled very different behaviours.

In this study, we will use Holland's paradigm to study the acquisition of a conditioned response. We will use tone and light stimuli, and examine the change in behaviour to the CS as a function of experience.

Methodology

Subjects:

Our Sprague-Dawley rats will serve as subjects.

Apparatus:

We will be using six custom-constructed chambers to condition our animals. Each chamber will contain a small dish, through which we can deliver a food reinforcer. These pellets are 45mg pellets of food, manufactured by the P.J. Noyes company. In addition, each group will be provided with a switch-controlled light and tone.

Behavioural Coding:

We will use Holland's behavioural coding categories. They are:

Perambulate: Change in position involving all four feet, including walking across chamber, circling and/or jumping suddenly to another position; often accompanied by sniffing.
Rear: Standing on hind legs with both forepaws off the floor, usually (not always) stretching to full extent, forepaws usually (not always) on top of side walls of chamber, often pawing walls; may be accompanied by sniffing or slow side-to-side movement of head. Does not include grooming movements, even if performed while standing on hind legs.
Magazine: Standing motionless in front of food magazine (the food cup) with nose or head within magazine, sometimes (rarely) gnawing on edges of magazine opening.
Head-jerk: Short rapid horizontal and/or vertical movements of the head, usually oriented toward food magazine; hindquarters motionless. Infrequently occurring with rear: in those cases, only head-jerk scored.
Head-jerk/hind: Head-jerk plus movement of hindquarters, either side-to-side or forward-backward. Simultaneous display of head-jerk and perambulate (rare) also scored as head-jerk/hind.
Other: Grooming head, body, or tail; scratching; gnawing sides of chamber; lying motionless on floor; sniffing, when not accompanied by other behaviours.

Procedure:

Each rat will be run for a 90 minute session. We will have five chambers, so this will mean two shifts. During the first 26 minutes, you will train your rat to approach the food cup when a pellet has been delivered. Deliver a pellet by gently tapping the side of the chamber and rolling a pellet into the food cup. Try to keep your own movements as unobtrusive as possible, so that they don't become a signal themselves. Wait until the rat is next to the food cup, and deliver a pellet. Repeat this procedure until you are confident the rat will approach the cup whenever a pellet is delivered. Then wait until the 26 minutes has elapsed.

Following the 26 minute "magazine training" we will deliver 16 trials of the light and 16 trials of the tone. One will be followed by food delivery (the CS+) and the other will not (the CS-). Each stimulus will be presented for 10 seconds. Record for each half of the stimulus which of the above behaviours was the dominant behaviour during the interval. That is, you will have an observation for the first 5 seconds of the CS and the last five seconds. If your rat has an even number, follow the tone by a food pellet; if your rat has an odd number, follow the light by a food pellet. As in the magazine training phase, tap the side of the chamber just before presenting the pellet. Present a stimulus every two minutes, in the following order: LTLT TTLL LLTT LTTL LTTL LLTT LLTT LLTT. This is a random sequence constructed by presenting blocks of two lights and two tones with the order of light and tone randomized within a block.

After the session, weigh and feed your rat.

Results and Discussion Assignment

As in any learning experiment, we are interested in a change of behaviour as a result of experience. Does the behaviour to the CS+ change as a result of pairings with the food pellets? Does behaviour to the CS- change?

To examine these questions, group your observations into blocks of four trials. Within each block, there are eight observations (two for each trial). Compute the percentage with which each behaviour occurred during a block, and plot this as a function of the four blocks for both stimuli. The format of this graph is up to you, but consider how to make the graph readable and informative

Obtain the data from a team with the opposite relation between light and food and tone and food (i.e., if light was an CS+ for your rat, get the data from a team for which light was the CS-). Plot this rat in the same way.

Have you established evidence for learning of the CS-food relation? In particular, do you think our study provides a sufficient control for nonassociative effects? If there is a difference in learned behaviour controlled by light and tone CSs, what do you think may underlie it?

References

Required:

Rescorla, R.A. (1988). Pavlovian conditioning: It's not what you think. American Psychologist, 43, 151-160.

Schreurs, B.G. (1989). Classical conditioning of model systems: A behavioural review. Psychobiology, 17, 145-155.

Sheffield, F.D. & Campbell, B.A. (1954). The role of experience in the "spontaneous" activity of hungry rats. Journal of Comparative and Physiological Psychology, 47, 97-100.