Shaping lever press behavior and fixed ratio training for water-deprived rats

Yong Ho Kim
Macalester College

Operant conditioning is a particular type of conditioning in which the subject associates particular behaviors with particular responses, and performs the behavior whenever it desires the response.

The leading psychologist to systematize behavior shaping was Skinner (Kalat 2002). He set up a box, in which a rat presses a lever or a pigeon pecks an illuminated disc. This box was called an operant-conditioning chamber. Based on his work, shaping is defined as “establishing a new response by reinforcing successive approximations to it”.

The standard method of operant conditioning consists in providing reinforcements for every instance of the behavior. This procedure is called continuous reinforcement. Once the subject has been fully conditioned, further restrictions on behavior-reward relationship can be imposed. For example, reinforcement can be awarded once every three times the behavior is performed. Or it can be given every five times. The procedure in which the reinforcement is not provided until the behavior has been performed a number of instances is called fixed-ratio schedule. Fixed-ratio schedules that involve two responses per reinforcement are named FR2, those which involve three are named FR3, and so on.

In the current experiment, Skinner’s method was reproduced using similar apparatus and procedures. A rat was conditioned to respond with a specific behavior (moving the lever) in order to obtain the reinforcement (getting water) and then it was induced to operate under FR2, and then under FR3, and in this pattern until reaching FR5.



For this experiment, ten non naïve, but new to this type of experiment (used only by Kim (2003) in the Morris Water Task experiment), male Sprague-Dawley rats raised in the Macalester College Animal Facility were used. The rats were 114 days old at the time of the experiment. The animals had been taken care of by Animal Facility Staff, having food available ad libitum, and deprived from water for 24 hours prior to the experiment.


Ten Coulbourn Instruments Large Modular Test Cages constituted the test chambers and were used to induce expected behavior on the rats. The chambers were equipped with a tiny dipper to drink water from, located in a receding box at the center of the front wall, and a retractable lever (Med-Associates) protruding at the right corner of the same wall, at about 2 inches from the ground. The sensors attached to the lever were adjusted to the estimated power that a rat could exercise.

The rat was placed in the chamber and locked. For every movement the rat performed that was likely to increase the chances that the rat might push the lever, it was rewarded with a visual stimulus (light in the box) and a tiny portion of water in the dipper. The rewarding process was automatized by using an external button, which was pressed by the experimenter. The actions rewarded included such movements as nose sniffing, standing on the wall, walking towards the lever. Those actions that did not involve location displacement, such as head turning (towards the lever), were not compensated unless the rat had become extremely inactive. Once if a particular response was rewarded too often, that response was dropped from the reinforced responses list and replaced with a more behavior more closely resembling to moving the lever.

Once the rat became conditioned, fixed-ratio scheduling was induced. After the rat began hitting the lever on a regular basis, one response was ignored and reinforcement was awarded only for a second response. This way the rat was conditioned to FR2. As rat’s response rate became stable for FR2, occasional FR3 reinforcements were mixed with FR2 reinforcements until the reinforcements were mostly FR3. This was to be repeated until reaching FR5.

Results and Discussion

Our experiment group does not have records of the conditioning latency because time was not measured. During the basic conditioning phase, a total of 67 reinforcements were awarded. A total of 15 reinforcements were awarded by the computer once it switched to the automatic mode. The subject was rather slow in acquiring the behavior, especially compared to the other rats in the colony which acquired the behavior. This can be attributed to the fact that at the beginning the experimenters didn’t know the precise instructions and administered the reinforcement as a lure to get the rat moving and not as a controlled conditioning apparatus.

During the shaping process, the rat displayed mostly the behavior of pecking at the ground bars, moving around, and standing up. Sudden turns of head toward the lever, movement towards the lever, and standing on the wall were reinforced behaviors in the initial shaping stage. In the middle of the experiment, the rat moved into the center of the box, encroached its body around its head, and stayed still on the ground for a long time. Experimenters hit the glass wall repeatedly to put the rat back into movement. After this, a more strict reinforcement pattern was imposed on the rat, meaning that less responses were reinforced. This proved to work and the rat reached FR1 after a few dozen trials. After 15 computer reinforcements, the behavior was no longer observable, probably because the rat was not thirsty anymore.

A week later, the same rat was put back in the box and was re-trained to FR1. The time required to be conditioned seemed less than the first time, and it was probably due to spontaneous recovery. Once the FR1 was stable, FR2 was induced. During the initial stages of FR2, the rat pushed the lever, put its head in the box (as it did before, during FR1), and usually it came back immediately to the lever to push it again. This way the rat was conditioned to FR4, but response for FR4 was not stable, nor were the occasional introductions of FR5 by the experimenter responded. The rat was not conditioned to FR5 during the entire period of laboratory time, which lasted 90 minutes.


Kalat, J. W. (2002). Introduction to Psychology. Pacific Groove, CA: Wadsworth
Kim, Y. (2003) Effects of cholinergic receptor blockade in place learning and cued versions of the Morris Water Task. Assignment for Introduction to Psychology Laboratory.







Leave a Reply

Your email address will not be published. Required fields are marked *