Introduction

    Learning may be defined as a relatively permanent change in behavior resulting from experience. This is a broad definition, which includes phenomena such as habituation, classical conditioning, operant conditioning, and modeling. In our laboratory work, we cannot hope to cover all of these topics. Instead, the laboratory will concentrate on operant conditioning. More specifically, the laboratory will concentrate on a particular approach to operant conditioning called behavior analysis.

    What is behavior analysis? Behavior analysis is an approach to experimental psychology originally proposed by B.F. Skinner. For our purposes, behavior analysis has two defining characteristics:

1) Behavior analysis involves the relationship between environmental events (sometimes called stimuli) and behavioral events (sometimes called responses). This is not to say that behavior analysis is "stimulus-response psychology" - it is not. However, behavior analysis does concentrate on the way that various types of stimuli control various types of behavior.

2) Behavior analysis is based entirely on observable events. When a behavior analyst is asked "What caused that behavior to occur" the answer is always given in terms of what can be directly observed.

Consider the following examples: A rat is deprived of food for 24 hours, and is then placed on a table top (an environment the rat has never been exposed to before), a short distance from a pile of food. At first, the rat does not move toward the food, but remains very still. Slowly, the rat moves toward the food, and eventually begins eating the food. The question is "Why did the rat behave the way it did?"

To a non-behavior analyst, the answer is simple: The rat was frightened when initially placed on the tabletop. However, the rat was also hungry, and slowly its hunger overcame its fear, leading it to eat.

You might be surprised to hear that the behavior analyst rejects this analysis. This is because events like "being frightened" and "being hungry" are not directly observable. "Hunger" and "fear" are things which go on inside the rat's head, and since we cannot "get inside the rat's mind" the behavior analyst chooses not to talk about these things.

How might the behavior analyst answer the same question? He or she might say "If a rat is given food in a novel environment following a period of food deprivation, the rat will initially freeze. It will then slowly approach the food, and begin consuming the food." What the behavior analyst has done is answer the question in terms of the environmental events present (food deprivation and presentation of food in a novel environment) and the behaviors which result (freezing followed by slow approach and consumption).

In the lab exercises which follow, you will use the principles of behavior analysis and operant conditioning to train a rat. As you progress, it is very important to remember the above definitions of behavior analysis. In particular, each exercise will ask you to describe what the animal is doing. Remember to only use observable stimulus events and behavior events when you answer these questions. Avoid saying "the rat knows" or "the rat thinks" or "the rat feels," etc, because these explanations are based on unobservable events going on in the rats head.

Types of Behavior Analysis

Behavior analysis can be further divided into two areas: The Experimental Analysis of Behavior, and Applied Behavior Analysis. The experimental analysis of behavior has traditionally employed animals as subjects, although experimental work on humans is increasing. The purpose of the experimental analysis of behavior is to explore the basic principles through which behavior occurs as a function of environmental variables.

Applied behavior analysis, in contrast, is the application of basic behavioral principles to solve human behavior problems. In other words, applied analysis takes experimentally derived principles and uses these principles to solve socially meaningful problems.

An example may clarify this point. Workers in the experimental analysis of behavior have discovered that reinforcers can control the behavior of rats. That is, rats will perform work when they are "rewarded" by reinforcers. An applied behavior analyst might be working with a mentally retarded client who refuses to brush his teeth. The applied analyst might give the client a reinforcer (a game to play with, for example) if the client brushes his teeth. Thus, the applied analyst has taken an experimental finding about reinforcement and applied it to a socially meaningful problem.

Graphing Conventions

Both the experimental analysis of behavior and applied behavior analysis rely heavily on data. Often, data are easier to understand if they are presented in a visual, graphical manner. Thus, one purpose of the laboratory exercises is to learn techniques of graphic data collection. In each exercise, you will be asked to graphically display your data. Part Three of this manual discusses the proper techniques for graphically presenting data. This section will be useful for graphing the results of your projects.


Due Dates and Laboratory Grades

In general, you will be allowed to work through the laboratory manual at your own pace. However, there are some time limitations. You may hand in any laboratory exercise as soon as you complete it. However, you must turn in each exercise by a specific date. Keep track of the following dates:

Lab 1 due
Lab 2 due
Lab 3 due
Lab 4 due
Lab 5 due
Lab 6 due
Lab 7 due
Lab 8 due
Final write up due

Grading Critiera:

Each laboratory exercise 1 through 8 will be worth 10 points. The final lab write up will be worth 20 points. Labs turned in late will have 2 points subtracted for each day late.

Although you work in groups, EACH student must turn in his or her own lab assignment. Obviously, you may share the data, but you must include your own version of the write up for each lab.

When you complete a laboratory exercise, put it in the box in the laboratory. All labs must be placed in the box no later than 8:00 PM on the due date. Labs turned in elsewhere will not be graded. Labs which are turned in late will have 2 points deducted. The following criteria will be used for grading the laboratory assignments:

Data recorded neatly and correctly 2 pts
Graphs drawn neatly and correctly 2 pts
Questions answered neatly and correctly 2 pts
Questions answered using objective behavioral language 4 pts

To get maximum credit, answer the questions carefully and concisely. Graph your data neatly, using the correct graphing conventions described in Part Three of the manual. Most importantly, restrict your discussions to directly observable stimulus and response events (that is, use "behaviorese" when discussing your rat's behavior.

Laboratory Attendance

You are NOT required to attend lab every day. However, you will need to attend lab often enough to complete the projects. Some projects will take longer than other projects. Use the following guidelines to plan your laboratory attendance:

Project Days Required

1 3 (approximate)
2 5 (approximate)
3 3 (maximum)
4 7 (maximum)
5 5 (maximum)
6 5 (maximum)
7 7 (maximum)
8 7 (maximum)

Thus, a total of 42 lab sessions are necessary to complete the assignments. In order to encourage planning, any student attending lab less than 42 times will have 10 points deducted from his or her final lab grade. The only exception is that students completing and handing in all labs BEFORE the May 1st deadline are NOT required to attend all 42 sessions.

Note that some of the labs are labeled with approximate requirements, others with maximumrequirements. For approximate labs, the number refers to the number of days it should take you to complete the lab. If you have not completed the lab in the estimated time, continue with the project. This is because you must complete these projects to complete the remaining projects. If you are taking longer than the estimated time, be sure to keep your TA informed of your progress.

For labs labeled with maximum requirements, you should run no more than the estimated number of days, even if you have not technically completed the assignment. For example, suppose you are doing Lab 4, and your rat still has not reached the discrimination requirement after 7 sessions. At this point, hand in your lab assignment and go on to the next assignment. You will not be penalized for this, but you will need to discuss (on your assignment sheet) why you think your rat failed to meet the criterion. It is also possible that you will finish before the required number of sessions. In this case, go on to the next assignment.


Laboratory Conduct

The guidelines below should be followed in all of your laboratory work:

1. Read the assignment before you come to lab. If you do not understand the assignment, ask you lab assistant for clarification before you begin working.

2. Check to see that your equipment is working before you start your lab. If the equipment is not working properly, report it immediately to your lab assistant.

3. When the equipment is ready, go to the animal colony to get your rat. Be sure to weigh your rat every time you work with it. If your rat has lost a substantial amount of weight (10 grams or more), it could mean that the rat is sick. Report this immediately to your lab assistant or to me.

4. It is probably easiest to take your rat's home cage into the laboratory with you. If you do this, be sure that you do not squash the rat's feet when you put the cage down. When you take the rat out of the cage, it will often grip the bottom of the cage. If this happens, pull very gently on the rat. To not "yank" it out of the cage - you might pull off the rat's toenail in the process.

5. Once you have picked up your rat, put it in the cage fairly quickly. While it is a good idea to handle your rat from time to time, keep handling at a minimum prior to the experimental session.

6. Keep noise to a minimum in the laboratory. Generally, you should not talk, or only talk in whispers. If you need to talk for a long time, go into the hallway. There is nothing more frustrating that to have your rat close to mastery on a task, only to have the rat freeze in the corner because someone is talking loudly.

7. When you have completed your laboratory, have your lab assistant sign the top of your data sheet. This is to assure that you have really been in lab doing the work.

8. Some laboratories will take a few days, others will take a week or more. If your laboratory seems to be taking a long time, bring this to the attention of your lab assistant. He or she will probably be able to help you make progress.

9. You are not required to come to lab every day, but it is highly recommended that you do. The rat will learn much faster if you work with it on a regular schedule, than if you are sporadic in your lab work. Be sure to pay attention to the due dates for each laboratory.

10. During all of the laboratories, you will be using drops of water as reinforcers. To make water an effective reinforcer, you must first deprive your rat of water. To do this, we use an 18 hour minimum deprivation schedule. The lab technician, the TA's and I will maintain the 23 deprivation for you. EACH night the water bottles will be placed on the rats for ½ hour at approximately 4:30 PM. DO NOT PLACE THE WATER BOTTLES OR REMOVE WATER BOTTLES WITHOUT EXPLICIT PERMISSION FROM DR. FARMER-DOUGAN OR BOB BOMA.

11. In general, the laboratory is not open on weekends. Each Friday, water bottles will be placed on the rats at approximaely 4:30 PM. The water bottles will be removed at approximately 5:00 PM Sunday night, so you can work with the rat on Monday.

12. Remember that your rat is a living creature. Treat it with respect. If you mistreat your rat, it will be taken away from you, which will make it very difficult to finish the lab assignments and pass the course.

Part I: Laboratory Experiments

Lab 1: MAGAZINE TRAINING

This is the first of several assignments involving the operant conditioning of your rat. That this first assignment is done properly is of utmost importance since the following assignments can only be completed satisfactorily if your subject is well trained to the magazine. This point cannot be over-emphasized.

Before you can train your rat to do anything, the rat must learn where the reinforcer is located and what sounds are associated with reinforcer delivery. This process is called "magazine training." In this particular exercise, magazine training consists of training the animal to go to the water cup (also called a "magazine") from any place in the chamber when water has been delivered. When the animal has learned to go to the cup immediately after the sound of the water magazine being operated, other behaviors are easily conditioned using water as a reinforcer.

Procedure: Choose a type of box and stick with it the rest of the semester. The old boxes are louder, but easier to use for both you and the rat. The new boxes are quieter, but more complicated and the water often gets stale!

1. Check the equipment to make sure it is working.

(A) OLD BOXES: Turn the power on. Place the program selector switch in the UP position (EXP). Switch the Reinforcer switch to MANUAL. Clear the Reinforcer and Response counters. Finally, turn the stimulus light on. Set the magazine timer (on the back panel of the control console) to its maximum setting. Fill the water reservoir and return it to its position.

(B) NEW BOXES: Turn the power on. Set the following settings on the box: Norm (Bar), Liquid. Clear the response and reinforcement counters. Set the MODE to CRF and set the Ratio/interval counter to 0. Be sure to set the liquid dispenser switch (on the box itself) to the highest setting.

2. Prepare a data sheet as outlined below.

3. Take the rat from its home cage and weigh it. Place it in the experimental cage with one drop of water already in the cup. You will probably want to manually prop the water cup in place for this first reinforcer.

4. Wait until the rat drinks from the water cup, then leaves it.

5. After the rat has moved away from the cup, operate the hand switch (on the box) so that water is again delivered to the cup. Wait until the rat drinks and moves away from the cup. Repeat this step 5 times. If it looks like the rat will not drink before the water cup is lowered, manually prop the water cup in place, as in step 2.

6. Record the time between operation of the hand switch and the initiation of drinking by the rat. This time is called the response latency.

7. Continue repeating step 4, but now avoid reinforcing the animal when it is near the cup or hovering over it. That is, wait until the rat is some distance from the water cup before operating the hand switch.

8. Continue magazine training and continue recording responses until the rat responds to the magazine within 3 seconds of operation for 5 consecutive deliveries. In each of these cases, the rat's head should be at least 3 inches away from the water cup before you operate the magazine.

9. When you finish your session, return your rat to its home cage and clean the experimental apparatus. If the rat is still not magazine trained after two lab sessions, ask me or the TA for extra help.

10. When you think you have achieved the criterion, ask me or the TA to watch your rat to confirm that it has learned the response. Have the TA sign the top of your data sheet to acknowledge this. Complete the data sheet and have it signed my me or the TA.

Precautions:

Take care not to "frighten" your animal (note that fear is not an observable behavior - hence the quotation marks around "frighten"). Do not slam the lid on the apparatus or make any unnecessary noise. A "frightened" rat will typically huddle in the corner of the chamber, and you will have to wait until the animal is moving about before you can even start. This can significantly increase the time it takes to magazine train the animal. NOTE: The water magazine is very loud (especially on the older boxes), and will likely "frighten" the rat at first. There is no way around this problem, but you should expect it to occur.

Always wait until the animal leaves the water cup before operating the hand switch again. Over trials, require the rat to be farther and farther away from the water cup before you operate the hand switch. If you do not require the animal to move away, it will start spending all of its time with its head in the water cup. This can seriously interfere with later attempts to condition new responses.

After about 50 drops of water have been delivered, the rat will become satiated. This means that water will no longer function as a reinforcer. When the rat approaches the satiation point, it will stop moving toward the magazine. Keep a running count of the reinforcers you deliver, and stop for the day when it is clear that satiation has occurred. Use your reinforcers wisely!

LAB 2: SHAPING THE BAR PRESS

Now that you have conditioned your rat to go to the water cup whenever the click occurs, you can train your animal to perform a new response by reinforcing with water. The new response will be pressing the bar or lever at the front of the chamber. You will observe that frequency of bar pressing will increase when the response is reinforced.

Bar pressing may be conditioned in two ways. You might place the animal in the chamber and wait for a bar press to be emitted, reinforcing that response when it occurs. This will probably take a long time, and it is possible that the rat will never press the bar. On the other hand, you can shape the bar pressing response by using successive approximations.

The method of successive approximations works like this: Since the rat is unlikely to press the bar immediately, you begin by reinforcing something which the animal is likely to do which is an approximation of bar pressing. That is, you choose a response to reinforce which the animal does fairly often and which is compatible with or even a component of the bar pressing response. When the animal is reliably performing the approximated response, you change the criterion and require an even closer approximation of bar pressing. You then require an even closer approximation, and so on, until the animal actually is barpressing.

For example, you might start by presenting a reinforcer every time the rat "looks" at the bar (in technical language, we say the rat orients to the bar). This is an approximation to bar pressing because most rats will orient to the bar prior to pressing it. It is also something the rat is likely to do, so you will have plenty of opportunities to reinforce it. After the rat is reliably orienting to the bar, you can change the criterion by requiring that the rat actually touch the bar with its nose. Then you might require the rat to put its paws on the bar, followed by requiring the rat to press down slightly on the bar. Eventually, you will require the rat to actually press the bar. As you can see, the method entails reinforcing responses which are initially quite unlike the desired response, then moving closer and closer to the desired response until it is achieved.

Procedure:

1. Deprive your rat of water for 23 hours before you plan to start.

2. Check the equipment to make sure it is working.

(A) OLD BOXES: Turn the power on. Place the program selector switch in position UP (EXP). Switch the Reinforcer switch to MANUAL. Clear the Reinforcer and Response counters. Finally, turn the stimulus light on. Set the magazine timer (on the back panel of the control console) to its maximum setting. Fill the water reservoir. Prepare a data sheet as outlined below.

(B) NEW BOXES. Turn the power on. Set the following settings on the box: Norm (Bar), Liquid. Clear the response and reinforcement counters. Set the MODE to EXT and set the Ratio/interval counter to 0. Be sure to set the liquid dispenser switch (on the box itself) to the highest setting.

3. Weigh your rat. Place the rat in the experimental chamber for 5 minutes, and record the number of bar presses occurring during this time (probably 0). This is called the subject's operant level or baseline rate of bar pressing. During these first 5 minutes, do not reinforce the animal for any response.

4. After the 5 minute period, begin the shaping procedure. On the OLD boxes change the program selector switch to PARALLEL or CRF setting. (This enables the response lever so that reinforcers will be delivered when the rat presses the bar). On the NEW boxes set the MODE to CRF. Deliver reinforcers (by pressing the lever ON the control panel of the old boxes, or depressing the switch on the cord of the new boxes) for successive approximations of the bar pressing response. Continue to record the number of bar presses which occur every 5 minutes, but do not record the number of reinforcers.

5. Once you have shaped bar pressing, allow it to continue until there have been at least 20 reinforced presses and the rat is pressing the bar at least 5 times per minute. Show the rat to your TA and have her sign your lab sheet, confirming that your rat is indeed pressing the bar.

6. Return your rat to its cage. Give the rat access to water for one hour before removing the bottle again. Complete your lab assignment sheet and hand it in to your TA.

Precautions:

Be sure not to reinforce behavior which will be incompatible with bar pressing. For example, turning away from the bar is incompatible with pressing the bar, so don't reinforce it.

When you change your criterion, don't make too big a change. And, if the rat seems not to be making the transition, move back a step in your chain of approximations. Otherwise, extinction can occur. Extinction is a reduction in response rate which occurs when a previously reinforced response is no longer reinforced. Suppose your rat is reliably looking at the bar. You decide that the rat should be required to put its paws on the bar next. But this may be too big a step: The rat may not put its paws on the bar for several minutes. All the while it is continuing to look at the bar, but you are no longer reinforcing "looking" - as a result, "looking" may extinguish.

By far the most difficult transition to make is from "paws on bar" to "pressing bar" - it is easy to "lose" the animal at this point and have extinction occur. It is best to use extremely small steps here. Require the rat to go from "paws on bar" to "pressing bar 1/4th of the way" to "pressing bar 1/2 way" etc. This way, the animal should be able to master the response without extinction occurring.

Lab 3 : CONDITIONED REINFORCEMENT

Up until this time, you have been concerned about the reinforcing effects of water on your rat's behavior. Traditionally, a reinforcer has been defined as any stimulus which, when presented immediately following a response, results in an increase in the probability or frequency of that response. When you make water contingent (or dependent) on bar pressing, the probability of bar pressing went up. With this finding you demonstrated that water is a reinforcer.

Water is called a primary reinforcer because it reinforces behavior without prior training. We might also call it an unconditioned reinforcer. It is possible that the rat encounters stimuli in the chamber that are not initially reinforcing, but became so during conditioning. Any stimuli in this category are called secondary reinforcers, or sometimes conditioned reinforcers.

The present experiment evaluates the reinforcing effects of the magazine click. You will investigate the possibility that the click has become a conditioned reinforcer due to its having been paired with the primary reinforcer, water.

Try to analyze the procedure, as you read it, to see why these steps are suitable for investigating this problem. Can you think of any other ways to investigate the conditioned reinforcing effect of the magazine click?

Procedure:

Important! Allow your rat to bar press for two days between the end of Lab 2 and the beginning of lab 3. This will strengthen the response, and make it less likely that you will need to reshape bar pressing later. You do NOT need to collect data during this time.

1. Deprive your rat of water for 23 hours before you plan to start.

2. Check the apparatus to make sure it is working.

(A) OLD BOXES: Place the programming switch in the PARALLEL or CRF position, and turn the stimulus light on.

(B) NEW BOXES: Set the MODE to DISC (discrimination). Use the shaping button (on the cord) to turn on and off the light. Water will be delivered for barpresses ONLY during the LIGHT-ON period.

3. Prepare your data sheet (below).

4. Place the rat in the experimental chamber. Reinforce every bar press for 3 minutes, recording on your data sheet how many bar presses occur in one minute intervals.

5. Extinction Period.

(A) OLD BOXES: Turn the programming switch to EXP and take the water tray out from under the dipper so that no water is delivered in the dipper. Do this as quietly as possible so you do not disturb your rat. Continue to record the number of bar presses which occur, but do not reinforce bar presses. This is an extinction schedule, so your rat should eventually stop pressing the bar.

(B) NEW BOXES: Turn the MODE to EXT.

6. When 3 successive minutes have passed with no bar presses, turn the programming switch back to PARALLEL (CRF) on the OLD boxes so that the dipper will now operate when the bar is pressed, but it will not contain water. For the NEW boxes, set the liquid dispenser knob (ON the box itself) to its lowest setting. Leave the box in the DISC mode and turn the light on. This will click as if it is giving a reinforcer, but nothing will be delivered.or turn the liquid dispenser back to the highest setting.

7. Continue recording bar presses in 1 minute intervals. After your rat has made at least 1 bar press in this condition, wait until it goes 2 full minutes with no bar presses. At that time, the extinction process is complete. Show your rat to the TA and have them sign your lab sheet.

8. Return your rat to its home cage and give it a 1 hour access to water. Complete your data sheet and hand it in.

Lab 4 : DISCRIMINATION TRAINING

You will now bring the bar press under stimulus control by training the animal to respond while the light is on and not respond while the light is off. A discriminative stimulusor SD is a stimulus in the presence of which a response is reinforced. The light will be the SDin this experiment. The Sdelta is the stimulus in the presence of which responses are notreinforced. The light-off condition will serve as the Sdelta here. Once the response occurs rapidly in the SD condition and not at all in the Sdelta we say that stimulus control has been established.

Procedure:

1. Check the apparatus and make sure it works.

(A) OLD BOXES: Put the programming switch in the PARALLEL or CRF position, and turn on the cue light. Get your rat and weight it.

(B) NEW BOXES: Set the MODE to DISC and turn on the cue light. Get your rat and weigh it.

2. Place your rat in the box. The box is set so the light is on and responses will be reinforced. Thus, this is the SD period. Begin timing the response latency (the time before the rat presses the bar for the first time). Record the latency on your data sheet.

3. When the bar press occurs, turn off the cue light, and, on the OLD boxes, turn the programming switch to EXP. This is the beginning of the Sdelta period: do not reinforce any responses, but record any bar presses which occur.

4. After 30 seconds have elapsed without a bar press occurring, turn the cue light on again, and, for the OLD BOXES, turn the programming switch back to PARALLEL or CRF position. Repeat steps 2 - 4.

5. Continue alternating SD and Sdelta (steps 2 - 4) until response latencies in SD are less than 5 seconds and you rat goes through 5 consecutive Sdelta periods without a response. When this happens, the rat is under good stimulus control.

Lab 5: SCHEDULES OF REINFORCEMENT AND

RATIO STRAIN

Up until now, you have been reinforcing your rat for every response it makes (except for the extinction conditions). There are some disadvantages to this, though. For one thing, you rat will satiate after 30 - 50 reinforcers. Thus, you only have a short period during which reinforcement is effective.

It is not necessary, however, to reinforce your rat every time it makes a response. The rat will continue to bar press at high rates even if you reinforce bar pressing only occasionally. But how do you determine which bar presses to reinforce and which to not reinforce? This is determined by the schedule of reinforcement. Schedules of reinforcement are simple rules which determine which bar presses are reinforced and which are not.

Schedules of reinforcement come in a number of types. Here, we will concentrate on Fixed Ratio (FR) schedules, which produce reinforcement after a certain set number of responses have been emitted. For example, on an FR 10 schedule, the rat will be reinforced for every 10th response it emits.

Animals responding on Fixed Ratio schedules typically exhibit a post-reinforcement pause. After each reinforcer is delivered, the animal typically waits for a period of time before it resumed responding. Then, once it starts again, it typically completes all of the required responses fairly quickly.

FR schedules can have another interesting property called ratio strain. Basically, ratio strain occurs when the value of the fixed interval gets too large and the animal quits responding. You might expect a rat to respond 10 times for a reinforcer, or 20, or even 50.....but would it respond 1000 times? Probably not. Instead, it would probably reach the point of ratio strain and stop responding.

For this experiment, you will probably want to have a watch with a second hand or stopwatch available. The lab will take several days to complete.

Procedure:

1. Check the apparatus and make sure it works.

(A) OLD BOXES: Put the programming switch in the SERIES position, and turn on the cue light.

(B) NEW BOXES: Put the MODE switch in the FR position.

2. Get your rat and weight it, and put it in the box.

3. If possible, complete steps 2 - 6 on the first day. We will begin the lab with a Fixed Ratio 2 (abbreviated FR 2) schedule. On this schedule, every second response is reinforced.

(A) OLD BOXES: When you first place the rat in the box, do not depress the button on the hand switch. This will mean that responses will not be reinforced. After the rat has made its first response, depress the handswitch button and hold it in. Now the rat will be reinforced for the next response it makes. Using the handswitch in this way, reinforce every 2nd (i.e., every other) response.

(B) NEW BOXES. Set the ratio/interval counter to 2. This will reinforce the rat on every second bar press.

4. During this first day of the experiment, you will be gradually working your rat up to a longer ratio value. You will also need to keep track of the post- reinforcement pause. To do this, start a timer at the end of reinforcement delivery (i.e., when the dipper retracts). When the animal makes its first response, you should stop the timer and record the time on your data sheet. Continue doing this for every reinforcer you deliver.

5. Continue steps 2 and 3 until the rat has been reinforced 10 times. Then, change the requirement so that only every 4th response is reinforced. This is an FR 4 schedule. Keep the FR 4 schedule in effect until the rat has been reinforced 10 times.

6. Change the schedule to an FR 8 schedule. That is, reinforce the rat for every 8th response it makes. Again, keep the FR 8 in effect until the rat has been reinforced 10 times.

7. Change the schedule to FR 16. That is, reinforce the rat for every 16th response it makes. When the rat has received 10 reinforcers on FR 16, return it to its home cage, and start step 7 on the next day. <It is possible that you will reach ratio strain before this. If you have, skip to step 8 and continue from there>.

8. You should start this step on a new day if possible. Begin with an FR 20 schedule. After 10 reinforcers are earned, add five responses to the requirement (i.e., it should be FR 25). Give 10 more reinforcers, and five more responses to the requirement. Continue doing this until the animal stops responding (i.e., it reaches ratio strain) or until the end of the lab period.

9. You should start this step on a new day. If your rat reached ratio strain in an earlier step, change the requirement to half of the value you were at when the rat showed strain. That is, if the rat showed strain at FR 30, decrease the requirement to FR 15. Then, continue as in step 7, increasing the requirement by five responses after every 10th reinforcer is delivered. Continue until you reach ratio strain again, or until the end of your lab period.

If your rat has yet to reach ratio strain at the start of step 8, start where you left off at the end of step 7 and continue adding 5 responses to the requirement after every 10th reinforcer. Continue until the rat shows ratio strain, or until the end of the lab session.

10. Repeat step 8 for two additional days.

Lab 6 : EXTINCTION REVISITED

During the first part of laboratory 3, you exposed your animal to an extinctionschedule. That is, you no longer reinforced the bar pressing response. As a result, the bar pressing response ceased occurring. During lab 3, extinction was probably fairly rapid. Most students report extinction to occur within the first session.

In the most recent experiment (lab 5) you exposed your rat to several schedules of reinforcement. The schedules you used are collectively called partial schedules of reinforcement because only some responses were actually reinforced. Partial reinforcement schedules may be contrasted with continuous schedules of reinforcement in which all responses are reinforced.

Numerous experiments had shown that partial schedules of reinforcement have some unique effects on subsequent extinction schedules. In particular, rats exposed to extinction training following exposure to partial schedules will take longer to extinguish than rats who have only been exposed to continuous schedules. This is called the partial reinforcement extinction effect (or PREE for short).

Spontaneous recovery is another interesting phenomenon which occurs during extinction schedules. Suppose you expose your animal to an extinction schedule, and it ceases responding during that first session. You return your rat to the home cage overnight, and then carry out extinction again the next day. What will happen at the start of the second day? Typically, the rat respond at the start of the second day even though it had completely ceased responding on the previous day. This resumption of responding following a rest period is called spontaneous recovery.

In the present exercise, you will examine both spontaneous recovery and the PREE. The exercise will take 5 days to complete.

Procedure:

1. Check the apparatus to make sure it works. Put the control switch on SERIES position for the OLD BOXES, or FIXED RATIO for the new boxes. Weigh your rat and put it in the experimental chamber.

2. For the first 5 minutes after the beginning of the session on the first day of training only, use an FR 5 schedule. Record the number of responses which occur each minute on your data sheet.

3. At the end of 5 minutes, begin the extinction schedule. On the OLD boxes you do not need to change anything on the apparatus. Simply stop delivering reinforcers. On the NEW boxes, switch the mode to EXT.

4. Continue the extinction schedule for 30 minutes, or until the rat stops responding for 5 consecutive minutes. Throughout this time, record the number of responses which occur during each minute.

5. After the session is over return your rat to its home cage.

6. Repeat steps 4 and 5 for the next 4 days.

Lab 7 : SHAPING A COMPLEX RESPONSE CHAIN

With this lab, you come full circle to the problems faced at the very beginning of your lab work. The process of shaping is again the object of investigation, as it was in lab 2. Here, however, you will be shaping a complex chain of behavior. in which several aspects of the environment play a discriminative function. Thus, the lab synthesizes and extends a number of principles you have previously encountered. It also gives you more practice with shaping, which may be the single most important procedure in behavior analysis.

The goal of this lab is to train your rat to mount an inverted tin can, reach down and pick a marble up from the chamber floor, and drop the marble into a hole in the can (see the attached diagram). In addition to providing further practice in shaping, this exercise will illustrate the technique of arranging the environment so that it facilitates the acquisition of a complex repertoire.

It might be possible to produce the performance described above simply by placing the animal in a chamber containing some marbles and a tin can, and then reinforcing successive approximations to the final behavioral sequence. The task will be easier, however, by breaking the total performance down into components, developing each component separately, and then reinforcing the correct sequence of well-developed components. In the present case, the final repertoire can be analyzed into at least two major components: picking up and dropping a marble, and mounting the tin can. The marble-handling will be shaped with the can out of the chamber. When this component is fairly strong, the marbles will be removed, the can will be placed in the chamber, and mounting the can will be shaped. When this behavior is well developed, the marbles will be returned to the chamber. The rat will then be reinforced for picking up and handling the marbles while on top of the can. Finally, the rat will be reinforced only for dropping the marble into the hole.

PROCEDURE

1. Check the equipment. Turn the dial to the EXP position, and turn the stimulus light OFF. Operation of the hand switch should produce reinforcement, but bar pressing should not.

2. Shaping marble handling: Put 5 - 10 marbles on the floor of the chamber. The can should not be in the chamber at this time. Reinforce successive approximations to picking up and releasing the marbles. It makes little difference whether the rat picks up the marble with its mouth or paws, but be careful not to reinforce chewing the marble, rolling it around on the floor, or long-maintained marble holding. Eventually require that the marble be lifted an inch or so off the floor, but only deliver the marble when the marble is dropped after being lifted.

3. Mounting the tin can: Remove all marbles from the chamber. Place the can on floor of the chamber, and reinforce successive approximations to mounting the can, eventually requiring that the rat remain on top of the can for several seconds before it is reinforced.

4. Combining the two components: Place both marbles and the can in the chamber. Reinforce being on top of the can and reaching down to pick up the marble. Once the animal is taking a marble up on the can, the next step is to require that the marble be dropped before reinforcement. At this point, it is possible to make a rather interesting mistake. When the animal is holding the marble, while on top of the can, it will usually drop the marble as soon as reinforcement occurs. In other words, the animal may drop the marble because the reinforcer has just occurred, not because of a direct contingency between response and reinforcement. Take care to insure that the rat drops the marble before reinforcement is delivered.

5. When your animal is completing the entire chain, make sure that your lab assistant sees this and signs your data sheet. Continue the lab until the rat has correctly completed the chain 20 times.

Lab 8: "DO IT YOURSELF" LAB

Up to this point, you have done a series of what are called "canned" labs. That is, I give you a can, you open it, and look at what is inside. The good thing about canned labs is that I know beforehand that the labs work - or at least should work. Unfortunately, canned labs don't really give you an opportunity to show any creativity.

So, with this final laboratory, you can demonstrate your creativity. The assignment is simple: Design and implement your own laboratory project. Virtually anything is acceptable as long as it does not involve making the rat sick or exposing it to pain. Here are some general ideas:

- Teach your rat an unusual discrimination (for example, can a rat discriminate classical music from rock music?)

- Replicate the "superstition" paradigm in your rat.

- Train your rat to do an unusual response of some kind.

- Expose your rat to a complex schedule of reinforcement.

- Expose your rat to a classical conditioning procedure.

These are just some general suggestions. The basic purpose is for you to come up with an experiment that you are interested in personally. You might want to look back through your book or notes to get some ideas. When you have an idea, clear it with me before you actually begin.

Because the projects will all be different, there is no data worksheet for this experiment. To complete this project, you should hand in a short, typed report. Be sure to include:

1) A brief description of what you are planning to do in your experiment.
This must be turned in for approval prior to conducting the project. Use the enclosed sheets from the IACUC committee to write up the proposed project.

2) A brief description of how you are planning to do it.

3) Your data.

4) A Figure showing a graphic representation of your data.

5) A discussion of whether your project was a success, and why or why not.

Part II: Plotting Your Data

Numerical "raw data" (the data which are recorded while the behavior is actually occurring) are rarely of any use just as a set of numbers. Raw data are critical, however, in that these data serve as the basis for all further analysis. Thus, raw data must be as accurate as possible.

Once numerical raw data are gathered, they are usually "manipulated" or "treated" to make them more easily understood. For example, raw data are often averaged, converted to rate measures (i.e., responses per minute) or transformed in some complex way - all to render them more understandable.

Even "treated" data are not usually very useful, though, because when we look at a table of numbers we usually don't understand them very well. Which numbers are important? What do they mean? Which number or numbers do you look at first? Last? What is the relationship between the numbers? It is very time consuming and difficult to make much sense of a table of numbers - to identify trends in those numbers and relationships among those numbers. To assist in identifying trends in the data, a pictorial representation of data - a graphis usually drawn.

A well constructed graph will convey a large amount of information in a small amount of space. Ideally, a person should be able to examine a graph and almost immediately understand what the graph shows. The following guidelines should help you design well- constructed graphs.

1) Although we have been using the term "graph" above, it is actually a technically incorrect term. Instead, we will use the term "figure." When a paper or report has more than one figure, refer to them as Figure 1, Figure 2, Figure 3, and so on.

2) Each figure should have a caption. The caption should include a) at title for the figure, and 2) all of the information necessary to read the figure properly. A well-constructed Figure should require no additional information - that is, all of the information you need to understand the Figure should be found in the caption.

3) If you are going to draw your figure by hand, you should always use a pencil. NEVER use a pen or colored marker to draw your figure. Mistakes are easy to make, and pencil drawings are easily corrected.

4) If you have the software available, and know how to use it, you can use a computer to generate your graphs. Almost all professional "publication quality" figures are now drawn using computers, so this is a good skill to develop if you have the resources available to you.

5) It is easier to read a figure if it is drawn on plain paper than if it is drawn on graph paper. On the other hand, it is easier to neatly draw and label a figure on graph paper. One good solution is to plot your data on the reverse side of a sheet of graph paper. Or, place a relatively thin piece of plain paper over the top of a sheet of graph paper, and plot your data on the plain paper using the graph paper as a guide.

6) Your graph axes must be neatly labelled. In general, it is best to type these labels. If you cannot type them, letter them very carefully using block printing. Never place the axes at the edge of the paper. If you do, you will be unable to label them.

7) Typically, you will be plotting "dependent variables" as a function of "independent variables." A dependant variable is the thing you are measuring in the experiment - for example, the number of bar presses made in a particular amount of time. An independent variable is usually something you manipulate as the experimenter (for example, the presence or absence of the stimulus light). In many of the experiments you will be doing the "independent" variable will be the passage of time or the presentation of a certain number of trials.

When plotting your figures, ALWAYS plot the dependent variable on the vertical y-axis. Independent variables should ALWAYS be plotted on the horizontal x-axis.

8) In many of the figures you will draw in this class, you will need to represent several different conditions or procedures. Typically, a vertical line in the figure represents a change in procedures.  Figure 1 is an example of a "well constructed" graph which has all of the characteristics we have discussed above. Note each of the following:

a) The caption below the figure gives all of the information you need to understand what is being plotted in the figure. As long as you understand the basic vocabulary, you should have little trouble interpreting the figure.

b) This figure has been drawn on a computer using the program SIGMAPLOT. Note how neat the graph looks. You should try to keep your figures looking this neat even when you plot them by hand.

c) The dependent variable in the figure is "rate of bar pressing." Note that this variable is plotted on the vertical y-axis. Also note how the axis is labelled, including the clarification that "rate of bar pressing" is measured in presses per minute.

d) The independent variable in this figure is "minutes." In other words, the experimenter has taken a measure of behavior once per minute across the session. Note how this variable is plotted on the horizontal x-axis.

e) Note that there are two different "conditions" presented. First, the animal is exposed to a VARIABLE INTERVAL (VI) schedule for 8 minutes. Then, the schedule is changed to EXTINCTION (EXT), and back to the VI schedule. Note how the vertical line separates the conditions.

Part III. : Final Write-Up:

Your final writeup should include:

1. A 3-5 page summary of your data. This should be written like a discussion section in a manuscript. Answer the following questions.
a. Briefly restate each lab activity and result that you obtained.

b. Describe why you believe you obtained the outcome you did. Try and tie the outcomes to material from the book and lecture (e.g., "as predicted by Breland and Breland (1966), the rat......).

c. Briefly summarize the overall data. What did you demonstrate, and why do you think I had you do these activities?

d. Describe what you got out of the lab experience on a personal basis. What did you learn? Do you have a better understanding of operant conditioning? What was the most important experience? Why? What was the least important? Why?

2. Any references or resources you used in your summary.

3. Order your lab write-ups and add these to (1) and (2) above to complete your lab portfolio.Part IV. Rules for the LAB