In my humble opinion, one of the goals of psychology is to gain understanding. Two ways to test or demonstrate one's understanding are prediction and control. For example, an astronomer might say that she understands the motion of the planets in our solar system because of her ability to acurately predict the positions of Earth and Mars in orbit around the sun. On the other hand, we might say that a physicist understands the sub-atomic interactions of various particles because of his ability to control those interactions to produce nuclear power.
Similarly, behavior analysts attempt to predict the effects of various environmental contexts (i.e., schedules of reinforcement or punishment) on the behavior of organisms. The degree of control is inferred by the cooccurrence of systematic changes in the dependent variable with changes in the independent variable. In other words, the term control is used when the experimenter is able to "turn on" and "turn off" an effect.
The procedure that an experimenter follows to demonstrate control is typically called an experimental design (See Chapter 2 of Powell, Symbaluk, and MacDonald for descriptions of a bunch of designs). The success of an experimental design rests in its ability to convincengly demonstrate experimental control. In doing so, they can help to answer our research questions.
In general, an experimental design should be taylored to address a specific question. However, some designs have become widely used and are frequently described in textbooks (maybe because they are able to address a wide varity of questions). One such design is an ABAB reversal design. This design contains two conditions, referred to as A and B. Each subject is exposed to each condition until the effects of that condition have stabalized. Then the subject is exposed to the next condition in the sequence, and so on, and so forth, until A and B have both been experienced twice. Hence the name, ABAB reversal, because the subject is exposed to A, then B, then the conditions reverse back to A, and then to B again.
The purpose of this assignment is to use an ABAB reversal design to demonstrate that lever presses tend to occur most often when they are followed by food compared to when presses are put on extinction (no programmed consequence).
2. Close the "Operant Associations" window. It's the one residing in the upper right-hand corner of the screen and is highlighted in the picture below. Email your TA if you want to know why he thinks you should close this window.
Figure 1. The Operant Associations
window.
3. Open the data file you created from assignment two. Basically, you should have saved it when Sniffy emitted enough responses to reset the cumulative record at least twice. If you didn't save it, go back and do assignment two again (you'll need that data file for future assignments too so you might as well do it again).
Just to recap, by now Sniffy should be pressing
the lever quite frequently. Currently, whenever Sniffy presses the lever,
a pellet is delivered. This contingency is referred to as fixed-ratio
1 (FR-1) or continuous reinforcement because one and only one response is
required for a pellet to be dispensed. We can conceptualize this contingency
as the A condition of our experimental design. In the next step, you'll
move to the B condition (extinction). During extinction, lever presses
will not produce pellets and we would expect lever pressing to drop out.
5. Click on "Experiment" located on the menu bar at the top of the screen. Then click on "Design Operant Experiment..." A window will appear with four options on the left. Click on the last option , "Extinction" then click "OK." Now, pellets will no longer be delivered contingent on lever presses and you should see the words "Extinction (Muted)" appear in the cumulative record.
6. Sit back and watch. At first, you'll notice Sniffy alternate between pressing the lever rapidly and grooming or exploring. Slowly, you'll start to notice less and less lever pressing and that will be evident by a near-horizontal line on the cumulative record. Keep Sniffy in this extinction condition until his responding is stable. Basically, there should be no more than one reset per four intervals (intervals are represented by the vertical lines on the cumulative record). If you keep Sniffy in extinction for too long, it will be difficult to get it to press the lever again. You've now finished the first B condition in our ABAB design!
7. Now that responding has decreased substantially, it's time to reverse back to our A condition. That means that pellets will be delivered according to a continuous or FR-1 schedule of reinforcement. To do that, click on "Experiment" located on the menu bar at the top of the screen. Then click on "Design Operant Experiment..." A window will appear with four options on the left. Click on the second-to-last option, "Continuous" then click "OK." Now the FR-1 is in effect and you should see the letters "CRF" appear in the cumulative record.
8. Sit back and relax again. Sniffy should begin to press the lever more and more. If you want, you can try shaping him again, or you can do what I would do: just wait. Continue this condition until the cumulative record resets at least four times.
9. Repeat steps 5 and 6 to finish the second B condition.
8. Save your data file!!! Make sure you name it something different (e.g., "assign2").
Your grade will be determined by the degree to which your cumulative record maps on to the instructions above. Specifically, I'll be looking to see if you've met the criteria that were underlined.
Once printed out, I should be able to see the following:
1) Sniffy's acquisition (this should have been completed during assignment 2 but it also needs to be turned in for assignment 3).
2) A transition to extinction, denoted by the words "Extinction (Muted)" or "Extinction (Unmuted)" accompanied by a decrease in responding such that there is no more than one reset per four intervals(intervals are represented by the dashed and solid vertical lines on the cumulative record).
3) A transition to FR-1 or continuous reinforcement, denoted by the letters "CRF" or "FR-1" accompanied by an increase in responding such that the cumulative record resets at least four times.
4) And finally another transition to extinction as described in number two just above.