Hey everyone, welcome to Chapter 6. So in this chapter we're going to talk about withdrawal designs. So withdrawal designs are a basic experimental procedure for demonstrating treatment effects and they're widely recognized across disciplines, not just in the field of ABA. We've got a lot of learning objectives in Chapter 6. So identify mechanics of withdrawal designs, describe the procedures for and uses of the ABA design, same thing for ABAV design.
describe the principles of prediction, verification, and replication in ABAB designs, identify the advantages and disadvantages of withdrawal designs, and then describe adaptations of the typical withdrawal designs, including the BAB and AB, AB, AB, AB, I think I got them all there, repeated withdrawal designs. So we're going to cover all of these things throughout our recorded lectures for chapter six. Okay, so we're going to start.
first, which is covering the mechanics of withdrawal designs. All right, so reversal designs are also known as withdrawal designs. So when we say withdrawal, that refers to the withdrawal of treatment during one or more phases of a study.
And this is done to demonstrate the effects that it has on the target behavior. The term withdrawal is the preferred name versus reversal because it describes the mechanics of the design. So again, withdrawal of that intervention rather than the intended outcomes of the design, such as like reversing direction of the target behavior. The intention is to examine the effect of introducing and then subsequently removing or withdrawing the intervention, not to actively reverse the effect of the intervention. When the term reversal is used, typically there's an active attempt to use procedures usually some differential reinforcement procedure to reverse the effects of a treatment.
But true reversal designs are actually pretty rare in professional literature. However, you'll still hear the term reversal design often used to mean the same thing as a withdrawal design. So you're going to hear those interchangeable.
It can also be used sometimes when you hear reversal design to refer to like reversing the focus. of the independent variable or intervention and that target behavior. So the basic goal of withdrawal design is to show that there is a functional relationship between the target behavior and the intervention. The typical withdrawal design that you'll see is designated by the letters ABAB. So if you remember, A, baseline, right?
So we're collecting our data, no interventions been... put in place at this point we simply want to see kind of where we're starting and then B would designate a treatment or intervention phase so that's when we're implementing our intervention or our treatment continuing to collect data to see what happens a we go back to that baseline so we're withdrawing that intervention or treatment continuing to collect data and then be bringing back in our treatment or intervention one of the most powerful things about withdrawal designs is that it really allows the researcher to directly and easily demonstrate a cause and effect relationship between that behavior and that intervention. So this helps rule out variables like history and maturation that we talked about earlier, because they can demonstrate that the behavior change occurred only when the treatment was introduced or withdrawn.
So especially with more replications though. that we have. A researcher might withdraw the intervention to see if the behavior changes toward or returns to that baseline level that we saw before. So that typically is designated by like an ABA design. So we've got our baseline intervention baseline.
However, it's not recommended in educational or clinical settings because it's concluding with a non-treatment phase. We're ending in a baseline phase. So the ABA design shows obviously more experimental control than the simple just A, B, where we've got baseline intervention that we talked about in earlier chapters, but it's still not recommended again since it concludes with that, the subject being in a non-treatment phase. This is particularly true for successful studies since it leaves that subject in the unwanted pretreatment levels with their behavior. Not.
where we saw success during the intervention. So this is why researchers should introduce the treatment to determine whether the pattern of behavior will change again, like we do with that ABAB design, which again is that typical withdrawal design. Withdrawal designs are also the most straightforward and generally the most powerful within subject design for demonstrating that functional relationship between our independent variables or our interventions and treatments and that dependent variable or those target behavior or behaviors.
Okay, ABAB design steps. So ABAB design follows these steps here. We kind of briefly just outlined them in the last one, but we'll go through them here. So first step would be of course to collect your baseline data. You collect data on that target behavior before the intervention is introduced.
And again you're going to see that labeled as like A or A1 on a graph. Second step is introduce intervention for a specified period of time and then collect data. So while you introduce your administrator intervention you are collecting data and seeing if it's making an impact.
Again that typically is designated as like a B or a B1 label on your graph. 3. Withdraw intervention for a short period of time and collect data. So the interventions withdrawn, you want to determine whether that target behavior returns to those baseline levels like we saw during that A or A1 phase, the previous one.
And then 4. Reintroduce the intervention and collect data. So again we want to see if we bring back this intervention and reintroduce it. Is it going to have effects on the target behavior?
All right, let's take a look at some ABAB graphs. All right, so here's this first one. So this graph, data that indicate a functional relationship, data on this graph are going to demonstrate that functional relationship between the target behavior and the intervention. So for this example, of course, my mouse here it is, the goal was to decrease the target behavior.
Okay. So note that the behavior decreased during the presentation of the intervention and then again increased during its withdrawal. So we had kind of a higher level of occurrence of that behavior in baseline levels. We introduce our intervention. We've got our phase change line here and we can see the level pretty dramatically drops.
That's a pretty significant decrease that we saw here in our level. So the level goes down with the behavior. We remove that. Withdraw that intervention to see what happens and we can see our behavior goes back up to about baseline levels that we saw in the previous baseline phase. We bring back our intervention and boom, there it goes right back down, shows a nice functional relationship.
Okay, on this graph, on the other hand, shows data that does not demonstrate a functional relationship. In this case, the withdrawal of the intervention did not... increase the target behavior towards baseline levels like the other graph showed. So we've got our baseline.
Here we go again. High levels, right? Intervention, it looks like it worked.
It dropped the levels down. But then when we remove it, it doesn't change much, right? Our levels are still not too far off of where we were when the intervention was in place.
And when we bring it back, it does start to trend down a little bit here. We see a decreasing trend, but still about the same. level as where we were at before.
Okay, and so it could be due to a variety of different reasons, which we'll talk about later, but there's a chance that this intervention might not be what caused this change in behavior to occur. So something that we need to think about and keep in mind.