Understanding Alternating Treatment Design

There are several advantages of alternating treatment design. They are ideal for determining which of several interventions is most effective. The efficacy of interventions can be determined pretty quickly. So as you are testing out the different interventions or treatments, you can pretty quickly tell if one is setting themselves apart from the other as being more effective. Another advantage, you don't need that baseline data. And if you do collect it, it doesn't need to be stable before the intervention is initiated, which we talked about earlier. It's also nice because it doesn't require the withdrawal of treatment. That's not necessary with alternating treatment design. You spend less time administering an ineffective treatment. So if you're recognizing that one is just not working, you can remove that from your study pretty quickly. And counterbalancing can help eliminate sequencing effects. So remember making sure you're alternating and randomizing the order that you are implementing the different interventions will help eliminate some of those sequence effects. So when to use an alternating treatment design. So a lot of these go right along with the advantages, right? When you want to determine the relative effectiveness of more than one treatment on a given behavior. So maybe you've came across a couple that you're really curious, you know, which one is going to be the most effective with this particular behavior skill. It's a great design to test those out. You can also use it when baseline data are either unavailable or might be unstable. So that's a little different than some of the other designs that we've talked about. When the treatments are sufficiently different from each other, again the subject needs to be able to distinguish between them. So they do need to be sufficiently different enough. That's the next one when the subjects can discriminate between those treatment conditions. Those kind of go hand in hand. And when the effects of sequencing the interventions might obscure the results. So these are all reasons when or why you might want to use an alternating treatment design. And of course, we have disadvantages as well that we need to make sure we talk about. The possibility of multiple treatment interference. That's that rapid alternation of interventions that can lead directly to this situation occurring. Multiple treatment interference might mask the results of the effects of a specific treatment because either treatments may influence or confound or carry over. They can also have an effect on the no treatment condition too. They can also have an effect on that. So what can you do? If you have multiple treatment interference, there are some things that you can do to minimize those effects. You can use treatments that are considerably different from one another. You can also present the most effective treatment at the end as that final treatment phase to help evaluate the presence of multiple treatment interference too. It's also another disadvantage is it's not appropriate for target behaviors that are not reversible. So again, think about if you were trying to two different methods to teach phonics, like for reading, for example, and one student learned a sound syllable relationship. And it'd be very likely that that's going to be retained, right? They're going to remember that because they've learned that. And so the differential treatment effects could not be adequately determined in that situation. It's also not effective for evaluating independent variables that produce change slowly or need to be administered over a continuous period of time. The treatment really should be able to produce change on a session by session. basis. You're using this very quickly and rapidly. Implementation also can be a disadvantage. It can be seen as artificial and not typical for a natural environment. So because you're rapidly alternating the treatments, that's not typical, right? And if more than three treatments are used, that may also be seen as excessive. So you typically don't want to use more than three treatments in this design. It also makes it difficult to counterbalance in most natural settings. If too many treatment conditions are involved, the participants might have difficulty discriminating among those different conditions. And lastly, it's inappropriate for individuals who may not have that ability to discriminate between or among the treatment conditions. So again, to be able to use this, they need to be able to discriminate. between the conditions as that's a prerequisite for using alternating treatment design. So when not to use it, again, kind of aligns with those disadvantages, but when treatments might interact, thus obscuring the results, you wouldn't want to use that. They need to be different from each other. When the subjects cannot discriminate between the treatment conditions, you wouldn't want to use it when treatments typically don't work. We want things again that can move rapidly and you might see change session to session. When the treatments need to be administered over a continuous period of time to be effective and when it becomes difficult to counterbalance the various aspects of the study. So those would be times that this type of design wouldn't be the most appropriate to use. There's also a couple different adaptations of the alternating treatment design. As I mentioned earlier, the terms alternating treatment design and multi-element design are often used interchangeably and there are some similarities between the two designs, but there's also some subtle differences. So one main difference is the nature of the independent variable that is being investigated. With the alternating treatment design, it's typically used to evaluate the effectiveness of like two or more interventions with the participant. The multi-element design is used to determine the factors that influence a participant's behavior to develop an effective intervention. And that's why it is used as part of a functional behavioral behavior assessment in which the purpose of a specific behavior is identified so that that behavioral intervention program can be developed. So the multi-element design can be used to evaluate whether a response occurs differentially across conditions in order to identify a functional relation. So once the purpose of the behavior is determined, then an appropriate intervention can be developed that should incorporate a functionally related replacement behavior that would serve that same purpose. The multi-element design, it's often used as that first. phase of a larger study. The next adaptation is simultaneous treatment design. I had trouble with that one today. This adaptation to the alternating treatment design is also called concurrent schedules design. So you might hear it referred to that as concurrent schedules design. And this design is actually pretty rare in the professional literature. You won't come across it very often. But as the name implies, the treatments are provided simultaneously or concurrently instead of being alternated. It does have several advantages over some of the other single subject designs. It best approximates the conditions of the natural environment and it takes less time to determine treatment effectiveness than other designs including the alternating treatment design. You can kind of determine effectiveness a little bit faster than a trip traditional alternating treatment design. On the other hand, the design does require more skill planning and organization than some of the other designs and it can be difficult trying to analyze the effects of several interventions that are being presented at the same time. To be a little tricky. The last adaptation is adapted alternating treatment design and this one has been used more often in professional literature. You'll come across this one a little bit more. With this design, each intervention is applied to different behaviors that are considered to be of equal response difficulty, but functionally independent. So that can be time consuming and problematic. But because that process of equating behaviors, again, it can take a lot of time. It can be a little bit problematic, but it's a very important prerequisite to the use of this particular design. Okay, last we have one example from the textbook. This study here involved a fourth grader named Leticia. She wanted to increase the number of correct spelling words from a list of 20 randomly selected words and the study compared the effectiveness of using a computer program for instruction that provided instruction and a peer tutoring intervention. You can see they started by collecting the baseline data first, so that the effectiveness could be better established, and we had a pretty stable baseline, holding steady at about 6 and 7. Then we moved into our treatment phase, and so we rapidly alternated between our two, starting with B, which was our computer instruction program, and then we moved or alternated that with C, which was our peer tutoring intervention. You can see they kind of went back and forth, counterbalanced that so it was different. We didn't have to worry about sequencing effects. And you can see that that computer instruction definitely has this nice positive increasing trend and set itself apart as more superior than the peer tutoring. Even though the peer tutoring you can see still had an effect, right? It did increase when you compare that to the baseline level. We do see an increase slightly, but definitely that computer instruction proved to be a little bit more effective. So then we use that final treatment phase to validate that effectiveness of that superior intervention and demonstrated results were maintained when the more effective intervention was used exclusively. So here's that final phase. It continued to have this nice positive increasing trend showing that that particular intervention was more effective. Thank you. So, while alternating treatment can be somewhat complicated, as you probably have picked up on throughout this chapter, and it can be difficult to manage and to implement, it still is a really great design to use when you have two or more interventions that you want to evaluate for effectiveness. So, you might also want to make sure that that target behavior is reversible, you've got that capacity to counterbalance the different treatments. and to also collect data across all those conditions. But despite all these requirements, alternating treatment design is still a very useful and practical choice that many researchers and teachers choose to use.

And if you do collect it, it doesn't need to be stable before the intervention is initiated, which we talked about earlier. It's also nice because it doesn't require the withdrawal of treatment. That's not necessary with alternating treatment design.

You spend less time administering an ineffective treatment. So if you're recognizing that one is just not working, you can remove that from your study pretty quickly. And counterbalancing can help eliminate sequencing effects.

So remember making sure you're alternating and randomizing the order that you are implementing the different interventions will help eliminate some of those sequence effects. So when to use an alternating treatment design. So a lot of these go right along with the advantages, right?

When you want to determine the relative effectiveness of more than one treatment on a given behavior. So maybe you've came across a couple that you're really curious, you know, which one is going to be the most effective with this particular behavior skill. It's a great design to test those out.

You can also use it when baseline data are either unavailable or might be unstable. So that's a little different than some of the other designs that we've talked about. When the treatments are sufficiently different from each other, again the subject needs to be able to distinguish between them. So they do need to be sufficiently different enough.

That's the next one when the subjects can discriminate between those treatment conditions. Those kind of go hand in hand. And when the effects of sequencing the interventions might obscure the results.

So these are all reasons when or why you might want to use an alternating treatment design. And of course, we have disadvantages as well that we need to make sure we talk about. The possibility of multiple treatment interference.

That's that rapid alternation of interventions that can lead directly to this situation occurring. Multiple treatment interference might mask the results of the effects of a specific treatment because either treatments may influence or confound or carry over. They can also have an effect on the no treatment condition too.

They can also have an effect on that. So what can you do? If you have multiple treatment interference, there are some things that you can do to minimize those effects. You can use treatments that are considerably different from one another.

You can also present the most effective treatment at the end as that final treatment phase to help evaluate the presence of multiple treatment interference too. It's also another disadvantage is it's not appropriate for target behaviors that are not reversible. So again, think about if you were trying to two different methods to teach phonics, like for reading, for example, and one student learned a sound syllable relationship. And it'd be very likely that that's going to be retained, right? They're going to remember that because they've learned that.

And so the differential treatment effects could not be adequately determined in that situation. It's also not effective for evaluating independent variables that produce change slowly or need to be administered over a continuous period of time. The treatment really should be able to produce change on a session by session.

basis. You're using this very quickly and rapidly. Implementation also can be a disadvantage. It can be seen as artificial and not typical for a natural environment.

So because you're rapidly alternating the treatments, that's not typical, right? And if more than three treatments are used, that may also be seen as excessive. So you typically don't want to use more than three treatments in this design.

It also makes it difficult to counterbalance in most natural settings. If too many treatment conditions are involved, the participants might have difficulty discriminating among those different conditions. And lastly, it's inappropriate for individuals who may not have that ability to discriminate between or among the treatment conditions.

So again, to be able to use this, they need to be able to discriminate. between the conditions as that's a prerequisite for using alternating treatment design. So when not to use it, again, kind of aligns with those disadvantages, but when treatments might interact, thus obscuring the results, you wouldn't want to use that.

They need to be different from each other. When the subjects cannot discriminate between the treatment conditions, you wouldn't want to use it when treatments typically don't work. We want things again that can move rapidly and you might see change session to session. When the treatments need to be administered over a continuous period of time to be effective and when it becomes difficult to counterbalance the various aspects of the study.

So those would be times that this type of design wouldn't be the most appropriate to use. There's also a couple different adaptations of the alternating treatment design. As I mentioned earlier, the terms alternating treatment design and multi-element design are often used interchangeably and there are some similarities between the two designs, but there's also some subtle differences. So one main difference is the nature of the independent variable that is being investigated. With the alternating treatment design, it's typically used to evaluate the effectiveness of like two or more interventions with the participant.

The multi-element design is used to determine the factors that influence a participant's behavior to develop an effective intervention. And that's why it is used as part of a functional behavioral behavior assessment in which the purpose of a specific behavior is identified so that that behavioral intervention program can be developed. So the multi-element design can be used to evaluate whether a response occurs differentially across conditions in order to identify a functional relation.

So once the purpose of the behavior is determined, then an appropriate intervention can be developed that should incorporate a functionally related replacement behavior that would serve that same purpose. The multi-element design, it's often used as that first. phase of a larger study. The next adaptation is simultaneous treatment design. I had trouble with that one today.

This adaptation to the alternating treatment design is also called concurrent schedules design. So you might hear it referred to that as concurrent schedules design. And this design is actually pretty rare in the professional literature. You won't come across it very often.

But as the name implies, the treatments are provided simultaneously or concurrently instead of being alternated. It does have several advantages over some of the other single subject designs. It best approximates the conditions of the natural environment and it takes less time to determine treatment effectiveness than other designs including the alternating treatment design. You can kind of determine effectiveness a little bit faster than a trip traditional alternating treatment design.

On the other hand, the design does require more skill planning and organization than some of the other designs and it can be difficult trying to analyze the effects of several interventions that are being presented at the same time. To be a little tricky. The last adaptation is adapted alternating treatment design and this one has been used more often in professional literature.

You'll come across this one a little bit more. With this design, each intervention is applied to different behaviors that are considered to be of equal response difficulty, but functionally independent. So that can be time consuming and problematic. But because that process of equating behaviors, again, it can take a lot of time.

It can be a little bit problematic, but it's a very important prerequisite to the use of this particular design. Okay, last we have one example from the textbook. This study here involved a fourth grader named Leticia. She wanted to increase the number of correct spelling words from a list of 20 randomly selected words and the study compared the effectiveness of using a computer program for instruction that provided instruction and a peer tutoring intervention.

You can see they started by collecting the baseline data first, so that the effectiveness could be better established, and we had a pretty stable baseline, holding steady at about 6 and 7. Then we moved into our treatment phase, and so we rapidly alternated between our two, starting with B, which was our computer instruction program, and then we moved or alternated that with C, which was our peer tutoring intervention. You can see they kind of went back and forth, counterbalanced that so it was different. We didn't have to worry about sequencing effects.

And you can see that that computer instruction definitely has this nice positive increasing trend and set itself apart as more superior than the peer tutoring. Even though the peer tutoring you can see still had an effect, right? It did increase when you compare that to the baseline level. We do see an increase slightly, but definitely that computer instruction proved to be a little bit more effective.

So then we use that final treatment phase to validate that effectiveness of that superior intervention and demonstrated results were maintained when the more effective intervention was used exclusively. So here's that final phase. It continued to have this nice positive increasing trend showing that that particular intervention was more effective. Thank you.

So, while alternating treatment can be somewhat complicated, as you probably have picked up on throughout this chapter, and it can be difficult to manage and to implement, it still is a really great design to use when you have two or more interventions that you want to evaluate for effectiveness. So, you might also want to make sure that that target behavior is reversible, you've got that capacity to counterbalance the different treatments. and to also collect data across all those conditions. But despite all these requirements, alternating treatment design is still a very useful and practical choice that many researchers and teachers choose to use.

Transcript for:Understanding Alternating Treatment Design

Transcript for:
Understanding Alternating Treatment Design