 I think we'll start this out with the demo. We've got a hell of a chain, as you're going to have eight hands. I wonder if I'm in the clouds. You grabbed by the pumpkin full of her. We might be wrong. We're funny, but not always a joke. Alternating treatment designs. Now, keep in mind, first things first, a lot of people will call these a type multi-element design, and I suppose they are in the strictest sense. If you've watched the other design types, I tend to separate them, multi-element designs from the alternating treatment design. They can be mushed together, but there's reasons why I separate them. It has to do with just kind of the understanding of logic. Remember, multi-element designs, as I talk about them in other videos, are kind of more along the lines of an ABAB design type logic. Whereas in an alternating treatment design, I tend to think of them like this. It sounds a little crazy, and I don't know if we got something up back there for you or not, but we probably did. The point is that an alternating treatment design literally alternates treatments. So, let's say we've got three treatments, B, C, and D. I don't know why, but we have three. We can just start randomly throwing those things. Notice I said randomly. That's a special type of an alternating treatment design. But we can just start throwing the treatments at you. Imagine if I put you under one condition. I don't know. It doesn't matter what the conditions are. There's a billion examples, and it just confuses people sometimes. Let's wait here in a second. Let's start off with a B. We get one data point on B, then we can go to C, then we can go to D, and each one just gets one data point. Then we can go back to C, then we can go back to B, then we can go to D, and C, and D, and D, and C, and D, and C, and A, and then C, and B, and C, and you just bounce back and forth. You can do this for quite a while, as you might imagine, and it's alternating treatments. You get a whole bunch of lines, a whole bunch of phase changes in your data, and we don't draw lines across those phase changes typically, right? Everybody knows that you never draw a line across a phase change. Unless it's an alternating treatment design, then you draw lines to each of the similar types of conditions. So even though you're phase changing all the time, in order to interpret this, you connect all your triangles, which whatever, that might be your B's, and you connect your squares, which might be your C's, and then you connect your circles, which might be your D's. You draw those together. You don't have to draw the lines. It just helps make it the, and it helps you interpret what's going on to see which one is the most effective. So, why would you use something like this? Well, you don't have to worry about the reversibility issue right off the bat. You don't even need a baseline, you don't need stability. I mean, there's so many things when you're like, by now you're going, well, that's the entire logic of single-subject research, and I don't know what's going on now, while you're darn right, because it's about repeated measurements, folks. I can't tell you that that reaction is completely wrong. I can't tell you that this is a very functional design type. It's highly useful. Do you lose some of your ability to talk about functional relations? Of course you do. But again, pragmatically speaking, maybe that's not always the most important thing. Sometimes it is. It depends on what your goals are, right? Functional analysis, you bet that's one of your goals. It's like your only goal. But if you're just trying to figure out which treatment works best in this particular environment, with this particular person, dog, worm, or whatever, then you don't need to worry as much about that, because the data will tell you which one's working the best, right? So there's variations that you can do. The randomized alternating treatment design, I did this one once, I actually did an experiment with this. Some day I think I'll publish it and I'll try. It was really cool. We had, basically we had online, we had a class and we had online components in that class for some weeks and not for other weeks. So some weeks you came in, you got your lectures live, some weeks you went on to YouTube and got your lectures. And then I measured their exam scores, but I ran, but I alternated everything and it was completely random. You never knew which week and I didn't tell the students until the night before class started, if it was going to be attend to class week or an attend to online week. And we randomized it, just made it complete, and just kind of rolled the dice and came up with this design type for a 16 week course and that's what we had. It was pretty cool. It basically showed that there was no difference. The students did fine either way. That was it. So sorry, but just completely like take the wind out of yourselves there, but no, it just didn't, and with that particular experiment it didn't matter. So we randomized, we did a randomized alternating treatment design. There's any number of things you could do here. You could actually add a baseline to get a little bit more of a, sort of pre-intervention information about where that was and we could make a comparison relative to that baseline. That's cool. So you get a baseline, then you start your alternating treatments. Let's see what else could we do. Oh, you could do the baseline thing. Then you could do your alternating treatments and then you could employ a best design or a best treatment option. So you can flip your alternating treatments to decide which of the three interventions that you have are the best and then you could put them into a, then you could actually start a whole phase where you're just doing the best intervention of all to give the client or wherever you're working with the most, kind of the best benefit of your work at the end. You can even do some variations on that and then you could start an AB-AB design based on that. There's so much you could do and I want you to realize after you've watched a bunch of these lectures on the methodologies that they don't have to stand alone. In fact, I'm going to do a whole separate video on putting these things together. Why? Because that's how it works. You take little bits and pieces of each one of these design types and you take the logic from here and the logic from here and the logic from here and you put them together to be able to draw some cool conclusions. So, alternating treatment designs are really cool. There's a whole bunch of stuff you can do with them. There's lots written on these things. They're pretty easy to understand. I covered most of it but I'm sure I didn't cover all of it. So, anyway that's kind of what you need to know about an alternating treatment design. Connect the like treatments together. That's one of the keys. Sometimes that's hard to remember. So, anyway, I think that's it.