 All right everyone. Welcome back. It's Professor Howard. Let's go through the basic anatomy of a graph. In order to communicate with a behavior analyst, you have to speak our language and more often than not the language that we're going to be speaking is data. So let me introduce you to our first graph. If you've watched the previous videos, then you recognize that this is a comparison design. What I'm showing you on screen here is a kind of pre-post measure. I have a simple line graph here where on the left axis or the Y axis, I'm showing you the number of smiles that the client emits. Along the X axis, the horizontal axis along the bottom is typically going to be a time measurement. Now this one is blank, but it could be sessions or days or weeks, something of that nature. What I'm showing you is a baseline condition here, four observations and a no praise condition. I'm showing you one, two, three, four, five, six, seven, eight observations and a praise condition. And we see that in the no praise baseline condition, our observations are in order two, one, two, zero. In the praise condition, seven, nine, eight, eight, nine, eight, seven, eight. And we have these conditions separated by a dotted line, which is pretty common to clearly communicate to the viewer that the condition has changed. So in this design or in this graph, what you want to know is on the Y axis, typically this is the dependent variable. This is the behavior that we were interested in measuring. In this case, we're talking about the number of smiles. In the conditions that we report here, you sometimes won't be able to tell which one is baseline and which one is treatment. The way that I can tell is that if you have something here like praise, and the comparison is no praise, like the absence of this variable, then typically anything that says no is probably going to be the natural condition. And so this is probably going to be the baseline condition. Now it's not impossible, for instance, for these to have been switched around and for you to start with a praise condition to immediately start with treatment and then remove treatment. But for the most part, I can conclude that anything with no is probably baseline. And then from there, this looks like my independent variable, which is present during the treatment condition. What we're looking at to evaluate here is the effect of the treatment on the dependent variable to see if there's any change in the rate of the behavior after praise is applied for the behavior of interest, the number of smiles. So this was a very simple comparison design. Let me give you another one. What I'm showing you here is a reversal design. And you can tell it's a reversal design because again, we conclude here that the the one with no name waking is probably the treatment condition and behavior analysts abbreviate the word treatment TX. So this is the treatment condition. And then on either side of the treatment condition, we have no waking conditions. So in this first condition, we have a no waking condition. In the middle condition, we have waking. And then in the last condition, we have a no waking condition. So it looks like a nice little sandwich of conditions with waking being the treatment. On my y axis, the left axis, remember that y is high, we have the behavior of interest, the minutes weighted. And we don't know exactly what the study is. So we're not sure what the phenomenon of interest is or why they're measuring minutes weighted, but just that this is the dependent measure. And again, remember on the x axis, x is that horizontal one, it typically refers to the time or how often the data were collected. And in this case, we see that days is how often data were collected. So here I know my dependent measure, my dependent variable is the number minutes weighted. I know my independent variable or treatment is whatever waking is. And then finally remember that many of these are line graphs. So I'm showing you a baseline condition, a treatment condition and a reversal to baseline condition. And this is probably nine, 12, 14, 16, 14. This treatment condition is probably one, maybe five, five, five, four. And then finally into the no waking condition, five, four, eight, seven, eight. And that will be important when we come back to visual analysis. Okay, one super quick reminder, when we say reversal, what that refers to is the environment. It's the reversal to baseline conditions. It is not a guarantee that the behavior is going to return to baseline conditions. So don't anticipate or don't count on behavior going back to baseline levels. Let's do one more because so far we've had a comparison design, we've had a reversal design. Now let's look at the multiple baseline design. Multiple baseline design is different in that it has, like the name would suggest, multiple baselines. So in this case, we have a multiple baseline across participants. In this case, we're talking about Miguel and Carla here. I'm showing you two line graphs stacked on each other. The y axis, remember left side y axis, y is high, shows me my dependent variable, which is the number of books read. My x axis, the horizontal on the bottom shows me the time measurement, which is how many books are read per week. And then in going back to our condition labels, we can see that we have one that says no, meaning that something is not in place. And we have just a condition label here that is that description. So in the left condition, we have no book at program. And in the right condition, we have the book at program. This was a popular program in the 90s and late 80s to get kids to read. I think it might still be being done in various points or various places throughout the US. However, remember, if we have no, we can have some pretty reasonable belief that this is the baseline condition. And then whatever that intervention is, is probably the treatment condition. The other critical feature of the multiple baseline design is not only do you have multiple participants or behaviors or settings, but you have to have, you must have stagger, you have to have the stagger. Because if you don't have stagger, if you don't have stagger, you can't show that the behavior changes when and only when treatment is put into place. So let me say that one more time. Behavior improves when treatment is put into place, but it changes only when treatment is put into place. We see that for Miguel, in the period of time when Miguel has not started treatment, his behavior remains relatively unchanged. But for Carla, when she starts the book at program, there's an increase in the number of books read. Now, if you would benefit from having a visually impaired version of these slides, something that's going to show you all the data in the information point by point, let me know. In this case, we have many of the same features. We have that y-axis tells you your behavior of interest or your dependent variable. We have the condition labels. So again, we have the no something in the something. So this is the baseline condition. And this is the treatment condition. We have participants here that's something new that's in multiple baseline designs. The dotted line, as always, indicates when treatment begins. And we have to have a stagger of three or more observations. We see we have exactly three here. We need three or more observations because three points make a trend. Alright, so that is super basic, super simple graph anatomy. Let me know if you guys have any questions. Otherwise, I'll see you next time.