 Should we, should we sally forth and continue to find out what's going on? Yeah, do you have to be sallying yourself? What did I just call you? An actual fire dancer? Morgan is not causation. But there's a twist. You ready? Psycho style. You can't have causation without correlation. So while one may not be the other, one is not a sufficient, and one is not a sufficient. How do they say that? One is a necessary but not sufficient condition for the other. Sounds like I've been drinking today. It's just, we're talking about statistics, and that has a certain effect on my body. Anyway, no, I love stats. Actually, I love talking about them, and it's really fun. So correlation, it's not causation, right? And that is absolutely one of the key things to keep in mind anytime you're dealing with behavioral change issues, anytime you're dealing with a new policy, anytime you're running into a new problem at work in that environment, is that, is that change? Causing the outcome, right? So if you have a problem with, let's say, let's say your insurance companies aren't paying as quickly as they are, and you think it's the result of some new process that happened at work. Well, that's quickly as they are. As quickly as they should be, or you expected, or as quickly as they have in the past. So you think that might be the result of a new process or a new change at work. So you go, what did I change recently? So we get back to that, what can you control things? We're going to look at only the things you can control. Then we're also going to go, is there anything in that environment or any change that I made as a manager or in my staff or in my policies and procedures that could possibly be affecting this outcome? So you're going to look at that and you're going to take a guess. And if you think you've found something, then really what you have is a correlation. I made a change on X, and it seems like it had an effect on Y. And notice I did this in A, B design, right? A, B designs are bad for scientific control and for experimental control. They're good for letting you know that you may have a correlation here that's going on. That if you do see a change here, or you make a change here and you see an effect here between your baseline intervention phases, whatever, then you're on the track to finding out if possibly the thing that you're doing is having an effect. Now, an A, B design by itself is not going to give you the information you need. Correlation is not a causation. However, if that thing that you're worried about is affected by the policy change, the new behaviors at work, the new system, whatever it is that you've done, the change that's made, if it is affected by that in the real world, then you will have correlation to begin with. So in order to show causation, you have to have correlation. But you can't assume just because you have correlation that you have causation. This sounds like a weird statistics argument that I'm putting in the context of work or OVM. But really, it's not even a statistics argument. This, folks, is about just basic understanding of scientific principles and the application of rational conclusions and empirical arguments. It kind of all matches together in this particular one little thing. So when you have, again, so correlation is we're all cautious with it and you're going to run into this all the time. People are always going to be saying, well, this happened because of that or that happened because of this and we did this thing and that caused this effect and we made this change in the software and that caused this problem and like, whoa, slow down. Not only if you're just changing things crazy, crazy, crazy all the time, you're going to get sequence effects, order effects, all these other things that you don't know anything about. You might be on some weird path that has you doing 57 different things when all you needed to do was one. So be careful when you're doing your interventions at work or when you're trying new policies and procedures. Change one thing at a time. That's kind of that scientific part, right? And then observe to see if that has an effect. Now, in some instances, you're going to be able to go back and forth and I've done that in my work and I've had the lucky opportunity to do ABAB type interventions. So sometimes, and we're going to come back to that when we talk about research methods in the OBM setting and why it's so important, but there are opportunities for you to practice all this stuff and keep all of these things in mind. It's just like if you're working with a client. It's no different. Now you're working with the processes and the business and the behavior is the compiled behavior of many, many people all put together into one thing. So if you will, I don't know why that word compile came in there but I just popped it. Anyway, correlation, not causation, but you can't have causation without it. So it's overrated, but it's still important.