 How do we affect this stuff? How does this actually, you know, how do we improve these things? Improve generalization. Right off the bat, train in the target situation whenever possible. Makes it a lot easier. In other words, if you're in that natural environment when you're doing the training and when you're trying to teach the person what to do, then it's going to kind of maintain itself. It's going to stick in place and it's going to stay there. But if you have to train in a different setting, then you have to remember to make sure you then purposefully get it to bridge that gap. Let's say you can't train in the unnatural setting. The idea then is you train in as many conditions as possible. That way you're teaching that this behavior works in all sorts of scenarios. And when you get to that natural environment where you're wanting the person to actually do the behavior, then they'll likely be prepared to do that because they've learned in these other five scenarios that it just worked. Program common stimuli. The more similar you can keep these environments in these contexts, the more likely it is to reinforce the particular behavior that's going to come about. And again, it's using these common stimuli. Make sure that those stimuli are going to be present in that target situation. Train sufficient stimulus exemplars. You're teaching concepts and functional equivalence classes. The idea here, again, give as many examples as you can. Reinforce those particular examples. As you reinforce the particular behaviors in those new particular contexts, again, sufficient stimulus exemplars, then you're going to produce much more behavioral momentum. That momentum is going to carry that behavior over into that new setting and likely come into contact with that natural reinforcer. Remember, concepts and functional equivalence classes. Concepts, multiple stimuli producing the same response. So that's response generalization, functional equivalence. The same many responses will produce the same outcome. We can think of that in a sense of general case programming. So teach the types of behaviors that will work in general types of situations that they will encounter. So I can teach you how to deal with money. In fact, when I was in Afghanistan, that's one of the first things I learned, was that learn how to deal with money. Learn the language for money. I was never really fully fluent in the language over there, but I was fluent enough in money. So in dollars, or not in dollars, but the money that they use is called afgani. So I was able to use my numbers and things like that across the board, and that made it very functional for me. So I could go shopping for my food if I wanted to. I didn't have to have a translator with me. I could walk down to Market Street or not Market Street, but walk down to a vegetable lane or walk down to Butcher Street or whatever it may be and get the food that I needed to get. So I was taught the type of behavior that will work in many situations that I was going to encounter, and that made it very functional for me. And later, I slowly added more and more language, but I was never fully fluent, and I'm still not fluent today. And I actually have a hard time with the numbers, and unless I'm talking, like if I was to just start getting into the numbers right now, I would be really clunky with it. It wouldn't be very fluid, or even fluid for that matter, but if I was talking to one of my Afghan friends, I would be much more fluent with it. So there is some context-specific stuff there, but the idea is that we were programming for the general, or I was programming myself for the general case. We also want to train sufficient response exemplars. What are the different things that you can do in order to achieve the reinforcer? There's 10 different ways to say thank you. So as long as you teach as many of those as you can, you're more likely to get one of those to occur in the right context. Hopefully, the learner will be able to generalize this response to new cases as well. So one person, in terms of the thank you thing, I may be able to say that in different intensities, or at different times when it's appropriate, as long as I've had enough examples of how to do that, and the appropriate responses that I need to do in order to make that happen. Very the acceptable responses. This is essentially how you develop creativity. Some people think creativity is this inborn thing. It's not. It's trained. It's learned. I can teach you to be creative. I can require a creative response. In other words, I can require that you do something new, do something that I haven't seen before, and you're going to get a reinforcer for that. And then if I continue to do that, keep requiring that same thing that will select for creativity. So whatever we can have people do, these different responses change it up a little bit, but make it still acceptable. Again, this leads us to the idea of functional equivalence responses. There's 50 ways to screw in that screw. They all work. There's 100 ways to turn on the light switch. They all work. So we want to focus on those, rather than focusing on the individual specific little things.