Differential Reinforcement

So without further ado, sit back, relax, and always analyze responsibly.

You know, don't always praise the Zoom, and a lot of people like talking about how fatiguing it is these days, but in terms of convenience and flexibility, we'll offer it a little praise today, huh? So I'm excited, man.

We had a couple good podcasts on the criticisms of ABA and looking at them, but getting back in the comfort zone and looking forward to talking about some good old, you know, meat and potatoes, ABA, and some ways we can improve or just... some ABA strategies, so super excited for today, Mike.

Dan, give us a little background, a little history, a little knowledge there.

Yeah. Kate came about with talking with you about kind of how linear things get in ABA and how we can maybe make it a little more contextualized and a little less lab and more living room and more just naturalistic and social with differential reinforcement and putting things into ABCs, but maybe a little bit different than we might be comfortable or accustomed to. So look forward to talking about that.

And yeah, I think that I guess getting started that lot of times with our assessments we ask parents skills that their kids do and skills that their kids don't do and parents are really good at letting us know all of the things their kids can't do and the maladaptive behaviors they're they're emitting because parents people are hardwired to look for these maladaptive behaviors but when i can't tell you how many times A parent's like, they just won't stop hitting themselves or they just won't

stop using this language. And I'm like, well, what do you want them to do? And the parent's pause and they have to think about it because they don't even know what the appropriate replacement behavior is that they're trying to seek, which means they have no idea what they're trying to differentially reinforce. All they know that they're trying to stop something. You can't stop something. You always have to replace it, right? It's like telling an alcoholic not to drink.

Well, that's not going to be too successful. So I'll pass it back to you, Mike, and we can go ahead and get started. But I think this concept of differential reinforcement is going to be exponentially important to any parents, practitioners. If you want to see more constructive behavior management and behaviors morphing with your kids and clients.

And at the same time, from an applied perspective, I actually feel quite skilled at applying differential reinforcement. And I think a lot of that lies in the readiness to look at multiple three-part contingencies active at once.

So one of the things that we discussed in preparation for this podcast were some of the obstacles that are naturally part of ABA, especially in current day practice in the way we train a lot of our behavior technicians and the way a lot of our basic training in ABA, especially professionally speaking, comes about. We talk about discrete trials. We talk about three-part contingencies.

And how many reinforcing consequences in your experience, Dan, do those reinforcing contingencies usually involve? How many items? Are we looking at many possible ways to reinforce? many different behaviors, or are we looking at one SD, one response, one consequence?

And again, I'm not saying it's incorrect to conceptualize it that way. It's just not living room practice. So what do you think about that?

You're doing all sorts of things at the same time. So for you to say like you're vocalizing a man, well, there's so much extra that goes into that. So many, I guess we could call them auxiliary behaviors that to say, Oh, there's this, just this one behavior. Well, what if I say my man's in a level nine voice versus a level two voice? You know, what if I say it, um, with a cuss word before it or not?

Like there's all sorts of different behaviors that kind of break from the linear, uh, You know, I hold a phone in front of you. The kid says I want phone and I give him the phone. There's so many other behaviors in there that I think we need to start acknowledging and looking at the complexity of that, because then we can start to find things to differentially reinforce.

Because if I'm giving the SD of holding a. a phone in front of you, and my case manager's told me that I need to withhold that phone until you look at me in the eye and you say, I want phone, and that's all you can say, and then I can give you the phone, there's going to be plenty of times where we're just going to be both stuck in neutral because you're not doing it, and then I have to put you on extinction until you do the behavior I want.

But there's so many other behaviors in there that can be differentially reinforced.

Change in experience as a result of your behavior that makes you go, oh, like, just the simple idea that when we withhold items, right? You know, the simple idea that it can come closer if you're getting closer, that you might be able to, and I get to practice this more easily with younger children, but just in doing the joint attention paradigms, which I love for this reason. I think they involve a lot of natural differential reinforcement.

Because the idea is that if you naturally grab an item that's in my hand and I just squeeze it tight so you can't take it out of there, at some point that child's going to look at my face and be like, hey, aren't you going to let this go? And then I can let go that first time. Because now I've taught them to look at my face, which means the next time I can insist on a little bit more with a different item that's now got more novel value or what have you.

Those things that weren't our best practices are usually very good at contriving. And I think that's maybe for a later date, but I think that that's where we've spent a lot of time in ABA measuring and documenting our data maybe a lot of times toward best practices, things that we should just be doing no matter what, but we're checking these boxes because that's what we've been trained to do.

And again, this will be for a later date, but are we really defining a lot of times the real socially significant behavior or have we taken exemplars for targets? And I think we've alluded to this topic before, we'll get into it at a later date, but I think, again, it has relevance here with the idea of differential reinforcement.

You alluded earlier, and I don't know if you want to bring this back, but even just the idea of replacement behaviors, functionally equivalent replacement behaviors, falls prey to this linearity of a three-part contingency. Replacement behavior or behaviors, there's a big difference in those two terms. And I think the singularity tends to be our habit.

It doesn't matter what it is, but if you're saying, if you're establishing your point and you're saying, And I disagree with your stance on something. I'm not just going to be like, nope, I disagree. You're on extinction. we've got to find some common ground, right? It's got, I have to differentially reinforce some parts of your stance so we can establish a common ground, right?

So like, hey man, I might not agree with your stance on taxes, but that abortion stance makes sense, or that stance on sexuality makes sense, or something like that. So just because you're not exhibiting the entire behavior or belief in this stance doesn't mean that I have to just put you on extinction. We can find areas to differentially reinforce. Once you find those areas of differential reinforcement, Within a conversation, you can then get on common ground.

It seems kind of one of the things with a divided society, and we won't get too much into that, that people aren't willing to differentially reinforce anything, and now they're just getting stuck in it's all or nothing. And I see a lot of that similarity with ABA, that it's the entire behavior I want or nothing. I can't accept anything else. So that just came to my mind as we were talking. I think it's super relevant.

As people try to put it in the differential context into the differential reinforcement into the context of just everyday interaction and how we're constantly differentially reinforcing people through our conversations and all sorts of different interactions.

And a lot of times, again, we're establishing this idea that we're going to have one set of those in linear fashion and then we're going to get stuck on prompting that one over and over to get the one response that we're after that we've called the target when really it's just one exemplar of, you know, say a larger vocal behavior semantic category or something like that. So, it's... It's interesting how pervasive this linearity is in everything we do.

It just seems to really disallow us from looking at the greater range of consequences, the greater range of allowable behaviors that do occur because we dig our heels in one singular linear contingency most often. Even when we do apply DRI, DRO, DRA, or differential reinforcement of incompatible other or alternative behaviors, How many times in practice or in theory and reading do you see those being suggested in tandem or in conjunction? We're also using those in a singular application.

So are you going to do a DRI? Are you going to do a DRA? The idea of this now moving away from conversation as an analog in this interchange of mans and introverbals where we might not necessarily get reinforced for that man we put out or that introverbal doesn't get offered back to us as a speaker, it doesn't necessarily mean that we give up and then not put anything else out there.

So instead of having a conversation, often what we're having is, I'm not even sure what you would call it, but I know we had a chuckle as we were preparing for this and looking at your basic greeting paradigm and it's like, hello, Dan. And then you don't say anything, so let me prompt you, Dan, now you say hello to me. I mean, how natural.

does that sound and how differentially reinforcing is that by nature it isn't it isn't so how we sort of pursue this idea as more of a conversation that it's really our role to make the conversation happen so we're gonna have to put that manned out there and that hook and that engagement signal out there to try and get a response. And then we can reinforce that response. And then we're going to try to get another one and reinforce that response.

And it doesn't necessarily mean that we have to prompt it and model or prompt it every time. So yeah, I know I've said a lot there. I'll hand it over to you. I don't know if anything, any of that made sense or inspired any.

I can't do other things because the more things I'm doing at the same time are going to add confounds, and now we've got new independent variables. So I'm losing the functional relationship between the DRA and the behavior chain. So in the lab, that's necessary, right? In EAB or maybe even ABA. But in BSD, behavior service deliveries, we deliver it. We don't need to prove...

It's the whole reason that we don't do really withdrawal designs because we don't need to prove necessarily which one is working. We just need something to work. And if we can make it more natural, then the parent doesn't care if it's a DRA or DRI. As long as we can do a treatment package probably consisting of all of them in some regard, I'm probably going to put somebody cussing at me maybe on a DRO-ish treatment And then do a DRA when they ask appropriately.

And a DRA for you can't cuss and not cuss at the same time. So a DRA and DRI is kind of the same thing in that point. Again, it gets kind of into lab semantics. So without getting too much into that, my point being that let's not focus so much on which one we're doing. because we're past the lab stage. Let's do them all. Let's get out of this linear.

Let's get out of this we need to prove it and just get into what's going to work most likely in a living, functioning individual because we're not dealing with pigeons and lab rats.

So one example I like to look at is that of a... tantruming three-year-old. And they... Let me describe this, and I'll try to do my best here and do it quickly. They tend to start crying uncontrollably. They also drop to the floor and, in a pretty obvious attention-seeking manner, bang their head on the floor. You know, occasionally. Of course, as a parent... one omission of that behavior is concerning enough.

So let's kind of try to work together to dissect that or task analyze some of the incidentals that might happen or some of the things that we might prescribe that might happen towards really arriving at what would be, you know, again, looking at my ideal British child standing straight up and their hands behind their back and poised and looking straight in your eyes and speaking very properly with perfect etiquette, right?

So we can get from the floor crying, kicking, screaming, hitting my head and seeing if you're paying attention to that point.

But a lot of times in practice, as we've discussed before, we are often conceptualizing waiting for the reinforcement to occur until some much greater semblance of the desired behavior like them standing fully or like them being completely quiet and standing and verbally communicating before we're offering some reinforcement where would it be reasonable for the sake of discussion that you might offer reinforcement to the child for still being on the ground but not crying?

And how often have we conceptualized things that way? And then further, what if they stand and they're trying to catch their breath and talk to you, but their tone still sounds inappropriate? How often do we see, you know, professionally, well, no, that's not appropriate. Your tone or our goal is an appropriate tone. Have we missed a differential reinforcement opportunity there? I would say we have every time. So now the answer that we've been kicking around here is functionally equivalent.

We're working toward functionally equivalent replacement behaviors But we're saying that along the way, there are other scenarios that we might not offer five marshmallows for. but we're going to offer one. Get a little taste of something good because that part was right. But usually we're saying, nope, that part was cool and that one was cool too, but you didn't get the right tone and you're crying still, so no, you don't get any reinforcement.

And we've lost a whole bunch of desired behavior or prerequisite behavior toward the goal and just left it without any attention.

But, you know, what they're willing to accept is all across the board. And that's got to be how we start this differential reinforcement with the appropriate... God, and you brought... We should almost never really use the word appropriate replacement behavior anymore. It should be appropriate replacement behaviors. It should always be pluralized, right? Because there's not just one... behavior that we want to differentially reinforce.

It's multiple behaviors, some semblance in the direction of that terminal behavior that we're trying to achieve, but there's a lot of behaviors within there, so many behaviors. I think

various behaviors like Maybe you're still lying on the floor and kind of hitting your head or pretending to, and you're looking at me, but you're not crying. It intimates all those little scenarios along the way that you might lend a little love to. You might lend a little of that attention to or whatever the function is. You might lend a little of that based on that behavior as you work toward it.

But even in saying functionally equivalent replacement behavior, that shaping idea gets thrown out the window a little bit. So not only do we have this This focus on sort of a linearity and the three-part contingency by nature, as we were saying, almost eliminates the idea of many consequences, let alone many behaviors that might receive certain consequences in route to or shaping a functionally equivalent replacement behavior.

But then we also have this traditional focus on the aberrant behavior as what we need to focus and prove that we're reducing. So that's sort of our self-fulfilling prophecy. What's your thoughts on that?

But again, it's not necessarily what's in our language or the way we conceptualize because of the urgency, because parents need us there to reduce the obvious. So that the idea, the notion that we're gonna come in here and maybe say, which would be very practical, hasn't everybody else, if we're here, hasn't everybody else already defined the aberrant behavior?

What if we just place our focus on the stuff that makes that go away as opposed to continually focus on that stuff that allows us to be here? If that makes any sense.

And then you've basically almost gotten rid of the behavior for decrease, just almost thrown it out, both from a quantifiable perspective and like an application perspective that I think you think that by doing that, it leads the RBTs or the behavior technicians to focus on other things. So that was a great idea. It seemed to be pretty successful so far. So let me pass it back to you on why you've done that, because I think differential reinforcement is probably the answer to that.

So now Early Start and the younger kiddos kind of give me, and as they often do in our conversations, they give me a nice developmental premise and framework to start with because I can expect certain things to be online and then there's already some delays. So it's easy for me to realize that a kid that hasn't started talking before the age of three might cry a lot.

So I can explain that premise a little bit better in terms of its function or lack thereof from a social perspective or a diversified social perspective. So it's a little bit easier for me to get away with this. And I want to say that. By no means am I saying that measuring undesired behavior or the challenging behaviors, it's obviously a huge part of our task in many situations. But Allow me to sort of explain a little further and maybe other parts will apply to people that are listening.

Knowing that certain under three-year-olds are going to cry quite a bit if there's been a language and socialization delay, I'm able to then shift the focus to the verbal behavior, to the joint attention pieces, to all those other things that are going to naturally build awareness.

Some level of self-soothing and self-regulation that comes from being aware of your environment and understanding what's happening next and that someone like us is there kind of offering you the world and all sorts of hopefully enjoyable activity and experiences.

I've been able to then shift our form and function, our application to, like you said, really just focusing on teaching those things and then not occupying our technicians' hands with data or anything other than a simple behavior plan that we've discussed and that they have in their head and that they've also already worked with the parent and they're working in tandem and I've been there to work it a little bit. And these are the best practices that I alluded to earlier.

this is our bread and butter. Why do we have to measure and write this out so much? Again, I'm not saying documentation is not so important, but this is some of the linearity or a few of the foci that maybe take our focus away from more of the dynamic practice and the differential reinforcement aspects that we're trying to elucidate here and to shed more light on in that sense, exactly.

Again, I've been able to get away with this and it hasn't applied for every early start client or every early childhood client that I do with because on the other end and this will be really interesting I think in discussion I'm not sure if we'll be able to explore it more now or later then there are the other kiddos I'm discovering that once I'm able to help them self-regulate, if you will, to calm them down, to reduce their threat response, then a lot of the deficits that were examined and

evaluated are no longer proving relevant outside of having missed a little practice because there was a child with such a high autonomic response. So in that sense, those kids... only have a few behavior goals in terms of measuring reduction of this response or this behavior because by and large, once we do that, I see that all the other areas are okay for the most part.

So verbally or from a communication perspective, we're still going to look at things like one-step instructions because for a two-year-old, that's just cool from a receptive communication aspect and from a compliance perspective and making them more teachable.

So we can still measure those things, but again, now I have just a few developmental goals in those domains, and then now a real true focus on the quantifiable pieces of the aberrant behavior, but not necessarily measuring a whole bunch of aspects other than the decreased duration and frequency, having worked out a pretty clear but simple behavior plan with everybody involved. So I love it.

I know that I'm very grateful that you bring it up because it's something that I'm very proud of professionally over the past couple years that people like yourself have encouraged development of stuff like this and allowed me to try it out. And then further seeing it for what it's worth. So I'll be the first to say that it doesn't apply across the board. in terms of practice yet. I think that the concept and the construct though does.

I think that it's good for us to start thinking about the idea that we already know there's undesired behavior there. Do we really have to count it? And how often do we have to count it? How big of a focus does it have to be in our data tracking and our efforts outside of the procedures to reduce it? I

And she's like, well, I don't know, maybe, yeah, I'll go. And then I hear the, yeah, I'll go. I don't care about the other stuff, it's just noise. And so that's kind of what we're talking about in this situation is, yeah, the yelling and the screaming and the protesting, that's all noise. Let's not focus on that.

Let's focus on what's behind that and what could be reinforced within that and stop spending so much time on the, well, the fact my girlfriend said she'll do something that I wanted to do. Does it matter that she didn't say, yes, I'd love to do that, honey? Like, it doesn't really matter, right? Like, why does it need to be? Back in this linear context, right? There's just so much noise. So let's not focus on that. Let's focus on the message behind it.

We're really in that picture, or in that scenario, what you should be doing is wiggling that connection, trying to figure out why is it not grabbing, continuing to apply some sort of solution to it. But instead, the way we're practicing is, oh, okay, let me try to plug it in. Did it work? No. Okay, now let me reset my trial. Oh, let me try the same thing again. Oh, picture? Nope. Okay, let me try again.

So it's, you know, and again, I'm not trying to make fun of of our efforts because empirically, scientifically speaking from a methods perspective, they're solid, they're spot on. like we like to talk about in this brewing analog and comparison of ours, there's a lot of things sort of coming together and filtering and ebbing and flowing to arrive at the right taste in your living room.

So I think these are all really, really important constructs of this frontier of differential reinforcement, which I know even just a minimal exploration over the past couple years from a real pragmatic perspective has done very well for us in practice. We've had a good time today. Had a good brewing session. Sort of leads us to this idea that we're talking about reinforcement and we're still talking about consequence-based ABA.

And there's nothing wrong with that, but if it's one thing we've made clear in our brewing sessions is that Man, there's three parts to this contingency. And why are we only turning the C knob when that A knob can use some adjustment and some changes too? Something we like to talk to parents about and something that I'm going to say, we don't have time to get into today and that's okay. It's a whole other topic. But the idea of differential reinforcement, yes.

Then what else do we have to look at to apply in that recipe? Along with this now dynamic differential reinforcement that involves a lot of different C's or different B's that get different level of C that might be evoked by all sorts of different acceptable and diverse A's to promote generality. So we're talking about antecedents, behaviors, and consequences. What else, Dan, do you think might lie in this equation?

So, yes, it's done reactively to one behavior, but then proactively to a subsequent behavior in a behavior chain. So let's differentially reinforce some behavior before the likely, you know, if we're talking about headbanging, the headbanging behavior occurs. What behaviors can we reinforce before we get to that headbanging behavior? Because, you know, we can evoke that behavior. That's no problem. And then I can differentially reinforce after it. But now it's occurred.

And now learning is not really going to be that important. on board until that behavior stops and now we're back at blanket extinction. Which, side note, I think I'm going to get a blanket extinction blanket. I'm going to make that.

Let's make that a priority as opposed to looking for those instances of headbanging, which are probably going to occur because we know they occur. Let's find the behaviors that are happening in its place that we can reinforce.

And now we're on consequence, basically.

We covered the idea that replacement behaviors and the application of even differential reinforcement of incompatible other or alternative behaviors continues to suffer or from this linearity or this blanket application. What else did we cover, Dan?

We gotta differentiate our approach to arrive at some other differential admission of behavior that is alternative, incompatible enough from the targeted undesired behavior that now we're rolling in that direction. So I think all these are very, very important things to consider as we are all out there helping the tradecraft and the field progress and evolve. Dan, always a pleasure, sir.

You can leave us a message on Facebook. You can email us. Please do reach out. Give us your feedback. What do you want to hear us talk about? What do you agree or disagree with? We want to hear from you. We want to keep this dynamic and a brew that everybody can... not only stomach, but actually enjoy. So, hey man, always good to see you. Cheers, brother.