Selection of Function-Based Interventions for Negatively Reinforced Problem Behaviour
- So Dr. Linda LeBlanc received her PhD from Louisiana State University, and she's published more than 90 articles and book chapters on topics such as behavioral treatment of autism, technology based behavioral interventions, behavioral gerontology, and system development in human services.
Dr. Leblanc is a licensed psychologist, and the president of LeBlanc Behavioral Consulting, which is a relatively new endeavor and adventure for her.
She's also the executive director of Trumpet Behavioral Health, and she served as professor at Claremont McKenna College, Western Michigan University and Auburn University.
In addition, she is an associate editor for one of our favorite journals, Behavioral Analysis In Practice, and a literature review editor of Education and Treatment of Children, and Dr. Leblanc was the 2016 winner of the American Psychological Association Nathan H.
Azrin award for distinguished contribution in applied behavior analysis, and so we're thrilled to have Dr. Leblanc to present for us this afternoon.
- Thank you very much.
Does the audio sound good? Can people hear me? Wonderful.
I'm delighted to be here with you today, and Behavior Analysis In Practice is also one of my favorite journals, as is Education and Treatment of Children, and what I'm gonna be talking with you about today is really designed to be a model that can assist behavior analysts and educators in collaborating to select function based treatments for negatively reinforced problem behavior.
So, I hope that this is a topic that is gonna be useful to you, and I definitely want to just couch this in terms that I'm an advocate of a collaborative model, as opposed to an expert model, and so I think that when you use this algorithm or the one that I've published for attention maintaining treatments, you really need to be viewing it as you are selecting a treatment collaboratively as the developer with the implementer, so that you are best able to do a pro con analysis of the likely barriers to implementation and the likely success points for implementation.
So, let's go ahead and get started, and we're gonna focus on treatments for negatively reinforced problem behavior today.
So, what we're gonna do is first talk a little bit about the features of a treatment that actually make them function based, and some of the most common evidence based, function based treatments for negatively-reinforced problem behavior.
And then finally decision-making guidelines for selecting among these different treatments for negatively-reinforced problem behavior.
And at the end I'm just going to spend a little bit of time talking about what I call managing logistics.
Anytime you're going to implement a treatment, a behavior intervention plan, there are certain things that just can be expected to be a little bit challenging, especially in the initial implementation of that intervention.
And the better that you can prepare for those, kind of shore up your resources in advance of that implementation, the better you are likely, the better off you're likely to be in the long term.
So let's take a little bit of a look at some of the literature, and this pie chart illustrates the distribution of functions of problem behavior in a large epidemiological analysis that was done by Iwata, and a very large number of colleagues.
So, the standard experimental functional analysis methodology was used in all of these cases, and there were 152 cases, and those were with children and adults with developmental disabilities, all of whom had self-injurious behavior as the topography of their problem behavior.
Now this is important because when you look at other kinds of topographies, you may see a different pattern of the distributional functions.
But in this incidence we're talking about severe problem behavior, self-injurious behavior in all individuals with developmental disabilities, many of whom may also have had autism, albeit though these data sets and these cases were treated at a time when we were less likely to be talking about autism per se, and instead generally talked about intellectual and developmental disabilities.
But let's take a look at the pie chart.
So, the largest piece of the pie is the blue piece over there.
Hopefully you can see my cursor, and that is the escape or negative reinforcement function.
So, in 38 percent of the cases, they identified that escape was the maintaining variable.
Now, second was automatic or sensory function.
Remember, this is self-injurious behavior as the topography.
Attention was the third most common at just around a quarter of the cases, and then there are just a handful of either acts as to tangibles as a function or multiple controlled problem behavior.
Now, there's a very small small percentage that multiply controlled or undifferentiated, and one of the things to keep in mind is that this methodology is generally quite successful at identifying the function of problem behavior.
So, only a very small percentage for whom that couldn't use this methodology to explicitly identify the function and not a lot of multiply controlled problem behavior.
In more recent analyses, you're more likely to see multiply controlled problem behavior.
So, let's look at some of that.
This is an analysis by Hanley, Iwata and McCord published in 2003.
And what they did was they looked at all of the published functional analyses in JABA in a certain window of time, and that turned out of be approximately 500 functional analyses of individuals with a variety of developmental disabilities, certainly including autism.
Now, they included functional analyses of any topography, so we're talking about SIB, aggression, disruption, tantrums, kind of that host, whole host or array of problem behaviors that you might see as a practicing clinician, and again what you see is this is a robust methodology, only four percent being undifferentiated, but up to 14 percent being multiply controlled.
So when you have those other topographies in there, and kind of the different kinds of environments that are more likely to be in play when you're comparing the '80s to the 2000s, you see kind of a different array, but one thing that does not change is escape is still by a margin the most common function, so that negative reinforcement function is still showing up in one third of cases with a attention now being the second most common one at 24 percent.
So still about a quarter, but the automatic function is smaller, the tangibles and multiply controlled are larger.
So, the negative reinforcement function for problem behavior is the common one, the most common one, and that makes a lot of sense when you kind of think about how we might respond to problem behavior, and also about how difficult learning might be for some one with special needs, and how often we are trying to teach them which could potentially inspire our best intentions become a somewhat adversive situation.
So, when we think about decision-making guidelines, and I'll be presenting a set of these here today.
One of the things that's true is in graduate school we're taught how to search the literature, and how to implement the procedures.
Here you do your literature research, you found functional communication training as an intervention, and here's how you do FCT.
But you also kind of need to know when you were going to have a better shot at success if you select this one versus that one.
And there is some guidance available from the discussion sections of some literature reviews in empirical articles, but most of those literature reviews, they're developed for very different purposes.
They're almost designed to facilitate next research questions rather than hey, you're in the trenches, you need to make these decisions everyday, you need to make them in a well-informed collaborative way to increase your chances of success in implementation, because you're going to be around and in this relationship with the person you're consulting with for a long time.
So, here's a couple of examples of literature reviews.
One on DRA, one for non-contingent reinforcement, so those are great, but they don't necessarily tell you about all the other interventions.
Their purpose is to tell you about that intervention.
So, we created this article really kind of as a distillation of some of our kind of years of experience leading to why we selected certain treatments over others.
So, we were professors and training students and trying to train them to be really well-prepared, to function independently after graduate school, and to do so in a positive and collaborative way, and so when we would work with a client, do a functional analysis and identified a function, we kind of had our certain go to interventions and when students would ask why are you picking this one instead of that one in this situation, you know, the answers to the questions were always based on something about perceived safety or ability to implement or client characteristics or implementer characteristics, a variety of those variables.
So, we tried to distill kind of those responses and reflected on what's controlling our behavior when we pick this intervention versus this intervention in this situation.
And what are the strengths and weaknesses of each of these different interventions that would make them more suitable to certain situations versus others.
So, think of it this way.
Problem behavior presents itself, you do a functional assessment, hopefully a functional analysis because you have so much better of a chance to get to a great level of confidence about that function of problem behavior, and you identify that escape from instruction, adaptive skills, some kind of perceived task or work as the controlling variable.
Well, you're in a great position because there are a variety of different interventions that are explicitly designed to address the escape function.
Now, there are other interventions that aren't necessarily designed to address the escape function, they may be dealing with positive reinforcement or what have you, but these ones all directly act upon the escape function itself, and they do so in different ways, so these are some of the most common ones with large literature bases, and I've separated these into columns.
In terms of how it is they operate upon a function, such that you should be likely to see a positive effect.
So, there are really three ways that an intervention can be function based.
The first is that it can eliminate the existing reinforcement contingency, and one way to do that is extinction, no longer provide reinforcement for a previously reinforced response, and that definitely interrupts the prior contingency.
Another way to do it is with a procedure that is sometimes called DRO or DR zero rates, where basically the absent rather than the occurrence of a behavior resulting in the escape reinforcer, the absence of the behavior for a specific amount in time results in the reinforcer, and if the behavior occurs, it eliminates the opportunity for that reinforcer, which is how you're eliminating reinforcement for the problem behavior.
Now, another way that you can directly address that function is to try to reduce the motivation, the MO, the motivating operation.
This is one of my favorite columns, because it really speaks to that notion of quality of life.
If you have fewer adversive things in your life, you are less motivated to have those bad things get out of your life.
So, curricular revision, demand fading, noncontingent escape, and choice based procedures, all address that motivation to escape the instructional environment.
And then on the right side, this is also a really appealing column.
You're basically establishing alternative access to the reinforcer.
Either by virtue of a direct request demand for that escape reinforcer, that's called functional communication training.
Or, by selecting another appropriate behavior that is substitutable or that is differentially reinforced.
And so, for example, you might differentially reinforce compliance or accuracy with escape, so the important thing is the reinforcer that's being accessed is escape.
On the first column, the reinforcer that's being eliminated is escape, and the middle you're reducing that motivation for escape.
So, all of these directly address the escape function.
There are other options that are sometimes added in.
So, for example, if you differential reinforcement of appropriate alternative behavior and you provide, let's say if you get three of them correct you get a break, and access to tangibles, don't forget, access to those tangibles, that's positive reinforcement and it's being added to the break.
You're not going to have the ongoing instruction while they have access to those tangibles.
So, you're intervention probably works in large part because it's a differential contingency to get that break, which was the reinforcer for problem behavior in the first place, and you have perhaps enhanced the effects of that by introduced some preferred items into the break.
Well, let's talk a little bit about is we reflected on why do we like the interventions that we like? What is it that we like about them? What is it that makes the intervention appealing in terms of what my values are as a professional? And you don't have to share these values, or they don't have to be in this order, you could put them in a different order.
The important thing is that you know your values.
So, for example, the things that are most likely to control my selection of one intervention over another are the quality of life of the client is very important to me, that if we have a client who was in situations that they find adversive, that we not set up interventions that just eliminate their opportunity to escape from that and have them just tolerate things that are not of high quality.
We don't want to leave people in miserable situations and just make them less of a bother to us.
It was interesting and I think valuable article on this, that we are not having them sit still and be quiet.
That's not our purpose in the world.
Instead, we wanna do things that kind of meaningfully enhance the quality of life of our clients.
The second thing that's very important to me is safety of both the client and whoever is going to implementing this intervention.
Not only because safety is valuable in and of itself, but because when people do not feel safe, they are less likely to implement the protocol with a high degree of integrity, and we know beyond a shadow of a doubt that procedural integrity for an intervention is one of the key determinants of its effectiveness.
So, if you are not thinking about safety, you may have someone who you think is going to implement an intervention, and they are not going to do so when a situation escalates and they feel frightened.
So, I think safety is a critical guiding value.
Similarly, related to that notion of procedural implementation, I think you have to really think through the suitability of an intervention given the available resources.
You know, there are interventions that if I were implementing one on one with a client, I would almost always pick this one or discipline, but guess what, this individual is not gonna have Linda LeBlanc implementing their intervention, and the whole team may not have Linda LeBlanc around beyond a few days as a consult, and so I have to be very mindful of what are the available resources in the environment, and that I cannot overtax those resources, because I'm not in it to make a little bit of change for a small amount of time.
My goal is to make long-term sustainable change that's going to move towards improving quality of life, keeping everyone safe, and having that intervention be successful for a long period of time.
So that notion of considering the available resources, it could be the student teacher ratio, whether there's an aid in the classroom, whether there is available and extra space where the client could earn a little bit of time away or what have you.
All of those details of the available resources, I think need to impact your selection of a treatment.
Now, another thing that's critically important to me is what is the client already able to do.
I have never met an individual no matter how impaired we might think of them as being that did not have some strengths.
Maybe even their behavioral rigidity is a strength because of their persistence.
So it's important that you think a little bit about a variety of behavioral patterns and how those things can be conceptualized as the client's strengths, and use that information to guide what your next steps might be.
And then, of course, it's critically important to prioritize what are the skill, needs and educational opportunities.
That's kind of why we're in this game in the first place.
We're kind of doing this type of intervention for escape maintaining problem behavior in particular, because we want to not only minimize problem behavior, but we wanna give ourselves the opportunity to be more robust in our efforts to teach and establish new skills.
So those are the guiding values that I distilled were controlling my decisions, and not necessarily in this order.
Sometimes safety is not an issue because you're talking about someone who's small, sometimes you have great resources and these other kind of guiding values are really driving the decision making.
In other instances, it might go differently.
It's gonna really, but these are the things I am looking at in any given situation when I am consulting within an educational system to assist them with problem behavior.
So, the most important thing is that you reflect on your own values.
If all of these values feel right to you, then these questions are going to likely seem intuitive and kind of you'll see why they flow in the order that they do.
If you have different values, like I really care most about this or that, then you may need to modify the decision tree based on what you think appropriately guides your consulting efforts.
But in the absence of strong feelings and that other values are far more important than these five, I'm gonna give you a model that you can use.
Shelly, do we wanna pause now? If there have been any questions that I can answer.
There was one question, and I have a feeling you'll probably come to this as you go on.
But a question about a real world example, particularly some ways that you might be able to reduce the motivation to engage in those behaviors that escape or avoid situations.
- [Linda] Yes, I will certainly be.
As I talk through each intervention, I'm going to talk through a variety of different situations.
I'll show you an example from the literature, but I'll also talk you through several instances in which that intervention was one that I felt was right and why.
- That's perfect.
And just for folks who may not have participated in our webinars in the past, just to let you know, that if you click on the chat icon at the bottom of your screen, a chat box should pop up for you if you haven't already done that, and if you do have a question as Dr. Leblanc goes along, you can just type your question in the chat box.
I'll keep an eye on the chat box and then when we have some breaks along the way, I'll pop in with your questions.
So, right now that's all.
- Okay, great.
Well, then let's move on.
So, this is the clinical decision making model that was distributed in an article for you, and I view this as a series of questions and answers, and, again, is important not to use this as an expert model.
Now, certainly, the behavioral analyst is bringing a lot of knowledge to the table, but you want to try to answer these questions in collaboration with the implementer of the intervention, and, you know, there will, and I find that the conversation and the process of going through these questions one at a time is in itself therapeutic upon the environment.
So, it may be that some of these things just have not been thought about before.
It may be that an educational environment feels like no one has listened to them about the safety issues or their concerns or resource issues, and it has lead them to be resistant to you're just gonna dump the behavior intervention plan on me and I know I'm not gonna be able to implement it.
So, don't, say behavioral analysts out there, don't think of this as something that you sit in your office and circle the answers for yourself.
Think of it as I'm going to maybe come up with my own answers, but I'm gonna actually go through this decision making model with the implementer, that you're kind of jointly thinking about what's gonna be our most likely effective for the effort that we have to put in, and we're gonna select among the good options together.
If you take that approach, I think it really transforms the experience from I'll let you know what I develop next week.
So, each time you ask a question, if you answered yes, you're able to go onto the next question.
If you answered no, you've identified a barrier that leads to a specific intervention being the place where you need to start.
Now, barriers can be overcome.
So, your goal is always to turn that no into a yes, perhaps as a step one of your intervention, and then be able to continue moving down the series of questions.
If you're able to answer yes to everything, then you get to ask what I call the gold ring question.
What's the most important clinical or educational goal that we need to teach you, and let's do it, because every intervention is an option.
So, just because you're able to move on beyond a here's your best option given this barrier, it doesn't mean it's no longer an option, and some of the interventions early on are quite effortful, and so you only want to use those interventions if in fact you don't have an option that might be lowered down in the model.
Whereas, if you're able to go all the way down the model, you were in a position to use some of our most valuable interventions like functional communication training and differential negative reinforcement.
So, what I'm gonna do is walk you through each question.
So, here's the first question, and I think it's, for the most part, if this question's not in the algorithm, we kinda just scoop past it, and so I think it's important to always ask this question, and I encourage people always say no, always answer no and then turn it into a yes for your good efforts.
So, the question is this.
Is the curriculum appropriate, and is instruction optimal? Are we doing all of the very best instruction on all of the skills that really matter, and we haven't missed any prerequisites along the way.
So, to be honest with you, there's always something that could be better, which is why I feel comfortable saying always say no to this, and then make things a little bit better and then you can turn it into a yes, but always start with that no and really do that careful analysis of what you're teaching and how you're teaching it.
So, let's take a look, and if you do say no, your best option becomes curricular and instructional revision.
So, what is curricular and instructional revision? Well, with curricular revision, we're talking about what you are teaching.
Is the task too difficult? Is it too easy or boring? Because either one of those, above instructional level or below instructional level can lead to the instructional context being more adversive and off task behavior, and then escalating problem behavior.
So, each of those are really meaningful questions to ask.
Making sure you're really having a vast majority of the instruction being right at what we might call instructional level.
Now, another question that I always ask is if we're working hard to teach a skill, we're having problem behavior which means it's clearly an adversive instructional experience, we better make sure it's really meaningful for their future before we invest a whole lot of time and effort into teaching that.
Now, let's say what you're teaching them is something that is gonna be critical for all of the rest of their skills.
Great, that is meaningful.
So let's say what you're focused on, maybe this an individual with significant cognitive impairment, and you're focusing on some very important adaptive skills.
So, you gotta brush your teeth, you gotta take care of your personal hygiene and toileting, and even though this is not a fun experience, it's absolutely critically meaningful for their future, you have to keep that in there.
It's got to remain as part of the curriculum.
But I will say this.
Too often I have encountered students that are beyond what we might call the early learning window for early intensive behavioral intervention, and that window kind of in the even one-year old these days, all the way up to seven or eight years old, we're teaching a lot of skills that are learning to learn skills, and we are trying to teach a lot of them because we're trying to catch this individual up, and to be in a position to learn for years and years and years to come, all the other things are gonna learn.
So, for example, imitation is one of those core learning to learn repertoires.
Matching is a core learning to learn repertoire.
But here's the thing.
Being able to sort things is a form of matching, but if you're working with a student who's maybe 10, 12, 14 years old, and they have not learned basic matching skills, even with visual visual stimuli, you need to think about the fact that matching or sorting in and of itself is not necessarily meaningful for their future.
Learning to be able to identify their name from others.
Learning to be able to use maybe a picture, change communication system and to recognize highly preferred items in a corresponding picture, that's meaningful for the future, but just sorting or matching kind of random names probably is not gonna build a meaning repertoire that's then gonna turn into a thousand more meaningful repertoires for them.
In which case, you might choose to put your time into getting to very strong skill sets on some other skills.
Another thing to think about in curricular revision is to whether you're missing prerequisites.
So that is let's say you're trying to teach some receptive language skills or a listener responding might be how you refer to it, where the person says something and the individual maybe finds it or does it in response to what the person says.
What I found more than a few times is that there is a missing prerequisite for simple auditory visual discriminations, and the team is already working on conditional auditory visual discriminations, and there's an absolutely core prerequisite skill that's missing completely.
There's no way you're gonna get those conditional auditory visual discriminations if you don't already have simple auditory visual discriminations, because a conditional discrimination requires multiple sequential and simultaneous simple discriminations.
So if you can't do the simple discrimination, auditory to visual, we don't have any chance at all with the conditional.
So, all of these things would be looking at what you're trying to teach, and why that might be too difficult or too easy for them.
Now, the other side of the coin is instructional revision.
So, that is not what we teach, it's specifically the procedures we're using to teach.
If the pace is too fast or too slow, that can make the instructional environment more adversive.
If the error rate is too high, let's say you're using a least to most prompting procedure, and this is a new skill, a difficult skill, and basically the person has to make multiple errors before they get prompted for the correct one.
This could mean they're actually engaging in a lot, a lot, a lot of responses, and a lot of errors, and that error rate, that hearing no or kind of getting the same task, like represent it, represent it at the next prompt level can be part of what makes the instructional environment less adversive.
So then in that instance you might decide we are going to flip to an errorless learning procedure with prompt fading, and then that's going to really dramatically increase the level of reinforcement and decrease the number of errors.
All of these changes that you would make, that is based on your answer to this question, are all gonna be designed to optimize what you teach and how you teach, and will decease the adverseness of the instruction, either because the pace is more optimal, the prompting procedure is more optimal, the tasks themselves are more optimal.
Here's another kind of real world example of what I've seen.
Sometimes you happen upon a situation where you're trying to teach a specific skill, but the response format is kind of bringing to bear other repertoires that are also fragile.
So, one of the things that is often said is that K through second, third grade is about learning to read.
After that, you're reading to learn.
It's why we put so much time into teaching people literacy skills, because their ability to read is gonna open up whole new worlds of venues for learning.
And so, some of those common tasks that might happen from third or fourth grade on can be very difficult for maybe an individual who has their higher functioning, they're kind of near grade level, they've got some of those literacy skills, but now, maybe with the diagnosis of Asperger's, now you're in fourth or fifth grade, and even though you have a lot of great skills, more of your tasks require you to read something, generate a narrative answer and write it with hands writing.
But hand writing is a relatively fragile repertoire due to fine motor skills being problematic.
Sentence construction can be somewhat difficult, and then, of course, readings just maybe gotten to readiness, and so it feels like the task is write about what you just read.
What that's actually kind of three very difficult tasks all worsened by a hand to writing as a response format.
So then, you know, what might you do? Okay, we're going to read the paragraph, and then I want you to say your answer out loud, and we're gonna type it.
Maybe I'll even type it for you and then you edit it.
So, you're decreasing some of the response effort for the whole task on the things that aren't critical.
What you most want is them to read the passage, get the right information from them and generate a meaningful answer based on that.
Whether they say that out loud or write down something on paper is not necessarily the critical reflection of whether they were able to read and answer the questions.
So, changing that response format can be a modification of curriculum instruction in that way that really minimizing that MO, it decreases the adversiveness of instruction and minimizes that motivation to escape.
Last question I always ask is even we're generally doing good instruction on the right things, is everyone teaching the same way.
Because that contrast that can arise for someone who is maybe differentially reinforcing correct responding versus someone who's not, even that can really make one or the other's instructional procedures very adversive, so that's really an issue of is everyone following the instructional procedure the same way.
And if not, is one of those instructional procedures resulting in a lot less problem behavior.
So, if you're getting this kind of all over the place, some days are great, or mornings are great, afternoons are not great, that kind of thing, you wanna look at is it implementation related, is it that there are these harder tasks in the afternoon, you're kind of doing that functional assessment of what is it about what we're teaching and how we're teaching that might itself be contributing to the aversiveness of the task.
And that really is a functional assessment approach.
So, I'm gonna walk you through this article by Dunlap and colleagues that kind of kid a functional assessment of disruptive behavior for a few different students.
And they used that information to modify and make curricular revisions.
So, first, what they thought was this student, that they were working with, might have more disruptive behavior during certain kinds of tasks that require different fine or gross motor abilities.
So, what they did was they did sessions where the tasks really involved fine motor or gross motor and they just alternated fine motor, gross motor, gettin' the data points, and overtime what you started to see is more disruptive behavior with fine motor tasks than gross motor, and correspondingly, here's the on task behavior, and you see pretty steady rates of on task behavior with gross motor but highly variable and decreasing overtime rates of on task behavior with fine motor skills.
Okay, now we've identified something about the tasks themselves that are a part of what's making it aversive.
Now, here's a second hypothesis.
The duration of the kind of task itself.
And what they looked at was relatively short duration, long duration.
Short duration, long duration, and what you can quickly see is that disruptive behavior is significantly higher during a long duration task, and in fact on task behavior is much lower as well.
So we're having a whole lot of problem behavior and not much learning when the task duration is longer, whereas short duration tasks are not producing that same effect.
Well, we all know in the long duration tasks can be chopped into short duration tasks or they wouldn't have been able to do this analysis.
And so then that leads you to, well, if I've got many more short duration tasks with frequent breaks, that change in kind of how I'm arranging the instruction might make a big different.
Now, here's hypothesis three.
Does it matter if it's a functional task or kind of an analog related to that skill, but not clearly a meaningful naturally anchored task, you see the analog tasks are problematic on both fronts, and then they also did an analysis of what if you get to choose some aspect of what we're gonna count, which worksheet we're gonna do versus no choice.
And what you see is no choice is inconsistent, but does result in more disruptive behavior whereas choice leads to more consistent on task behavior.
So several little pieces of this owe a kind of combining together to really increase the aversiveness of the curriculum and instruction.
So, here's their intervention data.
And what you see in the gray shaded portion is on-task behavior, and what you see with the squares is disruptive behavior.
So highly variable on both fronts.
The disruptive behavior and the on-task behavior in both the afternoon and the morning.
They used a multiple baseline design and this was the only change.
They revised the curriculum based on their hypothesis, so they did not have a lot of consequence based interventions.
Instead, they altered the curriculum to decrease that MO for escape and decrease of the aversiveness, and what you nee is an immediate reduction, basically to zero, of problem behavior and big increases, although there's still a little bit of variability on the on-task behavior.
So, one of the real strengths of that kind of intervention of the curricular and instructional revision is the fact that it produces an immediate suppression in a problem behavior.
So, if you've got safety as an issue, you need a fast result, and this gets you a fast result.
Now, I wish I could every time you do curricular instructional revision in your data, you're good look this beautiful, and you never even need to go on, like you just fixed it, ta-da, and no more problem behavior.
That might happen, and if it does it means you're done, simply by really having focused on the quality of life and instruction for that student.
Get all of the bang you can for that intervention buck, but if in fact it does not result in as complete a reduction as you need, then you move on and ask yourself the next question.
Can the environment tolerate any level of problem behavior? If the answer to that is no, and I have been in this situations, where the answer is no, then you only have a couple more options.
First, make darn sure you did curricular and instructional revision and got everything optimized.
Now your better options are demand fading or non contingent escape.
Now, what are some instances in which the environment is saying we cannot tolerate any level of problem behavior? In instances where it is very severe and high-rate SIB, I have seen this as a meaningful answer and had the treatment partner tell me that this isn't possible.
So, for example, I've worked with individuals with eye and head related poking, gouging or pounding, and there's just real danger with severe SIB that you're gonna detach a retina, you're gonna lose an eye, you could cause brain damage.
You know, head to wall are a very hard surface.
If that occurring at high rates, you've just gotta get that level of problem behavior down right now really fast in the amount of time it might take for extinction to work is not gonna be safe.
There's a real safety issue.
Similarly, if a student, a client is very large and powerfully aggressive, people have been put in a hospital, people have really been hurt, people are very, very frightened, the environment may feel like we can't have, we can't have this kind of aggression occurring, or we can't keep the student in this classroom in this environment.
That safety issue is just we can't do it.
So, I don't take this question lightly.
There are instances in which it is not safe for the problem behavior to occur.
If the person is already in some physical restraints.
Arm, splints, head gear, I am gonna be really worried about this, and if the team is just frightened of the student, I'm gonna be worried about this, and a shared benefit of both demand fitting and non contingent escape is it focuses heavily on that motivation, just like curricular and instructional revision.
So it really decreases that motivation to engage in the problem behavior, to get the functional reinforcer of escape.
So then you're in a great position to start with some of these three, and you might put yourself in a position to later feel like a less intense form of problem behavior or very infrequent problem behavior won't jeopardize the person's ability to be out of restraint or remain in the environment.
So, let's go to demand fading and concontingent escape, which just entered the ballpark as doable interventions.
Demand fading starts with elimination of all demands, pretty much down to zero, and what I often find is if you've got these safety issues, that step's already been taken care of for you.
Instruction has pretty much been put on hold, the student may be spending really long periods of their day just not able to be in instruction because of the worry of harm to self or others.
The next steps of demand fading involve slowing fading the demands back in.
Now, the real benefit of this is that produces an immediate reduction in problem behavior to zero very often, and over time, as you're just gradually, gradually increasing those demands, it's gonna increase tolerance to the demands.
Now, it will be important that you already did that curricular and instructional revision and that the demands that you're introducing are meaningful ones, because everyone that you introduce is gonna increase slightly that chance of problem behavior.
So, one of the requirements of this kind of intervention, demand fading, is that you have to kind of fade those demands in, so you're gonna go and introduce that one demand, two demands, three demands, four demands, it's gotta be slow and steady, and it may require some additional involvement of the behavior analyst to manage.
Are we ready to move forward, or are we ready to add in another one.
One advantage is you can use this without extinction, and what that means is if the really severe aggression occurs, and the person feels like they need to back up or run away in order to stay safe, that's actually problem behavior resulting in the functional reinforcer of escape, which means you don't have extinction in effect, but this intervention generally still works because you are catalyzing so much on motivating operation and slowly increasing the tolerance for demands.
Now, a non contingent escape is also heavily focused on the motivation for escape, and so what it basically involves is scheduled breaks.
Fixed time delivery or variable time, but that's just harder to do, so most people do fixed time delivery of the breaks, and there's no response requirement.
Even if you didn't get a single one of the task demands correct or independent or both, it doesn't matter.
When the work time is up, timer goes off, break is provided.
Not contingent escape also has the advantage of producing that immediate reduction, because you start off at what's called continuous non contingent escape, which means no task presentation.
So this is somewhat labor intensive at first, and as you're moving through the schedule thinning, but you can also use this without extinction.
It's generally done with extinction, but you can, again, it fits that big aggressive client that's producing that fear response, engage in the problem behavior, and then the staff backs off.
That is escape.
You wanna minimize how often that occurs, but because the motivation is so much lower, you're going to see a much lower rate of problem behavior.
So remember, you're usually doing this when you have very intense aggression, very high rate, maybe self-injurious behavior that could jeopardize limb more, or sensory capabilities or what have you.
So I'm just gonna walk you through a couple of examples.
Gary Pace and Brian Iwata have done some good work in this area that you can refer to.
Now this one, I'm gonna talk with you about.
Just because their data really showed this effect, but you can obviously look back at this article as well for a protocol.
So, with this one, they're working with an adult with a brain injury, and in task situations a lot of obscenities occurring, and so here what you got is the functional analysis, and what you can see is here's the demand condition compared to a reprimand condition and control, so definitely differentiated in demand, and here's the number of demands that are being presented per minute.
Right around two to three every 20, 30 seconds.
Now what they do is demand fading.
So, they decrease.
Oh, I'm sorry, this is demands per session, and it's around 45 per session.
So then what they do is they decrease the demands per session down to one per session.
No more demands.
And look at the suppression of problem behavior.
So here's your obscenities and they're at zero.
Then gradually they increase to four or five demands, six, seven demands.
Here's a reversal where they try to go back to demands at a similar level to baseline, and a problem behavior comes up.
They go back into demand fading, lower the number of demands to way beyond continuing that slow steady fading, and problem behavior remains low.
Another reversal and look at this.
They're actually able to get the number of demands per session up to exactly the same level as in the functional analysis baseline.
They have not changed the contingencies.
It's just that slow, steady progression is increasing the tolerance for those demands without the problem behavior.
So this can present a good option.
Now, this is a study where they were treating SIB.
I actually, I think was a data collector with this particular study, and this young man had a high rate, head directed SIB, and we just were not in a position, we needed to produce pretty immediate suppression.
You know, it was challenging enough to do baseline, and fortunately behavior became pretty efficient at super minute to get that 30 seconds of reinforcement, and went into continuous non contingent escape, the whole session, using immediate reduction.
Even increasing to 10 seconds of work and the rest of the minute is a break.
You can see resulted in just a little bit of SIB.
So, one of the things to keep in mind about non contingent escape is you very often have to start at continuous and have very slow progression.
Now, if you have very high rate SIB, you may not have a lot of other good treatment options anyway, and you're in a jeopardy situation, you need to do something.
So this does give you a good option to start with, and then by the time you get to five minutes, 10 (coughs) of tasks, now maybe you've turned each of these single instances of SIB into a less jeopardy situation because some healing has occurred, and then now you can maybe even answer a yes and move forward with questions.
So those are just kind of real world examples.
Now, let's say you been able to answer yes to the prior questions, but you're in that situation where there are virtually no demands to which the client already complies.
So, are there any demands to which the client already complies? Yes, but not these.
If the answer to that is yes, you can keep going and ask more questions, because you do have some tasks that the client will comply with, and you can use those to kind of build some momentum and potentially get yourself to a position where you can do functional communication training, but if the answer to that is no.
Like, if I asked this person to eat your favorite cookie, but it came across as a demand, they wouldn't do it and they'd engage a problem behavior, then you are in that same boat.
Curricular or instructional revision better really matter that you gave those demands and demand fading and noncontingent escape are really still your best options.
But let's say you did get a yes.
There are some demands that if the student already complies with, we done curricular and instructional revision, they are doing some of that, but we still have some problem behavior, then you ask the question, is time away from instruction tolerable? Is there this capacity for the individual to be able to either ask for a break and get that break and not do their instruction? Sometimes classrooms are arranged where it's a situation where it's not feasible.
Your student teacher ratio is out of whack, the student does not have an individual aid.
And then that means like functionally we're now gonna be able to, he's not okay just staying in the classroom, he wants to get out of the classroom or he wants to whatever, and we can't really do that because I can't leave all the other kids unwatched while I go with him.
That is meaningful barrier, we're gonna have to come up with some different options.
What I often find is that the gut is, the gut reaction is no, he can not be away from instruction, and when I ask a little bit about like what makes that hard, it's more of like a values kind of thing.
Like, I want him to have a lot of instruction, and then when you explore like is he missing instructional time now, the answer is yes, he's been suspended from school, sent to the principal's office, go into a timeout room.
Like, their are all these times.
Or he's just doddling around all the time and not really doing any of the instruction.
In which case, he's kind of already having some time away from instruction, and so it's why you don't want to try to answer these questions as the expert.
You wanna really talk with that to the implementer.
Talk them through.
Well, maybe are they already in time away from instruction, maybe when they're going out of the classroom or this, that and the other, could we tolerate maybe some brief periods that occur based on their behavior? A lot of times that turns this no into a yes.
If it's not a true man-power barrier, but more of a I don't think it's right that they have that time away from instruction.
So, if the conversation, you can turn the no into a yes, that's a yay.
But if it doesn't, then you have just a couple of options.
Activity choice and incorporation of choice or extinction.
So, all of the procedures we've talked about prior to this point can be implemented with or without extinction, but generally work better if you implement them with extinction, but they can work even if you do not explicitly include extinction.
Well, let's talk about activity choice.
Activity choice is one of those middle column interventions that is addressing the motivation.
So, it could be that what you wanna work on is counting, and you then look at what kind of choice can I incorporate into this activity that doesn't alter the curriculum or the instruction.
Maybe I have three different sets of bears, and we can count blue ones, green ones or purple ones.
What do I care? You're still gonna be counting.
The counting behavior is the same, the materials differ.
Maybe I have different versions of worksheets, and you just happen to be obsessed with Thomas the train, and instead of using the boats worksheet I'm gonna use a train worksheet.
Both of those are just easy ways to offer choice that may not have been offered before, and, of course, it matters that what you were working on is important and your instruction is optimal, so it's always lather, rinse, repeat, go back to question one.
Sometimes you can also do, are here are our different things that we're gonna have to do in the next hour, let's pick which one we're gonna work on first, next, et cetera.
I always offer this caution if you're gonna do that.
Students are gonna do exactly what you would do.
They're gonna pick their easiest, most favorite one first, then their next easiest move is favor is second, and as the longer amount of time working goes on, the more non preferred the tasks get, and so it takes real self control to actually do the adversive ones first or to intersperse.
So, what I always say is don't just let them pick, let them pick, let them pick, let them pick.
Let them pick, then you insert one that you know might not be as preferred.
Let them pick, then you insert one, so you're kind of assisting them to not get in that hole of the longer I've been working, the worse and less preferred that these tasks are, which is what you'll often see blows up the afternoon of a school day.
So again, this is that motivation kind of intervention effect.
Now let's talk about this is a fully antecedent intervention, so you do not have to alter the consequences for problem behavior.
Extinction is certainly possible, you can add that in and no longer allow, you know, no longer send them to the principal's office, or what have you.
But if, let's say the school is not willing to get on board with something like that because these is an antecedent intervention and you get that suppression based on motivation, you can have extinction or if there is severe problem behavior there might still be sent out of the classrooms, into the principal's office or what have you.
One advantage is there is no missed instruction time, especially if you implemented this with extinction, and so this is a really appealing intervention, either so low or in combination with some of the other ones.
Well, let's talk about extinction.
The biggest benefit of extinction is it works like a charm.
If you no longer provide a functional reinforcer, and you have really identified a functional reinforcer, you will see an extinction curve, and problem behavior will eventually get down to zero even you don't change anything else.
So, what this involves is you simply maintain the instruction, keep presenting those demands and prompts, no visible change in the environment contingent upon problem behavior.
Now, the difficulty is this.
Extinction can be very difficult to implement, right.
If someone's hitting at you to not move away at all, to not give that even minimal interruption in the instructional sequence can be very difficult, it's more difficult when a student is a little bit bigger and stronger, and even if you can do it, it can be very unpleasant for both the school staff and the student.
So it is possible but is difficult and unpleasant.
You better have very well trained staff.
It's one of those things where it's like you better have a big staff, well-trained staff, and here's the thing.
Extinction implemented at 99 percent accuracy has another term that could be applied, and that is differential reinforcement of more extreme problem behavior.
So, even one out of 100 reinforcing that problem behavior can lead to resurgence of that problem behavior, and just you've just gotta make sure it's so difficult, but you had better make sure that that implementation is 100 percent on the mark.
So my preference is generally for extinction as a supplement to these other interventions, as opposed to a substitute.
Now, this is in that left column, which means it does not address the motivating operation.
The instruction is still as aversive for that student as it's ever been.
They just simply can't use problem behavior to get out of it, so it really does not do much for that quality of life and safety.
If you implementing into isolation, and so, again, in conjunction with some of these other interventions like non contingent escape with extinction, activity choice with extinction, maybe if we get to functional communication training with extinction, it's a good supplement.
Extinction bursts are possible, and you need to honestly talk with the treatment environment about whether they're gonna be able to handle that extinction burst, having in those other interventions that do address motivation typically decreases the likelihood and severity of an extinction burst, so it's another good reason to have extinction as a component, but with another big component in there for one of those other columns.
All right, so let's say we got yeses.
We got yeses, we got yeses.
Okay, now that you mention it, time away from instruction briefly is tolerable, it's probably already happening, then we get to ask the gold ring question.
Which is the most important clinical or educational goal? If the individual does not have a communication repertoire to be able to ask for help, a break, it hurts, it's aversive, please stop, establish that functional communication response.
This is quality of life.
We don't want people going through their lives without the ability to clearly communicate when something bad and aversive and miserable is happening.
So, then FCT becomes so important.
Could be they can already tell you all day long I don't like this, I want a break, and that's not the issue, and instead there are other curricular targets that are most important.
There's also this issue of tolerance of an aversive event, and I'll talk about the situations in which I think DRO is appropriate and useful But let's talk a little bit about.
You're picking the goal and that helps you pick your treatment.
Which prosocial behavior do you need to target? If it's communication, FCT, ask for that break, ask for that help, ask for something to make the pain stop, whatever it is that's making it aversive.
If it's other skills, like, they can ask for a break, but they do it too often or they don't do it even though they know how, and we gotta focus on these math skills, these handwriting skills, whatever it is, then DRA is gonna be a better option, and what they do is they earn breaks by learning or compliance or what have you.
Let's say the important thing is really more tolerance.
So, sometimes we have done everything we can, but this is still an unpleasant thing that we have to tolerate, and I need you to tolerate it without problem behavior.
Things like a haircut, personal grooming, your nails, dental work, it's not fun, it's never gonna go fun.
Like, we can try, you get to pick if you wear a blue cape or a red cape, but it's just still going to be aversive when you hear the sounds and then we put headphones on you.
We can do a lot of things, but ultimately you're gonna have to tolerate some of that experience.
We can make it as least aversive as possible, but it's still aversive.
That's an instance in which I do think differential negative reinforcement of other behavior is important.
So the reinforcer it not an edible it's not praised, it's a break, so that is, you know, the nail clippers are right next to your nails, and for two seconds no movement, okay, we're done, a little break, or that drill is there and it's one second break for 10 seconds, one second, break for 10 seconds, and then you're gradually increasing the amount of time they need to be still in order to get that functional reinforcer on the break.
So, you have good options, different instructional goals, different environmental constraints, different client characteristics or severity of the bahavior, safety issues, are all gonna maybe leave you to pick certain interventions, especially as to your first intervention.
So, remember, with every question that you answer as a yes, all of the others remain an option.
So I typically do this as a collaborative.
Let's me and you answer these questions together, we wanna get them as many yeses as possible 'cause it gives us the best options, but if we hit a real meaningful no, we're gonna talk about that.
Then once I've answered all the questions, I and the person who's gonna be implementing, teacher, parent, we look through the table in that article and every pro and con, and we pick which ones based on the pros and cons we think are gonna be most useful, and you know sometimes I'll say let's start with non contingent escape until we at least can get to the best point and then we're gonna add in functional communication training, or maybe let's have activity choice and extinction, and depending on how that goes, once we got enough instruction going, maybe then we can add in whatever.
But almost always have some focus on curricular instructional revision.
Well Shelly, let's take another break and see if anyone had questions that I can answer before I kind of walk you through the next portion.
- [Shelly] Sure, thanks.
There were a few questions in this section.
So, one of them goes back to the curricular and instructional revision, and one of the things that folks run into is that's often a place where we have to stop and to spend sometimes a great deal of time, and it can be easy in some ways and difficult in others.
So the easy part can be finding the right goals and identifying them, but the hard part can be explaining that what was being done previously was wrong or maybe it wasn't the best choice at the time.
So, we're wondering if you have any suggestions for having tactful conversations with teens about that.
- Yes, I do, and it's never easy right.
Like even if you were able to have a very successful tactful conversation, you were working for it.
It's just that you managed to frame it the right way, and could have said what I typically do is I frame it as a success and as a success without all the hullabaloo of a drawn out behavior plan.
So what I might say is, first I'm gonna try to pose it as functional assessment where I'm trying those different versions of a task.
Short versus long, so I've got data that shows with some clarity or let's say it's somewhat an early learner, I might do that task analysis portion of the (audio cuts out) which kind of shows you is this prerequisite there.
If not, you gotta teach that one and kind of watch should do these prerequisites.
And I also kind of present it as thank you so much for letting me explore with some of my assessments.
I have some great news in that I think I figured out why some of these tasks might be harder than we thought for him.
Let me show you my data.
Oh my goodness, look at this when I tried this and then tried this, what a big difference I got.
So I'm really letting the data speak, and just kind of like we just asked a question, and the good news is we found some answers and then it's like I really think if we make these (audio cuts out) changes, that's gonna be more than half the battle and we don't have to worry about bringing an extra step and restraint, and maybe some other things that might be on the table if things stayed the way they are and of course no one wants them to stay the way they are.
So, step one looks to me like let's move in these directions and focus on we accidentally missed one before, and I always, always, always use a we.
You know, we made that choice to do this assessment and look at some, and thank goodness we did because let me show you what I think we found and this is how that like I'm always a fan of easier as long as it works, so let's try some easy first, but, of course, if this wasn't the real issue and you make these changes and there's still problem behavior, I'm gonna have other options that we can try.
So I kind of tend to frame it that way, and let the data kind of show those effects, and so it's not that it's...
It's never my opinion that the curriculum or instruction is bad.
It is simply I tried these couple of different options and you're not gonna believe what a difference I saw who'd have thunk it that look at this, and look at this data.
So, I generally approach things that way because I find that it's not my opinion versus your opinion, it's just here's what we saw and this one sure looks like the session I'd rather be in, how about you? So, it's easy to agree with.
Is there another question? - [Shelly] Yes, thank you so much for that.
The second question was when you're using noncontingent escape as your intervention, what should or would the learner be doing during that break? - Yeah, so that's a great question.
It will kind of depend on how long the break is, and that learner.
So, you often start off with very little instruction and a chunk of break.
It's gonna probably be at least 30 seconds, but it literally could be the rest of a three minute interval or something like that.
The longer that break interval is, the more important it's going to be to kind of enhance that break that you're not having to do the instruction, that's great, but let's not have you get bored and go to your own devices, 'cause you can start to see other functions of problem behavior like I'm bored, I want some attention, or you start to see a lot of stereotypy which could be disruptive in the classroom.
So what I find is if the break is gonna end, like sometimes you're barely like, okay, let me put away the material and just stretch for a minute and get comfortable and maybe here's just a little quick high five, and now it's time to put the materials back, then don't worry about adding something in during your break.
What if it's gettin' a little bit longer, then you can either go with some social interaction if that's appropriate.
If it's in the classroom and that's not really viable, then you can't do that, but if can add in some kind of supplemental, you know, even if it's like here's your stress ball that's quiet, but it's a little sensory stemmy thing, you could do that.
What I find is if there is multiply controlled problem behavior, it's definitely escape, but there's also this attention or tangible function.
I use that to guide how I'm gonna supplement my noncontingent escape or my FCT or DRA, so it's not only a break, but maybe it's a break with, and I'll usually go with maybe a moderately preferred toy as opposed to the most favorite, 'cause then when I try to take it back from you to introduce the next demands, I could be in a little bit of a problem again.
When possible, I like to add in some social interaction to that break.
All right? - [Shelly] Yeah, which relates very closely to the third question and maybe you may have something to add or you may not, but do you have any advice on how you can minimize reinforcement and attempt to implement extinction when there is that combined function of escape and attention for the problem behavior during instruction.
- Yes, so let's say you've got like escape is really the bugaboo here, but you have a little bit of a tension maintaining behavior.
Like, if I get your goat, I sure will take that as a secondary reinforcer, and I tend to really try to avoid the extinction protocol as the primary driving force under those circumstances, because it is typically the case that when you're providing the escape inadvertently, that reaction is a double whammy of an escape and attention reinforcer.
So then I really wanna bring the big guns on that decreasing the motivation, that middle column or right column, and then you just work hard with someone on in those few instances there's still gonna be some where problem behavior is occurring and you do have to implement escape, that what you're gonna do is voice neutral, face neutral, next demand.
And it takes a lot of work.
So, that's one thing that I do, and then the other thing is do your best to enhance the break reinforcer interval with those functional reinforcers not contingently or even use that as the opportunity for some FCT for those.
So, let's say we're gonna be working and if you ask for a break you get it, and they ask for a break and you say all right, take a break, and you wait two seconds and then you say did you wanna hang out with me or have a toy, and now that triggers the opportunity for the other FCT.
So, all of those things, I think can really make the intervention a little bit more successful if you have multiply controlled problem behavior.
- [Shelly] Great, and if you have time, there's one more question, and I think I'm understanding it correctly.
So, if a learner is engaging in some problem or challenging behavior in order to access a tangible item during instruction, in addition to maybe an escape maintained behavior, would you provide them with that tangible item during the break or would you avoid that? - That's a great question.
And let me say this.
I don't know that there are tons of studies out there that are guiding.
This is just my years of doing it.
What I probably would do, is I would work, I would really work my reinforcement contingencies, and I would also outside of the instructional context really get to strength with you can have it, my turn, you can have it, my turn, and really kind of get that exchange response occurring smoothly outside of the context of the demands which is gonna drive up the motivations for everything, but you get that going and you got a lot of praise, maybe you even have some who knows what.
Edibles, whatever it takes to get that smooth, easy, exchange occurring, then insert that into the instructional context where it's like, okay, you either earned your break, or it's fixed time break.
Hey, did you want this? Or noncontingent access and then that exchange occurs, but rather than kind of trying to also work on that in the context of you just had to do a bunch of work and you're about to do a bunch more work, I get the exchange acts when in what I might call a low jeopardy situation where exchanges don't always result in here's a bunch more work.
The other thing you can do is this.
Like, let's say you have a DRA protocol, which means like you have to complete five tasks independently, and then you get your break, and you can ask for your toy and what have you.
When I say my turn, if you hand it over immediately no problem, the next time's only four.
You actually lessen the work burden just a little bit, you're still gonna be gettin' more done than probably when we started on this journey, but you're really kind of using those different reinforcement schedules to get yourself in a better spot.
- [Shelly] Great, thank you so much, that's all of the questions for now.
Okay, so what I wanna do is just walk you through an activity that you're gonna actually take away with you.
Again, I think it is best to do this activity and to do it jointly.
So you should have a handout for group activity one.
It is the escape decision making algorithm.
So what you would do is look through that algorithm, think a little bit about what you think the answers are for one of your clients or one of your students that has escape maintaining problem behavior, and then meet with the implementer.
Maybe it's the classroom teacher, maybe it's teacher and parents, however it is you're gonna do that collaboration and go through those questions with them, and the reason why I say go through of it ahead of time is because you may be able to predict where you're gonna get some notes, and whether those are real and meaningful notes, or whether you're gonna have that conversation in a therapeutic enough way to turn a no into a yes, and give yourself more options.
And I'll always start the conversation with the more yeses we can have, the more good options we have to pick from.
And if we're only stuck with one, fine, we're gonna do that together, but let's be in that position to try to give ourselves as many honest yeses as we can, and then what that's gonna do is it's gonna lead you to wherever it is you stop at a no or your last selection of a goal tells you the intervention that's probably best suited, but all of the others also remain an option, and so then you pick are we gonna do only this one, are we gonna do only FCT or are we gonna add in an extinction, are we gonna add in activity choice.
So that's gonna allow us.
So, your first activity is use the model at least one time, see how that collaborative approach works for ya.
Now, I wanna talk a little bit about implementation planning, and then I'll walk you through the second activity.
Because this really needs to be a meaningful peers activity, I'm gonna have you take it away.
All right, let's talk about making it doable.
Whatever it is you picked, all of them have, you know, the pros are it works well, or whatever.
The cons are maybe it's harder or it takes more time or it takes more expertise on the part of the behavior analyst.
So, there are logistics to be managed, and if you are thinking systematically about managing those logistics, you increase increase the likelihood that you're going to get good long-term implementation.
The other thing is this.
You're almost never able to start your intervention off at a level of treatment intensity that's ultimately desirable.
So, when you implement your plan, it's like, okay, well we just picked whatever, FCT.
You're at point A, and you can see just a little ways down the road.
Okay, we're gonna implement it, we teach it, let's go.
But from the perspective of the long term kind of environment, the teacher, the parent, that's their road, right.
It's gonna wind for months and months and months, and you may already be gone as the consultant or less of your time is available.
And so, you gotta plan what they're gonna be doing at point B.
In six months, a year, all day everyday, rather than just what is doable for short bursts.
Extinction may be doable for a short burst but is not doable all day long.
If you don't have other components.
So, now, you may have to start with something that feels like oh my God, that's too hard, I could never do that.
Well, you gotta start with success because there's no path forward from failure, that's kind of where you're already at when you're called in to do the behavior plan.
So you have to do something that's gonna be successful, and don't underestimate the A, you know.
Demand fading, you start with no work.
Noncontingent escape you start with no work or a few seconds or one task.
It may be very, very, very limited, and then you have to know where you wanna end up.
That's your point B, and you've gotta be reasonable.
And sometimes I kinda collaboratively negotiate a point B, and I'm thinkin', we're really only gonna get to point C.
So, it's that notion of like well could you provide a break.
Could you provide a break once every five minutes? And of course it's absolutely not.
We have to get to, it's gotta be no more than once a half hour, and I say, oh, okay, and I'm thinking that's gonna be point C or point D at best, now I'm starting at point A, and we're at a break every 30 seconds.
I'm gonna go to point B and then C, and I gotta map that out.
So, sometimes do you have to kind of be appropriately responsive, but still help someone kind of see it's gonna take us a while to get to that level of reasonableness.
Here's our first kind of check it out point, because here's the thing.
Most times people in that instructional environment, they can't imagine what that's gonna be like when there's not all that problem behavior.
They're struggling with it now, and so the notion of this is gonna be more effortful and what have you, they're almost kind of in the back of their mind imagining all of that effort superimposed on all this problem behavior, as opposed to that effort is in exchange for all of this problem behavior, and that it's probably actually going to be easier to have that effort than this problem behavior, and they have to kind of experience that, and the success point, and so what I often find is even if someone says I need to get to, he can only have a break every 20 minutes, by the time I've got it to five minutes they're like oh, this is more doable than I thought, and we end up with 10 minutes as an ultimate kind of it's okay point, or maybe we get all the way to 20 minutes.
So, you gotta have a plan for decreasing the intensity, and I do think this generally has to be run by the behavior analyst, because if you go too fast or too slow, it blows up.
Remember that stimulus fading intervention I showed you with the adult with the brain injury, the reversal was just going back to too big a jump in the number of demands, and it blows up.
So, slow steady, but also no stalling.
So what you do is you map out the steps from A to B, even if you know there's gonna be a C later.
Get an A, get a B, move forward slowly and don't stall, because of these kinds of either stimulus fading or delayed reinforcement fading or scheduled fading, part of why they work is that the next step is not wildly different than the prior step, and so it doesn't suddenly feel tremendously more aversive, and so slow and steady.
The more you stall at one level, the more noticeable the change to the next level becomes, and you use your data to guide you and make the decisions.
I always have a very explicit response criterion, and if I hit that criterion, you go up a step, and part of why I do that is so that people will be able to make decisions in the absence of Linda, but careful decisions.
So maybe it's gonna be an 80 percent reduction in problem behavior, or an 80 percent reduction in problem behavior and at least one unprompted functional communication response for a break.
Maybe that's or no problem behaviors.
So you're gonna figure out what level of problem behavior is an effective reduction, doesn't necessarily have to be perfect, and then if they hit that criterion, maybe your criterion's gonna be two data points in a row, three data points in a row, whatever it is that you're gonna say, they hit that criterion, you move forward a step.
So, here are the most common methods.
Number one, more work is required in the work context, so one task, two task, four task, six task, eight task, 10 task.
Notice they don't stubble each time beyond the first three.
It's a mistake to think of this like a progressive ratio schedule, because those jumps start to be pretty darn big.
Method II, is to have a delay between the functional and communication ask and the delivery.
So, this is typically called the wait period.
Zero, five seconds, 10 seconds.
You can find a lot of implementations like this in literature.
I'm gonna talk with you about something that is a better option that has been discovered recently, and then Method III is kind of that longer interval, either for the free NCR or the DRO intervention where it's how long they maybe have the break.
So it's the duration of the break itself that's changing in correspondence to the duration of the tasks.
So, here's an example of how I would set up the steps.
Let's say you're gonna do noncontingent escape.
The first thing is how do you know what point A is.
You can always start at continuous.
If it's gonna work, it's gonna work at that level.
But you can also look at your rate of problem behavior in baseline, and calculate either an inner response time or just a rate derived average.
So, let's say problem behavior was happening in a 10 minute session, you're having eight SIB per minute, and so you know that problem behavior's occurring not quite at one per minute, so a one minute non contingent escape interval, well it means some of the times you would, problem behavior would have already occurred, so if you really wanna get that suppression below that, you might have your derived rate, and maybe provide the noncontingent escape every 30 seconds.
Then, each increase in the interval is proportional.
So, if you look at this, I go from 30 seconds, not 31, 32, 33, 34, 35, it's gonna take us all year to get to point B, instead I increase by a portion of the prior value.
So, 45 seconds is a 50 percent increase from 30 seconds.
So, the additional 15 seconds is half of the prior value.
So that's a 50 percent increase.
Early on you're likely to have increases that are closer to 50 to 100 percent if the values are very small.
Now, let's look at going from 45 to 60 seconds.
This is a 15 second increase again, but now the percentage has dropped and 15 percent is one third of the value of the 45 second prior interval.
So, you keep moving up, and I generally try to keep all of the percentage increases between a 25 to 50 percent increase.
So, if you look at 360 seconds, is a 90 second increase of over 270 seconds, and 90 seconds is one third of the value of 270 seconds.
So, the values in and of themselves are the total number of seconds, are getting bigger, but the proportion increase is going down from the beginning towards the end and staying in that ballpark of 25 to 33 percent for as many of them as you possibly can.
Okay, so now let me talk about kind of what's newer in the literature that I think is a good way to go if you're using functional communication training.
If you use what is called a multiple schedule, you can enhance the effects of FCT with extinction such that you never have to move on to more intrusive consequences.
So, I think it might have been in 1998, we published a study out of analysis of all the FCT cases at the Kennedy Krieger Institute, and what we saw was FCT did not work particularly well with that extinction, especially once you move on from you can always have a break anytime you ask for it.
So, definitely use extinction as a supplemental contingency for your functional communication training.
What we found in that early set was at some point in the delayed reinforcement, you asked for it but you have to wait three more minutes to get it, we lost the big reduction in problem behavior and had to add in some kind of additional reductive procedure such as a punishment procedure.
And a more recent analysis, Wayne Fisher, the first author on that other study found, was if you add in this multiple schedule, they never had to move on to punishment in order to get to kind of that terminal point B where it's very doable.
So what do I mean by multiple schedule? It's alternating periods where reinforcement is available if you ask, and not available if you ask.
So, problem behavior's always on extinction, ask for it under a green stimulus, schedule correlated stimulus, ask for it when the green card is out, you can always have it immediately.
Ask for it when the red card is out, you can not have a break right now but when the switch is back to green, you can ask.
So, you're kind of introducing these schedules that are correlated with the different schedules of reinforcement and extinction.
This is a great way to manage too many asks for a break, and extinction induced problem behavior and it keeps your functional communication requests very strong.
So this was the first great creative study, kind of investigating this use of the multiple schedule.
It's a little bit complicated and translational, but the big takeaway point is this.
When they used a mixed schedule, no signals, you don't know if now is when you could have it or not, they saw more problem behavior.
When they used that multiple schedule signaling, reinforcement's available, reinforcement's not available, then what you saw was lower rates of problem behavior, and what you saw for the communication was that over time it became discriminated.
So, here are the requests under the multiple schedule.
They're asking all over the place, and actually asking more during the extinction period, 'cause they don't get it, so they ask again and again and again, and eventually the problem behavior surges.
But when it's clearly signaled as red, and green, what you see is they start asking pretty much only in the presence of the green.
So, include that multiple schedule, that's gonna help.
Now, depending on what you're doing, if it's very labor intensive, if it's a large student, very aggressive, if you've got complicating sleep problems, anything like that, if the treatment environment is unwilling to put problem behavior on extinction and they're using time out, and you know that's actually the functional reinforcer, things are gonna be hard, and the early days are the hardest, so you have to figure out what you might need to do.
Switch your staff frequently, do you mean multiple staff, do you have two to one for certain targeted times.
Let's say you're doing activity choice.
The heavy lifting there comes in that you have to prepare all your materials, and so maybe you've gotta redirect a little bit of help or additional time to be able to help with the materials preparations, so you have multiple opportunities for the client to choose.
Now, I've also, I guess I'm like don't be afraid to ask.
They'll either tell you no or they might tell you yes, but they can't tell you yes if you never ask, so don't be afraid to ask for extraordinary efforts like space to conduct a functional analysis.
Sleep problems can really exacerbate escape maintaining problem behavior.
I've asked for space for my student.
If they get off the bus clearly visibly tired and irritable, go take a nap.
First 30 minutes, take a nap on a cot in a quiet place, that's gonna improve the whole day.
Ask for it if it seems meaningful, you never know.
I've often found that principals spend the very earliest part of the school day walking the halls, working the system, making sure things are going well.
They're not in their office anyway.
Whatever it might be.
But be on top of your data, because you've really gotta use your data to manage those logistics in moving forward.
Okay, I got just one or two more points before we wrap up, and I'll try to stick around and answer some questions.
It's very important that you monitor both progress and procedural integrity.
You could have substitute staff or new staff that get hired, and people just need periodic retraining.
You know, I generally have a goal of at least 80 percent accuracy of implementation, but certain parts need 100 percent, like extinction.
And so you gotta monitor that periodically.
This is just a little bit of a illustration.
In this instance, this maybe 10 or 12-year old female, autism, self-contained classroom had pretty intense aggression, primarily escape maintained, but a secondary function of attention.
And in baseline, what you're looking at here is the number of days between seclusion timeout in a padded room, so the last thing we ever want happening with a student, particularly a student who has escape maintaining problem behavior, and so zero days between seclusion, but really the student was going in multiple times per day and spending a vast majority of every day in seclusion timeout, and that was off the table.
No matter what I came up with as a behavior intervention plan, they were gonna use that seclusion timeout room if aggression occurred, and that was a non negotiable.
Okay, it's good (audio cuts out) selection model, and I knew what was on the table.
So, this is in the fall.
In the spring we implemented a plan, we did as many motivation, operation targeting interventions as we could, but we did not have extinction in effect, okay.
We, I think had functional communication training in there as well, just multi component intervention, and we got to the point where she was on average going to to three days between, two to three days between timeout, so not an hour or two at best, but a few days, and the team was really happy about this, and what I said was yeah, but go get some implementation data, and what we found was that was only at 60 percent accuracy on our other interventions, and what I said was long term, if that drops any, escape's still available for problem behavior, we will be gone and they'll be back in the same boat.
Go get implementation up to 80 percent.
So this is the next fall where we push hard to get implementation up to 80 percent, and now we're looking at basically a week and a half of school days between any trips to seclusion time out and a very brief duration of that time out.
And so, you know, don't assume that your best intervention plan just isn't working until you know that it's not being implemented.
So you do have this activity too which is a follow up to activity one, and it walks you through planning those steps of implementation.
What kind of schedule planning? Do you need a multiple schedule? What do you have to get ready to go for day one? How far do you have to get it before you kind of turn it over to the classroom staff if you're gonna do the early steps, and it walks you through the logistics planning for all of that.
Okay, Shelly, I'm gonna turn it back to you one more time before any questions that people might have, and we can wrap it up.
- [Shelly] Okay, there is one more question.
So, many of the examples that you spoke about sound like they involve learners who have a higher level of language or the cognitive ability to understand breaks and what that means.
Do you have any tips for individuals who are working with much younger children or with learners who have more significant impairments? - Yeah, so, certainly this young lady with the seclusion timeout.
She had no functional language at all, just some guttural sounds.
And so, this is absolutely, these interventions are going to work.
Now, usually the communication response is a very simple one.
Maybe touch the green card when it's out, and that's the only one that's out at any given time, or a hand raise, a simple motoric response as the functional communication response.
So FCT is something that's typically done often with all the literature with profoundly cognitively impaired and non vocal verbal individuals, so this literature, and the use of those multiple schedules are really well suited to that.
Noncontingent escape, let's say that example I talked through, that child was young, but very profoundly impaired.
You know, demand fading, I've done with teenagers who were completely non verbal, and those kinds of things.
So, I think you have to worry, the older they are, the more you have to worry about safety issues, and the more careful you have to be with that curricular and instructional revision, because I often find people are just pouring time into things that I could say their life's not gonna be any worse if they never learn to do this, let's not put anymore time into it, let's go do 10 other things that are gonna be more useful, or you have to be just really on it with your using that errorless procedure instead of trial and error or least to most prompting, so I think you have to be more precise, but I would not necessarily say that any of these interventions are off the table if someone is more profoundly impaired.
What I've often done if someone has, you know, no language, severe aggression, severe SIB, start of with noncontingent and escape.
Start off with something that's gonna work, everybody's safe, you don't have to teach them a new response, which functional communication training means have to teach you a communication response, and you kinda go through a little bit at a time and you get them to the point where they're tolerating less frequent breaks, or that that motivation might be starting to rise for I wish I had break before the scheduled one comes, and then boom you use that as that natural opportunity to add in functional communication training, but, you know, activity choice might not be quite as appropriate for someone who is quite impaired.
If they at least have, kind of, can respond to pictures as meaningfully substitutable for activities, you can still do an activity schedule with them and offer some choices.
That's probably the only one where it's going to be much better suited for someone with really more kind of cognitive and language abilities.
All the others can certainly be implemented well, regardless.
- [Shelly] Perfect, thank you so much.
And I'm just aware of the time.
I know there are a few more questions, but would it be okay if I gathered those and sent them to you in a e-mail and then distributed responses to the group? Would that be okay with you? - Absolutely, no problem at all, and I do see a question here.
Would having the students go to the office for severe behavior for the rest of the day be considered removing demands.
That's like that uber escape.
(laughs) It really is.
It's just that it's contingent escape, and it sure is.
So, any other questions, e-mail them to me, and I will respond and you can distribute that to folks.
Good luck with these activities, and I hope the information was useful, and probably run your question through Shelly and she'll them to be in a batch.
That way if people have the same question, she can get that question out to everyone, but thanks for having me and thanks for all the good work, and good effort that you are doing with your students in the Atlantic provinces.