Taking Charge of Your Health

[MUSIC PLAYING] DR. WAYNE FUQUA: Brian. Would you be
so kind as to introduce yourself to our viewers? DR. BRIAN IWATA: Sure. My name is Brian Iwata, and I’m on the faculty at the University of Florida. WAYNE: Great. Thank you, Brian. Brian is an internationally recognized expert on the assessment and treatment of self-injury. And that is the topic we have for this training video. So Brian, let me start off by asking you
to describe what we mean by self-injury and how it pertains to intellectual disabilities,
developmental disabilities, and autism in particular. BRIAN: Self-injurious behavior is sort of
an interesting disorder, because it really doesn’t consist of any particular behavior. It’s simply defined by its outcome– any behavior
that produces injury to the individual’s own body. Now, I guess estimates of prevalence
range somewhere between 10% to about 17% among individuals diagnosed with
severe or profound intellectual disability or autism. So maybe 15% might be
a good, round prevalence number. And SIB can take a number of different forms. In fact, I’ve got several slides here
that I can show you that will illustrate some of them. The most common form of self-injurious
behavior is forceful contact, usually against the head. So in this first slide, we see a young boy
banging his head against his walker. And the typical injuries that you see are
bruises, abrasions, and lacerations about the site of the injury. Now, here are some other examples–
an example of biting behavior. You can see the lesions
up and down this young woman’s arm. This is an example of eye-poking behavior. An example of severe and chronic scratching behavior. This is an unusual form of self-injurious behavior. It’s technical name is aerophagia, more commonly known as air swallowing. And so this young woman essentially has taken air into her stomach instead of into her lungs. And to give you an idea of the extent to which she’s done that– this is an adult female who weighs 65 pounds. And so what you’re seeing there is basically an inflated stomach. Now, here are some internal views of self-injurious behavior. This is a photograph of a tongue that has been partially bitten off. Here’s a view through a CT scan, and the yellow arrow points to a whitened area. That’s a calcified retina, and it was produced not through direct trauma into the eye but rather indirectly as a result of head-banging, again, one of the most common forms of SIB. Here’s an x-ray of a hand. It belonged to a young man who hit himself forcefully in the face and on the head, and as you can see, has fractured in two places. So if you can imagine hitting yourself hard enough to break your hand in two places, you might imagine the damage that’s done inside the brain. And finally here is an example of pica. This individual has somehow ingested a nail. So as you can see, self-injurious behavior involves a variety of different responses. The common feature is injury produced to the body. Now, some of the behavior is fairly
dangerous and may actually be life-threatening, but we don’t have good
prevalence figures about mortality rate. Because individuals who present
with significant risk are usually either restrained or sedated–
that is medicated. And so we don’t really have a good
figure on how many people actually kill themselves. WAYNE: It sounds like a serious problem
that must present all kinds of health issues and personal freedom
restrictions and stuff of that nature. How do we go about assessing self-injury? And how does that then lead into
some of the treatment protocols that we think might be effective, Brian? BRIAN: Well, for many years the major
focus on the treatment of self-injurious behavior was to simply stop it quickly. It’s a dramatic disorder.
Individuals who engage in it may cause severe harm. And so typically the field has focused
on interventions that eliminate the behavior very rapidly. Occasionally that has been effectively
treated with reinforcement-based interventions. But early in the development of our
treatment procedures for this behavior, a large amount of it was basically
treated by way of restrictive interventions, restraints, punishment, and so forth. Now, that’s simply an attempt to stop it. More recent research has
attempted to understand why it occurs. And one of the predominating theories
is that self-injurious behavior is simply behavior that is learned. It’s acquired as other behavior is. And so that begins to focus our attention
on contingencies of reinforcement. Now, some problem behavior tends to be
maintained by specific types of contingencies. So let me go through a couple of other
responses before getting to self-injury. Let’s take aggressive behavior–
also a significant problem. Well, a great deal of research
on aggression has shown that it’s primarily a social response. It is maintained by either
social positive reinforcement, people deliver things to those who are aggressive, or social negative reinforcement,
people who are aggressive effectively escape ongoing work demands. It’s basically never maintained
by automatic reinforcement– that is the felt aspects of aggression. By contrast we have very stereotypic behaviors–
these are non-injurious repetitive motor behaviors, like the twirling, and the flipping, and so forth. And these behaviors have been shown
primarily to be maintained by automatic reinforcement. Now, one might wonder why that’s the case,
and I’m not really sure, but one can speculate that aggression is a sufficiently dangerous
disruptive behavior that will immediately produce a reaction by people in the environment. They stop things that are going on.
They start things that aren’t happening. Whereas stereotypic behavior, although
it’s disruptive, it’s not immediately disruptive. And so evidence indicates that that
behavior doesn’t really enter into social contingencies. Now, self-injurious behavior is interesting,
because it’s a disruptive behavior. So it basically requires some
sort of reaction from the environment. But it’s also a behavior that
produces a number of felt aspects. And so self-injurious behavior is
one behavior that’s likely to be maintained by any sort of reinforcement contingency. And as this chart sort of illustrates,
it basically lays out the kinds of contingencies that could maintain self-injury. So for instance, if attention or access
to various items are rewards, and if they are not available, then an individual with poorly developed
language skills may learn that a good way to get access to attention and various tangible items
is to engage in self-injurious behavior. Similarly in the presence of work requirements
that are demanding and perhaps somewhat aversive, an individual who doesn’t have any socially
acceptable way to indicate that he or she would like a break may learn that bleeding is an appropriate
way to get people to stop what’s happening. And finally, self-injurious behavior
could be a self-stimulatory behavior. Not so much that people are producing
injuries because others will react, but simply because it produces a great deal of sensory
stimulation when none is available. WAYNE: In spite of the fact that most
people would identify that as painful stimulation. BRIAN: Yes, that’s true.
WAYNE: That’s the possibility of that maintaining, right? BRIAN: As an example, you could take,
for instance, let’s say a long-distance runner. If you were to stand at the finish line
of a marathon, you would not see too many people running across, simply as if they’re
having fun, smiling, and waving, and so forth. Most of them look like they’re
ready to fall over and die. Now, what maintains that behavior? Well, possibly it’s shaped up by
a way of some sorts of social contingencies– praise and various things like that
for learning to run well, but eventually most long-distance runners maintain that behavior
because of basically the felt aspects of running– a sense of accomplishment,
being in good condition, and things like that. So although it looks as though someone
engaging in vigorous exercise feels bad or may be subjected to aversive stimulation,
in fact they’re just engaging in the behavior because it produces some
sort of sensory stimulation. WAYNE: Interesting.
Now, you mentioned sensory stimulation. Does that imply there’s a biological
and/or genetic cause for self-injury? BRIAN: Well, that’s a very appealing possibility. Because biology seems so tied up in
that response, it sort of cries out for a biological explanation. And so a number of theories have been proposed. For instance, one is that individuals who
engage in self-injurious behavior may be experiencing some sort of chemical imbalance, or they
may have been exposed to some sort of environmental toxin. Now, some basic laboratory studies have
shown that if you expose laboratory animals to very high doses of certain toxins,
you occasionally can induce self-injurious behavior. Well, the question becomes then,
does that generalize to actual humans? And in fact, it’s been shown that it’s
very unlikely that most people have ever been exposed to these toxins in the doses
required to produce self-injurious behavior in animals. And so although there’s a preparation
that may lead to a biological explanation, that doesn’t seem to account
for self-injurious behavior in humans. Another possibility is that individuals who
engage in self-injurious behavior simply don’t feel pain. And if they don’t feel pain, then, let’s say,
banging one’s head– when ordinarily it would look like a response that’s directly deterred because of pain,
might be no different than, let’s say, raising one’s arm. And in fact, it might be
a more effective way to get attention. And it doesn’t really matter that I raise my arm
or bang my head, because I don’t feel the pain away. And of course, that requires all sorts
of pain sensitivity tests given to individuals who engage in self-injurious behavior versus not. And not a whole lot of evidence has shown
that individuals who engage in self-injurious behavior are less tolerant to painful stimulation. Now, the third biological theory
is kind of a very interesting one. And it suggests that individuals who engage
in self-injurious behavior may be activating their endogenous opiate system. So for instance, the body produces its
own opiate-like substances under conditions of high stress. And as the theory goes, people who
engage in a chronic self-injurious behavior are simply activating their own opioids, and in fact,
perhaps producing something akin to a biological high. And that’s probably the most interesting
biological theory, because it has led directly to the development of a drug regimen
that might selectively reduce self-injury, and that is the opiate blockades,
like Naltrexone and Naloxone. Now, unfortunately if you look at those
studies what you find is that the evidence is mixed. Some studies show partial effects,
some show no effect, others show relatively reliable effects that wash away rather quickly. And of course, it may be the case that
there’s something to this particular biological theory, in which case you’d have to
explain all the conflicting findings. Well, one possibility is that those types
of drugs would be effective for individuals whose behavior is maintained by that particular source
of reinforcement– the biological reinforcer. But of course, what they haven’t done is to
determine why particular individuals are engaging in self-injurious behavior
before they administer those drugs. And so that’s one possible promising avenue
that needs to be developed a little further. WAYNE: Brian, the endogenous opiate
theory sounds like it’s interesting but is not yet proven to be an efficacious treatment based on that. Are there any other biological, physiological
issues that are pertinent to our understanding of SIB. BRIAN: Well, the way that the medical
and the psychiatric community has typically responded to SIB is to prescribe drug regimens. And these could either be chronic
psychotropic drugs or relatively short acting sedating drugs. And often individuals who are diagnosed
with SIB are given another diagnosis for which these chronic drug regiments might be appropriate. Generally, however, it’s been found that
these drugs have no selective suppressive effect on SIB. Rather if they slow SIB done, it’s probably
a function of simply sedating the individual. WAYNE: And that, of course,
has side effects well beyond the SIB in terms of their
interaction with the environment. So let’s assume that
you have a case referred to you of SIB. Can you kind of walk us through the protocol
of how you do assessment, what you look for, and then how that begins to link
into some sort of treatment planning? BRIAN: Sure. Actually, over the years we’ve
developed sort of a general protocol or assessment sequence for dealing with self-injurious behavior. Now, the first thing has nothing to do with treatment,
and that is the identification of risk and protection against harm, because these individuals
are usually causing tissue damage. And what we want to do first is to determine
how severe that damage is likely to be by reviewing a history of the kinds of injuries that have been produced. We would have a physician check that
individual to make sure that the behaviors that are occurring are not likely to be life-threatening. For example, one of the most dangerous
forms of SIB is pica– ingestion of inedible substances. And pica per se is not necessarily that dangerous
except for the fact that it doesn’t produce any visible injuries. And so an individual who has a history
of pica, in theory, could swallow a broken piece of glass and cause major damage,
if not death, rather quickly. And so first we have the person evaluated
by a physician, we conduct a medical record review, and then we try to, I guess, prescribe
a short-term intervention that would reduce risk– such as the provision of protective devices. Now, these may be regarded by many
people as restrictive, but the alternative is sedating drugs, because there’s no way to stop SIB right now. And once we can protect the individual, then we
have more time to conduct other kinds of assessments. Now, in addition to having a medical
professional conduct assessments, there are several different kinds of assessments that
the typical therapist, even perhaps teacher, might be able to conduct on a regularly routine
basis to document the extent to which risk is occurring as a result of SIB, and we’ve
actually developed several of them. So for instance, here is an example of a rating scale that we published a number of years ago. It’s called the Self-Injury Trauma Scale. And it’s a way for non-professionals to simply observe an individual, to conduct a fully body check, and to actually document the kinds of injuries that are visible on the body, the severity of those injuries, to differentially weight them depending upon whether any injuries are in the location of the head, because those are more severe. And you can come up with a sort of general index of risk based on the number, the location, the type of injury, and so forth. And this might be done quickly on an intake. It might be done prior to treatment. It might be done after treatment has been in effect for a while to sort of document the changes and observed aspects of SIB. Because we as behavior analysts, typically,
when we treat a problem like SIB, we look for changes in terms of reductions and rates. But that’s not necessarily an indicator
of whether or not risk is being reduced. So we like to use these corollary measures. WAYNE: So it sounds like a supplementary
measure of a response product essentially– what damage are you doing to yourself? BRIAN: Yes. And it might be particularly good
for occasional out-patient use when you don’t have access to regular data on the frequency of problem behavior. Now, there’s another way you can actually
get a more fine-grained analysis of severity of injuries. And we’ve sort of worked on this one, too. Many years ago, we tried actually to
figure out how to judge how bad a wound is. And of course, the problem back then was
that you photograph a wound, and the distance from the wound determined size,
and there were a lot of problems. But now we are in the computer age. And so we’ve located some software that
you can download freely from the NIH website. And using that software and a digital
camera, you can get a precise measure on a wound. Now, that might seem sort of esoteric,
but then again if you’re working in an out-patient clinic, you don’t see the individual regularly,
you can take photographs of the wounds and actually document changes over time. For instance, here is a screenshot
of a lesion that we photographed, and we’ve uploaded it to
our software, which is called ImageJ. And all you need to do is to have a ruler
of known length in the photograph, so that you can adjust the resolution based on the ruler. And so we’ve got the wound in the picture,
we’ve got a ruler in the picture, and I’ve basically drawn a circle around a five-centimeter section of the ruler. And we could enter that into the software,
and that automatically adjusts for distance to the wound. You then, as illustrated in this slide,
can simply draw a border around the size of that wound, and that basically is automatically
calibrated by the software, which then yields a measure. So it gives you basically a wound surface
area measure of a wound simply based on a photograph. And what you can do is basically have
parents or teachers who are responsible for implementing treatment programs
send you these data, again, as corollary measures to determine whether or not reductions
that you see in the frequency of behavior are correlated with reductions in
wound size, or risk, or what have you. WAYNE: Do you sometimes
find them to be uncorrelated, Brian? Changes in wound size appears
to be unrelated to changes in behavior, SIB behavior? BRIAN: Well, one that we found,
interestingly, is that when you attempt to correlate wound size with response rate,
that response rate increase immediately translates into noticeable observed differences in wound size. But if SIB were to stop,
then wound reduction is usually delayed, so that the correlation is very good
one way, it’s delayed in the other way. So at any rate, the first thing we do
is to identify risk to see if we can figure out how to document it and then to prescribe
some interim strategy to reduce risk. We then move on to
looking at the function of behavior. As I indicated before, there is a good
bit of evidence to indicate that self-injurious behavior might be maintained by a variety
of different sources of reinforcement. And our field has developed
a number of ways for conducting what’s known as a functional behavioral assessment. A functional behavioral assessment
simply means any formal method for identifying environmental determinants of a problem
behavior, or alternatively stated, sources of reinforcement. And there are three general
methods that have been developed. The first is known as the indirect,
anecdotal, or verbal report measure. It basically consists of checklists
and questionnaires that we use to solicit information from caregivers
about circumstances under which behavior occurs. Now, the strength of these methods
is that they are efficient and simple to use, but of course, the data are highly unreliable. We then have what are known as
descriptive analyses– these involve objective observation of the circumstances under which
behavior occurs out there in the natural environment. And although these are highly objective,
they tend to only inform us what’s happening, but not necessarily what’s causing behavior. And therefore we have the third approach,
known as the functional or the experimental analysis. And it involves exposing individuals
to that general environment but only portions at a time. So we can basically tease out that part of
the environment that’s responsible for behavioral maintenance. So then we have the function of
problem behavior identified, but there are a couple of more things we need
to do before we move on to treatment. The next one is to evaluate
the individual’s adaptive repertoire, because eventually we will want to replace
self-injurious behavior with some alternative behavior. And one thing we’ve learned over
the years is that there are three particular aspects of adaptive behavior
that we need to consider. Now, of course, we could consider
all kinds of adaptive responding, and in general, you would want to do
that for educational purposes. But for the treatment of self-injurious
behavior we found that there are three key types of responses that
are very important to have. The first one is a method to
solicit positive reinforcement from the environment. And if the individual does not have
communication that produces that function, so to speak, we need to establish that. It doesn’t really matter what
the function of the behavior is. If that part of the repertoire is missing,
then self-injury, essentially, could evolve to produce that kind of consequence. The other one is basically
compliance with instructions. And if an individual has very low
levels of compliance, it’s likely that they will have difficulty performing tasks. They will therefore find those tasks
aversive and may engage in problem behavior as a means to escape or avoid. So we look at the individual’s ability
to engage in imitative behavior to follow instructions. And if there are any deficits there,
we know we need to do some strengthening. And the third one is, I guess for
the lack of a better term, a play repertoire. Does the individual have any
responses that will put the person in contact with reinforcing aspects of the environment? And I’m not talking here about social
reinforcement, but simply engaging in leisure behavior. Many clients do not have that type of repertoire. And if they don’t, then chances are they may acquire
an inappropriate or perhaps dangerous self-stimulatory response. And so we look at those kinds of
responses to figure out where deficits are, and we will need to establish them if they’re missing. Then the last part of
the assessment is, of course, motivation. Why would an individual
be motivated to acquire these new responses? And so we use a variety of methods,
which probably are going to be reviewed in another segment, for conducting what’s known
as a systematic preference assessment. We try to identify a series of items,
these will be various types of social interaction, access to edibles or leisure materials
that are highly preferred, and we can use these to establish those alternative responses
that eventually will replace self-injurious behavior. WAYNE: It sounds like
a comprehensive assessment, Brian. What do you do with this information? How do you then turn this into
a comprehensive treatment strategy for SIB? BRIAN: Well, everything is somewhat
keyed off with a functional analysis. We conduct the functional analysis,
which supposedly identifies the source of reinforcement for problem behavior. And then based on that outcome,
we can put together an intervention plan that neutralizes the source of reinforcement
that’s responsible for behavioral maintenance. So for example, I’ve got several slides here
that just lay out that strategy in a very general type of way. Let’s say that we conduct a functional analysis
and find that problem behavior, or SIB in this case, is maintained by social positive reinforcement–
access to attention or tangible items. Well, we know that there are three general
ways that you can reduce behavior with reinforcement. One of them involves eliminating
the antecedent event that serves as what we call an establishing operation
that makes a reinforcer valuable. If the reinforcer is less valuable,
the individual is less motivated to engage in the behavior. The second one is to eliminate the contingency
that maintains behavior, that is extinction. And the third one is
to strengthen an alternative response. So if we apply those three general strategies
to behavior maintained by social positive reinforcement, what would that look like? Well, the establishing operation is usually
going to be depravation from the positive reinforcer. Right? Because the person is
engaging in the behavior to get it. So once we identify the function,
we can identify the establishing operation. We basically deliver reinforcement for free. So we use a procedure that is often
called non-contingent positive reinforcement– increase free access to positive reinforcement,
decrease deprivation, decrease motivation to engage in self-injury. Now, the second strategy– eliminating
the contingency of reinforcement– we’ve also identified by way of the functional analysis. We find that self-injury is
maintained by access to attention. Well, we know the way to extinguish
that behavior is to not attend to occurrences of SIB or approximations thereto. Now, the third general strategy, establishing a replacement response, we know that positive reinforcement is valuable, we know that this behavior occurs because it produces positive reinforcement. And so what we therefore need to do is to establish another response that’s socially acceptable that will produce positive reinforcement. And we already have the reinforcer that we can use to establish that response. It’s basically the attention that’s used to maintain the problem behavior. And so that becomes the focus of our differential reinforcement procedure. And many people call that functional communication training. In this case, the function is positive reinforcement. Now, if behavior is maintained by
social negative reinforcement, interestingly, these same three general strategies
apply, but procedurally they are the opposite. So if problem behavior is maintained
by negative reinforcement, it is not the case that deprivation is the establishing operation,
but rather it is usually the presence of aversive stimulation. Having identified the contingency that
maintains behavior, we can then identify the aversive event that serves as the establishing operation,
and we can modify it in a number of different ways. We might simply give the person
free breaks from work, at least initially, because that’s a very quick
way to reduce problem behavior. That will then buy us some time
such that we can conduct further assessments. Those assessments would be aimed at
identifying what feature of work makes it aversive. So for some individuals it might be
the nature of the task, that is novel tasks are more aversive than maintenance or familiar tasks. So we can continue to present some tasks to
that individual, but they just won’t be as difficult as they used to be. Or it may be the duration of the work session. So at any rate, we could, through
some systematic assessments of curriculum, identify those features of the task that
make it aversive, modify those features, and problem behavior should decrease. We could then combine that with
a strategy known as stimulus fading. That is gradually reintroducing that
part of the environment that occasioned problem behavior very slowly, and the person
may be able to tolerate gradual return to the original situation that produced problem behavior. Now, another procedure has gone by
a particular name, and that is the high probability instructional sequence,
or as we call it, the Hi-p sequence. And this particular procedure is
usually implemented in a characteristic way. It involves first identifying a series
of instructions for which there is a high probability of compliance, and of course,
a corresponding low probability of problem behavior, which is why it’s called a Hi-p sequence. And so we sequence instructional
situations as follows– we start every session with three Hi-p instructions, assuming that
that does not produce problem behavior, we then kind of sneak in a Lo-p instruction. Now, of course, if we got compliance
with the Lo-p instruction, we don’t want to persist. Because that may then occasion
problem behavior, so we go right back to the Hi-p instructions. And so a Hi-p instructional sequence is
implemented by alternating, typically, three Hi-p’s to every one Lo-p throughout a session. There’s one final possibility, and this is sort of
an indirect approach to getting at the establishing operation. The other methods that I’ve described
sort of directly reduce aversiveness of a situation. Occasionally we’ll find that
you can’t change the nature of the work. I mean, it is what it is. You have to take the shower,
you have to perform the math worksheet, whatever. And so there’s no way to directly
reduce the work requirement, but it’s possible that we might make the general
situation more tolerable simply by throwing in free positive reinforcers. And so, yes, you have to work,
but while you’re working you get free access to better positive reinforcers. And that may make the situation
tolerable enough so the person is less likely to engage in escape behavior. And as long as they’re remaining
in that situation, they may actually work, and they’re compliance may increase. So there is a series of things one
might do that all fall under the strategy of eliminating the establishing operation. Now, we then get to the actual
contingency that maintains behavior, in this case it’s escape. Now, extinction remains the operative
method for reducing behavior, but in this case, extinction of escape-maintained
behavior is procedurally the opposite of extinction of behavior maintained by attention. So for instance, we’ve all heard of
procedures known as planned ignoring– time out. Those are all procedures that involve
terminating ongoing events, which might be appropriate as extinction for behavior
maintained by positive reinforcement. Well, if we use those same procedures
for behavior maintained by negative reinforcement, we’re not using extinction,
we’re delivering negative reinforcement. And so the way you extinguish that
particular function of behavior is to not terminate what’s going on– that is you continue
the situation and not allow problem behavior to produce escape. Now, that particular version of
extinction can become very effortful, because you’re attempting to maintain
the ongoing work task, the individual is trying to escape, and their behavior may escalate, and physically
it may become difficult to manage that situation. Now, the third approach is the replacement
behavior, and there are several options there. Occasionally people who engage
in severe problem behavior engage in other minor sorts of behaviors before actually
engaging in the severe problem behavior. For instance, let’s say you are
an instructor trying to get me to work. Well, eventually if it gets bad enough,
I will engage in aggression, or I’ll throw my chair, or something like that, but initially
I won’t do that. I might start basically by cursing, or by shuffling my papers, or by losing
eye contact, or groaning, or things like that. And these responses
we would call precursor responses. And so if an individual engages
in reliable precursor behavior, and you’re observant enough to see that,
then what you might simply do is to give an individual a break from work contingent
upon the occurrence of these precursor behaviors. Now, the reason I list that first
is that doesn’t involve any new teaching. The person already had that response
in their repertoire, you just are selecting it out, and you get to forgo the aggression
or the significant property destruction. Now, the bad aspect of doing that
is that sometimes these precursor behaviors aren’t a whole lot better than
the problem behaviors– the cursing or the throwing things. And there’s another subtle issue–
and that is that these precursor and target behaviors are often tied together because
they have a common history of reinforcement. And if you select that one that is,
let’s say, the minor behavior, then that’s the one you’ll get, but the major one
stays functional in their repertoire. And let’s say you are willing to tolerate
a slightly non-compliant student or one who curses, but that student goes to the next class,
that teacher doesn’t tolerate that. Well, that teacher is going to get beat up. And so a better approach, even though
it takes a little bit longer, is to teach the individual a more socially acceptable escape response. Now, that’s not difficult do, but you would
do it in a way that’s very different than teaching a new attention getting response. So we can pick the same response,
that is, let’s say, raise your arm, and if we deliver lots of attention or edibles for doing that,
we’ve taught the person how to get access to attention or edibles. You can’t do that with an escape response.
The reinforcer is terminating ongoing events. So the way you initially develop
that behavior is you start a work situation. Before problem behavior occurs,
you prompt what you would like to establish as the new escape response,
and you simply walk away. Come back in a few minutes,
restart the work session. Before problem behavior occurs,
you prompt the alternative response and walk away. And it may take several hundred trials,
but eventually, I hope you’ll notice, is that, let’s say, I, the teacher, approach you, and now
you’re engaging in that behavior, I give you a break. Now, that type of behavior is
valuable for everyone to have, because we don’t know when people will encounter aversive stimulation. So although to some it may seem like
we’re simply teaching people how to get out of work, that’s valuable because the next time
they are faced with an aversive event, it would be nice if they had
an appropriate rather than inappropriate way to escape. Now, of course, the final strategy for replacement behavior is to teach some sort of compliance with the task at hand, because that’s the behavior that should be occurring. And the thing about this function of problem behavior that is escape is that although we know that vastly improving positive reinforcement for compliance may increase it, we also know that escape is a valuable reinforcer. So we can combine negative reinforcement and positive reinforcement and make compliance more valuable than problem behavior. So for instance, you could hit your head and you don’t have to work, or you can engage in compliant behavior, that is perform
the task, and you get a break from work. And while you get a break from work,
you get access to these neat positive reinforcers. And so basically that’s the general way
you would formulate strategies for behavior maintained by social negative reinforcement. Now, there’s one more function.
WAYNE: Sure, go for it. BRIAN: And that is behavior
maintained by automatic reinforcement. Now, in this case, the ongoing
social environment is not particularly relevant to the maintenance of behavior,
because it’s maintained because it produces its own reinforcers. That being the case, we would
first ask– what’s the establishing operation? Why is self-injurious behavior valuable? And the answer is probably that
the person is not getting a lot of physiological stimulation. And so the first strategy of eliminating
the establishing operation would involve eliminating general deprivation to sensory stimulation. We would try to make access to sensory
stimulation available as often as possible. And to the extent that that sensory stimulation
is at least as good as, if not better than the stimulation produced by self-injury,
we would see some temporary reductions. Now, that’s not the solution,
but at least that’s a start. Now, then we have reinforcement
and how to limit reinforcement. And the problem here is that the person
is actually producing his or her own reinforcers. And so the types of extinctions that
one would use for behavior maintained by either social positive or social negative
reinforcement are totally irrelevant. They just won’t work. A general strategy known as
sensory extinction has emerged in the literature. It’s a little bit of a trial and error procedure,
but it involves attempting to interrupt the final part of the behavior or rather to attenuate
the stimulation produced by the behavior. So the behavior could occur,
but not its endpoint, or the behavior could occur to its endpoint, but the stimulation is reduced. As in, for instance, putting thickly padded
protective arm devices on the arms of someone who scratches. They can still scratch, but it feels
a little bit different, and to the extent that it’s less reinforcing, again,
behavior may temporarily increase. Now, there’s another group of
interventions that are not technically extinction, but they have been shown
to produce extinction-like effects. And they go under the category
of response effort interventions. So if we can make the behavior
more effortful to exhibit, although if they would exhibit the behavior,
they would still produce the consequence, it may be sufficiently effortful
that eventually they just stop. And a perfect case example that was published
many years ago in a case study by Ron Van Houten, who was working with a woman
who punched herself in the face, and what he simply did was to put
1.5-pound wrist weights on her arms. Now, it wasn’t extinction,
because she could hit herself, still feel the effects, but she would do that a couple of times
and then simply stop because it was too effortful. And the interesting thing about it was
that these arm weights then acquire discriminative properties. So while you’re wearing them you
know that it’s more effortful, and so you stop engaging in the behavior over time. Other examples of response effort interventions
have included things like these arm sleeves that you can fit onto individuals who engage in,
let’s say, forceful punching behavior. And they actually have channels in them
so you can place flexible Plexiglas rods in these channels, and you can actually dial in the amount
of effort required to reach one’s face. So for instance, if the problem behavior
is punching severely, I put several rods in these arm sleeves, and now I can reach my arm, and it’s
a little bit– my face, rather– it’s a little bit effortful to do that. So I can feed and scratch and things
like that, but it’s hard to punch very effortfully. So that would be another example
of response effort interventions. The final solution to this function of
problem behavior is to teach, again, a replacement behavior. But in this case, social consequences
are not particularly relevant. The important thing would be to teach
a response that will put the individual in contact with a variety of sources of sensory stimulation. So in essence you teach the person
new, socially appropriate, self-stimulatory behavior. Now, you may initially need to prompt
and reinforce arbitrarily with, let’s say, edible items– that response just to get it going. You may also need to block
self-injurious behavior for a good bit of time. But the hope is that eventually once
the individual starts engaging in that response a number of times, they will then come
into contact with alternative sources of reinforcement, and now you’ve built
another self-stimulatory response. Well, now that they’re interacting with
the physical environment, it should be easier to build another self-stimulatory response. And then before long, the individual has
what we would consider to be a socially appropriate toy play repertoire,
which is what most people have. WAYNE: That sounds fascinating, Brian.
There’s a whole range of effective behavioral interventions that need to be matched up based on,
it sounded like, the function of the self-injury. BRIAN: Right. WAYNE: Now, one that was notable
in its absence, I didn’t hear you mention anything about the use of punishment.
What is your perspective on that? Is that ever appropriate in this
armamentarium of interventions for self-injury? BRIAN: Well, possibly. Earlier I mentioned
that way back in the beginning when we began dealing with this problem,
the goal was to eliminate it quickly. And the procedure that has the fastest
influence on behavior is punishment. Although some people might argue
that that’s not necessarily the case, if you go look at all the published data,
the fastest way to reduce behavior is by way of punishment. Now, of course, punishment is
an intrusive intervention– many people object to its use. And so we’ve developed this sort of
technology of behavior reduction with reinforcement tied to a functional analysis of behavior. Now, that is not a simple technology. As you’ve just seen, it requires knowing
the function of behavior, having these different strategies, turning these strategies into procedures. But if we do that, then by and large we can
decrease a large proportion of self-injurious behavior through their use. Now, that doesn’t mean
that they are always effective. Occasionally you can’t use extinction
with severe problem behavior, such as you can’t ignore someone who’s bleeding,
or you can’t continue to require someone to work if they’re very large,
or you can’t interfere with the sensory stimulation produced by the response. And so occasionally extinction
is simply impossible to use. Well, then we have difficulties,
because the behaviors will always be reinforced. And under those conditions,
when all of the positive reinforcement-based procedures we have available aren’t effective,
we might consider the use of punishment, again, because it produces relatively rapid effects. And the idea is that we might
use punishment for a short period of time simply to suppress the behavior
long enough for the other components of intervention, namely the reinforcement procedures, to take hold. WAYNE: And I would imagine that
if you do move in that direction you have to be concerned with all kinds of
ethical and/or health issues about the use of the punishing stimuli. BRIAN: Sure. Many punishing
procedures have been sort of documented in the literature, and so it would
be a matter of going to the literature, finding them, paying very careful attention to
the limiting conditions of their use– like, what are the limits
of applying this procedure? And then making sure that
whatever review committees or review boards there are that oversee intervention
plans have an opportunity to review those procedures. WAYNE: That sounds very reasonable.
It sounds like a very effective technology. It has a fair number
of decision points in there. I’m wondering if you might be able
to help us understand how it’s actually applied with the help of a case study
or something of that nature, Brian? BRIAN: Sure. In fact, I have
a case study, and it illustrates self-injurious behavior maintained by
social positive reinforcement that is attention. So as you can see in this
introductory slide, this client was Henry. And he was a very young boy,
just three years old, one of the youngest children we ever treated. He didn’t
have any formal means of communicating, but he generally seemed
to be somewhat socially responsive. And his problem behavior
was forceful head-banging. That is he was banging
his head against the floor and against the walls to the point where he was
at significant risk for doing brain damage. Now, we conducted a functional analysis of his SIB, and these are the results. Now, if someone hasn’t seen a functional analysis graph before, let me just steer you through it. Each data point represents a session of a particular assessment condition. And they’re sequenced from left to right going from first to last. And the measure being portrayed, that is the y-axis, reflects mean responses per minute of SIB across the minutes of a session. And back when we treated Henry, our standard session length was 15 minutes. And I’ve color-coded the data so you can see the different assessment conditions. The red data points illustrate self-injurious behavior during the contingent attention condition. This is a condition in which the therapist does not deliver any attention except as a consequence for SIB. So the antecedent is deprivation, the consequent event is delivery of attention. The blue data points represent the demand condition, which is another typical test condition of a functional analysis. The antecedent event here is the presentation of work trials, and the consequence is the termination of those trials. So it’s the test condition for behavior maintained by negative reinforcement. And the black data points represent the alone condition. That’s a condition in which nothing happens. So we’re ruling out the effects of social reinforcement here unto the extent that behavior maintains, it’s probably being maintained by its own consequences. There is one control condition included, and that’s illustrated by the green data points. And in that condition attention is available for free, so there’s no deprivation from attention. There are no task demands, so that establishing operation is also gone, and there are lots of fun things to do– free access to leisure items. So we would suspect that regardless of the function of problem behavior, you should get low rates of problem behavior in that condition. Well, as you can see from the graph, there is clearly a higher rate of self-injurious behavior in this attention condition. So we would say, Henry’s self-injurious behavior is maintained by attention. Now, how do we turn that into treatment strategies? We implemented his intervention in several phases. In the first phase, we combined two of
these strategies for reducing behavior with reinforcement. We attempted to eliminate the establishing
operation, that is deprivation from attention, by simply providing continuous attention. Attention is available for free. We combine that with extinction, and that is
if you engage in head-banging, then attention goes away. And so it’s exactly the opposite of
the assessment condition of a functional analysis. You get lets of free attention.
It temporarily goes away if you engage in SIB. And the reason we combine those two
is because those two in particular typically produce an immediate large reduction in problem behavior. Now, we could add in that third strategy–
that is teaching the replacement– but that would have required Henry to learn
new behavior to get reinforcement. And by combining the “eliminate the establishing operation”
and “eliminate the reinforcement contingency,” Henry didn’t have to learn any
new behavior, he just stops his old behavior. And so once the self-injurious behavior
decreases, we then come in with the second phase. We can reduce the access to
non-contingent reinforcement, we keep the extinction in effect, and we
attempt to establish an alternative behavior. In Henry’s case it was simply
an arm raise– you just raise your arm and that’s the new replacement. And some authors have particular
biases about the type of replacement behavior that should be selected, like a vocal
response, or a PECS response, or something like that. And we don’t have any particular biases.
Simply we’re looking for a behavior that is easily shaped, and if it occurs, it will be
highly discriminative in the environment. Someone is likely to recognize that behavior. And so we frequently
prompt that throughout a session. When that behavior occurs,
we deliver enthusiastic attention, and temporarily we also deliver edibles. Now, edible reinforcement was
unrelated to why Henry engaged in the head-banging, but we were attempting to make
the arm raise more valuable than the head-bang ever was. And so in a relatively short
amount of time, as I’ll show in the data, Henry acquired the new
replacement attention-getting behavior. Now, at that point we have
an effective intervention that’s totally impractical, because people can’t
do this all day long. And so what we eventually do
is to come up with a maintenance strategy which involves gradually
reducing the amount of reinforcement we deliver for that replacement behavior. So let me show you some results. This graph shows two sets of data. The blue data points represent the rate of self-injurious behavior. The red data points represent unprompted– that is independent occurrences of the replacement behavior. That is this sign “the arm raise.” And as you can see during baseline, Henry’s averaging about six head-bangs per minute. This is like one every 10 seconds. This is an extremely high rate of head-banging behavior, and of course, never raises his arm. In the next condition we combine non-contingent reinforcement and extinction, and you can see SIB drops immediately, precipitously, to the point where by the third session SIB is occurring at zero rate. And we run that out for a little bit, just to make sure that we have a nice reduction in self-injurious behavior. In the next condition labeled as DRA, or differential reinforcement of alternative behavior, we are prompting Henry to engage in the replacement response, we are delivering the attention, we are delivering the edibles. Now, although we’re frequently prompting him throughout the session to engage in a behavior, we are only counting it in the data if it’s unprompted. So as you can see in the graph, during the first couple of sessions the behavior really is not occurring. Well, it’s actually occurring but not independently. But after about a dozen sessions, what you notice is that the rate of independent arm raising has reached the same rate that the original SIBs occur at during baseline. And so at that point I said we had a very effective intervention, totally impractical. So finally, in the last condition, what we simply do is to place the replacement for behavior on an intermittent schedule of reinforcement, so that we reinforce it less often. Now, on the slide you’ll notice that we were using interval schedules of reinforcement. And that could lead to a very long discussion about maintenance programs, but let me just give you the sort of CliffsNotes version. In typical behavioral acquisition,
let’s say we’re teaching someone new adaptive behavior, everyone knows the value of
the intensive prompting and the shaping and use of continuous schedules of reinforcement,
and we do that because otherwise people don’t learn. Well, once they have acquired
the behavior, we don’t need to deliver reinforcement all the time,
in fact, it’s very impractical. The most common way to thin out
continuous schedules of reinforcement during maintenance is to increase the ratio requirement. So that originally it would be
every behavior gets a reinforcer, which is basically a fixed ratio one schedule. So we might then go, OK,
now let’s put that on an intermittent schedule. We’ll go to fixed ratio two–
every other behavior gets a reinforcer. Or a fixed ratio of five–
every fifth behavior gets a reinforcer, until we get out to some terminal schedule. Now, why do people
use these ratio schedules? A couple of reasons–
they’re easy to use, and they have a characteristic effect on behavior.
They produce high rates of behavior. And so for relatively few reinforcers,
we’re maintaining higher rates of behavior. But of course, the only way to get
the reinforcer is to engage in that behavior that we’ve taught the individual. Now, we can encounter a problem
occasionally when we stretch that ratio too high. We encounter a situation
that’s called ratio strain, where– sort of using lay terms–
the reinforcer has gotten too expensive. We are requiring too many
responses per a reinforcer, that person stops behaving, we recognize that as a problem,
and we have to readjust that schedule. Now, that’s typical acquisition. When we replace problem behavior
with an alternative behavior, we also have an acquisition situation,
but the difference here is that the person already has a behavior that
is more well established in their repertoire to produce that reinforcer. So you’re attempting to teach them
the new response, but they already have an old response in their repertoire. We shape up the new response.
We put in a ratio schedule of reinforcement. We make the reinforcer
more and more expensive. The rate of the new response
goes up, the person hits ratio strain. I would submit much faster
than under typical acquisition conditions. Why? Because the person can
switch back to an old behavior. And so this seems counter-intuitive,
but I think it’s an important part of maintenance for severe problem behavior,
including self-injury, and that is we don’t want a high rate of the replacement behavior. Because then we will be able to
reinforce a larger proportion of them, and the behavior is less likely
to then be replaced by the problem behavior, which is why we use interval schedules. So that was not quite the CliffsNotes
version, but it was, I think, short enough. And so what we do is use
interval schedules in which the reinforcer is available periodically–
after 30 seconds, after a minute, after however long. But whenever that interval is up,
only one behavior is required, and so the reinforcer never becomes more expensive. And what you have there is
a rate of response that simply starts to decrease. And so by the end of treatment,
what we have is a zero rate of self-injurious behavior– Why? Because it’s being extinguished–
and a low rate of the replacement behavior. And so as long as Henry’s caretakers
can periodically deliver reinforcement for this arm raise and not deliver reinforcement for
the head-bang, he should be OK. WAYNE: That sounds pretty darn effective.
Now, what about the course of treatment, Brian? Is that a pretty standard course
of treatment in terms of how quickly he responds and what you do for maintenance? BRIAN: Well, that’s a good question. Because the typical way that intervention
is implemented is to come up with the plan, and then the plan goes into effect
everywhere, all day, across all teachers and therapists. And there’s a good reason why
that’s a common way to intervene. Because if it’s effective, life changed,
and life is now good because problem behavior goes away. But in a way, that model
of intervention requires success. Because if you are not successful,
there are several problems that you might need to solve. The first one is when we implement
interventions all day long by everybody, necessarily we have to come up
with a practical intervention, which means usually a maintenance schedule
rather than an acquisition schedule. And so if we have a failure,
it may be the case that we didn’t select the wrong procedure,
but we simply implemented it in such a watered down way that it was ineffective. Now, if it’s being implemented all
day long by everybody, you have a higher likelihood that people are going to–
what? Do it incorrectly. And the question becomes–
how effective does an intervention have to be in order to be therapeutic? In other words, what about 90%
of the time it’s being implemented correctly? Well, most of us would say
that would probably be pretty effective. What about 60%?
What about 50%? And if everyone is doing
the thing all day long, then chances are you can’t monitor everyone very closely. And so another possibility is,
again, not that the procedure was incorrect, but it was just not being done correctly. And of course, the third possibility
is you have the wrong procedure. And if you implement interventions
using the typical model, although, it’s seemingly efficient to start with, if you encounter
a problem it’s almost impossible to figure out what the problem was,
because there are at least three different problems. And so our preferred mode
of intervention is as follows– and that is we have figured out
how to temporarily reduce risk, and that’s in effect all day long. And several times throughout
the day we will bring that person into a session and implement the intervention
at maximum intensity, so we know that we couldn’t have done it
any more intensively than it was done. We only have one or two people
doing it, so we could monitor them and ensure they’re doing it at a high accuracy. And so very quickly we can
figure out if it’s that third problem– namely the procedure was ineffective. And if it was that third problem,
we didn’t go to the effort of training everyone to do it, so we could just train that
one therapist to change things, and we can keep on cycling through
interventions very quickly until we get the right one. And then at that point we start
to look for the maintenance and a generalization program. WAYNE: That sounds like it
is a perfectly good strategy. I wonder if implementing
the intervention across lots of people and lots of settings, which I know
you don’t always do, but if that adds to the possibility a satiation
problem with the reinforcer. Does that ever happen at all, Brian? BRIAN: That may, but if it does,
then it might be OK. So for instance, let’s say I engage
in self-injurious behavior to get attention. Now, I get a lot of attention,
and I get a sufficient amount of attention so that now I’m satiated to attention. Well, then you might not be able to
use attention to reinforce any particular behavior, but I’m also sufficiently unmotivated so that
I won’t be engaging in much attention-getting behavior. And usually satiation with reinforcement
is a more significant problem when we are using arbitrary reinforcers to compete with
the reinforcers that are maintaining problem behavior. Let’s use Henry as an example.
He banged his head to get attention. Well, if we didn’t do a functional analysis
and wanted to use reinforcement, we might say, OK, we’ll just deliver edibles
for Henry for not engaging in self-injury. And if the edibles, at least temporarily,
were more valuable than the attention, then he might refrain from
SIB to get the edibles. But as he has consumed
more and more edibles, edibles will become less valuable, and that’s
when attention will become more valuable. And so at some point when
we use these arbitrary reinforcers to compete with the reinforcers that maintain
problem behavior, that’s when satiation can become a problem. Initially the arbitrary reinforcer is
more valuable, but with prolonged use it loses its value. WAYNE: What you’ve talked about for
the most part, Brian, is really focused on treatment, and very effective treatment I should mention. Does any of this have implications
for the prevention of the emergence of self-injury? BRIAN: I think so and not
just for prevention of self-injury but for most problem behaviors. Again, way back in the early
development of our field, if someone is referred with a problem behavior, we’ll fix it. And of course, if you don’t have
a problem behavior, then you’re not of interest to us, because you don’t need our help. And so the point at which
we became involved with an individual was the point at which
the problem behavior became severe enough to do something about it. And so as a result we simply
didn’t focus very much on prevention. And of course, at that time
we didn’t really know much about functions of behavior, so we really weren’t
too clear about the origins of behavior, and we hadn’t really
developed any prevention strategies as a result. Well, now that we’ve identified
environmental circumstances that maintain behavior, instead of waiting until we have
the severe head-banger or the horribly aggressive client referred for treatment,
of course, we have to keep working with them, but instead of generally waiting
for problem behaviors to emerge, we can begin to look at the environments of
young children who are at risk for the development of problem behavior
but who don’t yet show it. And especially in early
intervention programs we can start to look at their environments very carefully. And when we begin to see
certain things happening, like deprivation from attention, failing to produce
a socially acceptable response to get attention, or aversive stimulation–
a student or a child many not tolerate that very well. When we begin to see these
sorts of environmental situations arise, we can act quickly to actually
use those occasions as a warning that we have to teach
these alternative adaptive responses before they’re actually needed. And so when faced with
deprivation, when faced with aversive stimulation, the individual already
has these responses in their repertoire, and they may become less
sensitive to the kinds of environmental arrangements that produce problem behavior. And I think that’s eventually
where our field wants to go. WAYNE: Yeah. The idea of early
intervention– that takes it even further towards the extreme of early, doesn’t it? BRIAN: Yes, it does.
WAYNE: Absolutely. Brian, is there anything else you
would like to offer in terms of advice for practitioners in terms of assessment
and treatment of self-injurious behavior? BRIAN: One of the things that
people often ask me or becomes apparent– when they make referrals, or just,
let’s say, send me information about their clients, is that often the situation has
become so bad, by the time people seek outside advice, that a clear emergency is occurring. And some significant, perhaps permanent,
injury is likely to happen quickly. And so if there’s anything that
I might recommend is that they pay attention to the early emergence
of self-injurious behavior and not simply dismiss it
as something that’s going to go away. Because if anything it’s likely
to intensify, because it becomes dramatic enough such the environment has
to respond to it and will maintain it. And so I would begin very quickly
when I see the emergence of anything that looks like it might be
self-injurious to start to do the assessment sequence and to start to establish
the replacement alternative behaviors. WAYNE: That sounds like
a great strategy actually. And just out of curiosity related
to that, do you see self-stimulatory behavior and self-injury as being
somewhat on the same continuum? Or do you see them as being
dramatically separate, distinguishable categories of behavior, Brian? BRIAN: That’s actually
an interesting question. There was a theory that was
proposed a number of years ago that self-injurious behavior
that’s self-stimulatory in nature always starts as
self-stimulatory behavior. And so there was a progression
of non-injurious, self-stimulatory behavior to eventually injurious. Now, there was very little
evidence to support that theory, although it kind of makes
sense if you think about it. I engage in mild behavior
that produces sensory consequences, and over time I require
more sensory consequences so the behavior may evolve to be self-injurious.
We’re not really sure if that occurs. Because there are a lot of people,
many more people, who engage in self-stimulatory behavior. And they do that for long periods of time,
over years, and they never become self-injurious. A much smaller subset of people
engage in self-injurious behavior. And so if it were the case that
the appearance of self-stimulatory behavior was predictive of a development
of self-injurious behavior, we should see generally about equal prevalences–
and we don’t, they’re much different. WAYNE: OK. Interesting.
Thanks, Brian. Could you give us a real quick
wrap-up and synopsis of your recommendations for practitioners in terms
of the management and assessment of self-injury? BRIAN: Sure. Again, I think the first thing
is assessment of risk and prevention of harm. So that is your first responsibility as a practitioner. And then once you’ve done that,
you have bought time to do a very careful job of assessment– the function
of behavior, adaptive repertoire of reinforcers. And so now you’re ready to intervene. And the functional analysis will,
in a sense, steer you toward certain intervention strategies. WAYNE: Great. Well, fantastic.
Well, Brian, thank you so very much for sharing your research and clinical
and scientific experience with self-injury. With some of these strategies,
assessment and treatment strategies, we can greatly improve the clinical
outcome for people that are afflicted with self-injurious behavior problems.
Thank you. BRIAN: Well, you’re welcome, Wayne.
I’m glad to do that. BRIAN: During the interview,
the case study presented on assessment and treatment of self-injurious behavior was Henry. If you recall, Henry was
the three-year-old boy who engaged in head-banging behavior
that was maintained by attention. And in this simulation, we will
sort of recreate part of Henry’s assessment, that is his functional analysis,
and then his treatment conditions. Now, how do we determine that Henry’s
self-injurious behavior was maintained by attention? We conducted what
is known as a functional analysis, which contains several
test conditions which examine the influence of various contingencies
on problem behavior and also a control condition. Now, the two key conditions
for Henry turned out to be the play condition, in which access to attention
is delivered freely, that is Henry does not need to engage in self-injurious
behavior to get attention, and then the attention condition,
which I will describe next. Now, Travis Jones is playing
the role of Henry. Jennifer Haddock is playing
the role of the therapist. And so let’s see what unfolds
during the playing condition. Henry has free access
to leisure items, free access throughout the session to attention,
no consequences are delivered for the target behavior, in this case, Henry’s
head-hitting behavior. JENNIFER: Henry, I really like
how you’re drawing and playing so nicely. You write your name really well. That’s a pretty cool puzzle we have here, too. [TOY PLAYING MUSIC] BRIAN: Now, Henry engaged
in a problem behavior that is property destruction, but that’s not the focus of the assessment. And so the therapist
simply does nothing as a consequence. JENNIFER: That’s really good
how you wrote your name there, Henry. I also have this puzzle
if you want to play with it. That’s really good. Nice job. BRIAN: As you can see,
the therapist simply delivers attention freely throughout the session. JENNIFER: It’s a really cool day outside, huh?
I bet you’re going to go play outside when you get home. I think this is
a “Dragon Tales” puzzle. I like “Dragon Tales.” BRIAN: When he engages
in self-injurious behavior, the therapist does nothing in terms of a reaction. JENNIFER: That’s a cute dragon. Oh, there’s your puzzle piece. Do you want me to play you a song? [TOY PLAYING MUSIC] Nice. It’s a really pretty day outside.
You could go swimming or skateboarding after this. It’s going to be so much fun. BRIAN: Now, problem behavior
may occur occasionally during the session. But if no consequences are
delivered following its occurrence, over successive sessions
we should expect that it would decrease to a near zero rate. [TOY PLAYING MUSIC] JENNIFER: I like that keyboard.
It’s pretty cool. [TOY PLAYING MUSIC] You’re such a good musician. I really like how you’re drawing there.
You’re writing your name lots of times so neat. Henry you can sit there and play,
and I have to do some work. TRAVIS: Play with me. Play with me, please. BRIAN: Now, Henry has engaged
in a socially appropriate response that is a request to play. But we’re not attempting to
identify the function of that behavior, so the therapist really
does not deliver any consequences for it nor does she deliver any
consequences for any other problem behaviors except for the target.
For example, property destruction also will not produce attention. JENNIFER: Oh, Henry, I don’t like it
when you hit yourself. Henry, I’m trying to read.
Don’t hit yourself, please. Henry, are you OK?
That sounds like it hurt. TRAVIS: Play with me.
Please play with me. JENNIFER: Are you OK?
You’re going to leave a mark. Henry, I really don’t like it
when you hit yourself. TRAVIS: Look at the triangles I made. JENNIFER: Oh, Henry, how’s your ear?
It’s looking a little red there. Don’t hit yourself so hard, buddy.
You’re going to give yourself a headache. TRAVIS: See? Look, a triangle. JENNIFER: Henry, I really don’t
like it when you hurt yourself. No, Henry. Don’t hit yourself anymore. BRIAN: Now, once we had determined
that Henry’s self-injurious behavior was maintained by attention, we developed a series
of interventions aimed at attention-maintained SIB. And the three general components
of treatment were, first, non-contingent reinforcement, that is the non-contingent
availability of attention, extinction for the occurrence of SIB.
If SIB occurred, the therapist briefly turned away. And then eventually differential
reinforcement of alternative behavior, sometimes called functional
communication training, in which the therapist actually attempts to establish
a socially appropriate attention-getting response. So in this first simulation we’ll see
the implementation of non-contingent reinforcement and extinction,
that is attention freely available except when self-injurious behavior occurs. JENNIFER: Are we going to play
a matching game? So you turn them over like that. Yeah.
And then you turn over another one. Good. You didn’t get a match,
so I’m going to try. Oh, no match. Look at that rabbit.
He’s jumping rope. Oh, he’s making s’mores.
Those look delicious. Look, I found a match,
so I’m going to put it over here. And when you find a match–
you have a match. Nice job. BRIAN: As you see, attention is
freely available, but when problem behavior occurs extinction is put into effect. JENNIFER: Did you find any matches?
Let’s see– oh, he’s carrying lettuce. He looks so tired.
Yeah, those rabbits match. Nice job. Look at all those cards you turned over.
You have Tigger and Owl and Winnie-the-Pooh. Pooh’s getting into some trouble there.
He’s making a mess. And then there’s Eeyore.
He’s planting a flower. BRIAN: Now, in this simulation
self-injurious behavior is occurring to demonstrate how extinction would be used. However, during typical application,
one would expect that over a series of sessions self-injurious behavior would
eventually decrease to a near zero rate. JENNIFER: You need another
match of Owl flying a kite. I like to fly kites, do you?
I bet you do. BRIAN: Now, after the combination
of non-contingent reinforcement and extinction results in a low rate of
the self-injurious behavior, we enter the next phase of treatment,
during which the therapist attempts to establish a more socially appropriate
attention-getting response. Now, frequently throughout the session
the therapist will prompt the pre-selected target behavior, in Henry’s case
recall that it was an arm raise. She will then deliver enthusiastic
praise and initially also deliver an edible. Now, edible reinforcement
really didn’t have anything to do with the occurrence of
self-injurious behavior, but we used it with Henry to initially make
the arm raise, the replacement behavior, more valuable than
the former behavior, the head-banging. JENNIFER: Oh, nice job
raising your hand to get my attention. Have a chip.
They’re delicious. How’s it going, bud? Hey, Henry. Good job
raising your hand. Have a chip if you want. I like how you’re making
all those matches. Those are cool. BRIAN: When Henry does
engage in the self-injurious behavior, the therapist explicitly
ignores that response, so it’s being placed on extinction. JENNIFER: Nice job, Henry.
I like how you raised your hand. Look at all those Piglets
you have over there. Those are cool. BRIAN: And so as the session
continues, the therapist will be frequently prompting the replacement
behavior, delivering reinforcement, or terminating all interaction
following self-injurious behavior. JENNIFER: Great job
raising your hand, Henry. Here’s a chip if you want.
Look at all of those matches you’ve made. Are you having a good time?
This looks like a cool game. BRIAN: Now, once Henry
begins to initiate independent arm raises, the therapist will then decrease
the frequency of prompting, will prompt as necessary and deliver reinforcement,
but will also deliver reinforcement following every independent occurrence of the arm raise. JENNIFER: Great job raising
your hand, Henry. Here’s a chip. You’re still having fun
playing this game, huh? Hey, nice job raising
your hand, buddy. You can have a chip if you want.
Those are so good. They’re from Jimmy John’s. My favorite. Hey, good job raising your hand.
That was awesome. Have a chip if you want. Nice job raising
your hand, Henry. Here’s a chip. This is a fun matching game, huh? Hey, nice job, bud.
Have a chip. I really like how you’re
raising your hand all by yourself, buddy. You’re doing such a good job
matching all of these things, too. Hey, awesome job raising
your hand, my friend. Have a chip. BRIAN: And as sessions continued
during Henry’s treatment, as you saw from the data, the frequency of self-injurious
behavior remained low, eventually going to zero, and the frequency of independent
arm raises continued to increase. JENNIFER: Nice job raising
your hand. You can have a chip if you want. That silly Eeyore.
He’s a funny one. BRIAN: This sequence will
illustrate the assessment and treatment of self-injurious behavior
maintained by escape from task demands. Jennifer Haddock is playing
the role of the student/client who is engaging in self-injurious behavior
consisting of hitting her head. And Travis Jones will
play the role of the therapist. Now, first, how do we
identify that self-injurious behavior is maintained by escape? Well, again we conduct
an assessment known as a functional analysis. And the two important conditions
in this assessment are, first of all, the play condition. In the play condition, the individual
has free access to leisure items, no task demands are presented, and no consequences are
delivered following the occurrence of self-injurious behavior. TRAVIS: Wow. I like your star.
Nice work, Jennifer. Keep it up.
Look at you. Those are some great stars. This puzzle is pretty
fun, too, over here. JENNIFER: She lost an eye. TRAVIS: Yeah, she did
lose an eye. Nice job. We’ve got some tape
there to fix her up. JENNIFER: Thanks. [TOY PLAYING MUSIC] TRAVIS: Wow. This music
over here is pretty fun. BRIAN: Occasionally occurrences
of self-injurious behavior may be observed in this condition, but as you can see,
they produce no consequences from the therapist. JENNIFER: That’s a cool dinosaur. TRAVIS: Yes. It is a cool dinosaur. Wow. This is going to be
such a pretty scene when we’re done with it. [TOY PLAYING MUSIC] TRAVIS: Wow. Great music, Jennifer.
You are quite musical. Yeah. Nice job filling
in that piece of the puzzle. Look at you.
It’s all done. Great work. [TOY PLAYING MUSIC] That’s some beautiful
music you’re playing there. Oh, wow. Look at that.
Putting the stuff on the puzzle. That’s nice. BRIAN: The second important
condition of this assessment is what’s known as the demand condition.
And the demand condition is the test condition for problem behavior maintained by
negative reinforcement, typically escape from task demands. Now, in that condition,
the therapist delivers instructions and/or prompts to do work of various
types throughout the session. And the therapist will continue
to go through that sequence of instruction, prompting, and if the individual
engages in compliance, will deliver praise. However, if at any point in time
during the sequence Jennifer, the client, engages in self-injurious behavior,
the therapist terminates that trial. Now, it appears to look like
a time-out, but in fact, it is escape from task demands contingent
upon self-injurious behavior, which equals negative reinforcement. TRAVIS: Put the pencil in the bin. Good work. Put another pencil in the bin. Put the pencil in
the bin like me. You do it. Put the pencil in the bin like this.
That’s how you put the pencil in the bin. BRIAN: Now, as long as
self-injurious behavior does not occur, the therapist continues
to go through the instructional sequence. TRAVIS: Hand me the red chip. Good job. Hand me the black chip. You don’t have to. BRIAN: When self-injurious
behavior occurs, the task is terminated. Touch the pink card. Touch the pink card like me.
You do it. Touch the pink card like this. You don’t have to. Hand me the black–
you don’t have to. Put the pencil in the bin. Good work. Touch the– you don’t have to. Hand me the red chip. You don’t have to. Touch the white card–
you don’t have to. BRIAN: Once we have determined
that self-injurious behavior is maintained by escape from task demands,
we develop an intervention plan aimed at differential reinforcement of
an alternative, that is socially appropriate escape response, and in this
particular case we have selected touching a break card, which is a very
easy response to shape up. Now, the therapist will initially
begin an instructional trial, and before problem behavior occurs,
will prompt that break response and deliver a break. Now, the other component
of intervention is extinction. That is if at any point in time
during the sequence Jennifer would engage in self-injurious behavior,
the therapist does not terminate the trial, rather he prompts her
to complete the task. TRAVIS: You don’t have to. BRIAN: Now, as you can see,
he put out the materials, but he didn’t even deliver an instruction before she had
the opportunity to engage in self-injurious behavior. He prompted the response,
followed it with escape from the task demand. TRAVIS: Break time. Touch the white card. Break time. Touch the yellow card. Touch the yellow card like me.
You do it. Touch the yellow card like this.
That’s touch the yellow card. Touch the white card. Break time. BRIAN: Now, on those two trials,
first you observed that Jennifer engaged in self-injurious behavior,
and that did not terminate the instructional sequence. On the subsequent trial, the therapist
prompted the alternative response and allowed escape. TRAVIS: Put the pencil in the bin. Put the pencil in
the bin like me. You do it. Put the pencil in the bin like this.
That’s put the pencil in the bin. Hand me the black chip. Like this. You do it. Hand me the black chip like that.
That’s hand me the black chip. Hand me the red chip. Hand me the red chip like this.
You do it. Hand me the red chip like that.
That’s hand me the red chip. Hand me the black chip. Good work. Nice job. Put the pencil in the bin. Break time. Hand me the red chip. Hand me the red chip like this.
You do it. Break time. Put the pencil in the bucket. Break time. Touch the white card. Break time. Touch the red token. Like me. You do it. Touch the red token like that. Touch the black token. Break time. Put the pencil in the bin. Put the pencil in
the bin like me. You do it. Put the pencil in the bin like this. BRIAN: Now, as sessions progress,
Jennifer learns that an effective way to get a break is to touch the break card,
and so eventually she begins to do that independently. The therapist will continue
to initiate learning trials. If Jennifer does not engage in
self-injurious behavior, he will continue those learning trials. If she engages in self-injurious behavior,
he’ll still continue the learning trials. If, however, she independently
touches the break card, she then gets a break. TRAVIS: Hand me the black token. Hand me the black token like this.
You do it. Hand me the black token like that. Put the pencil in the bin. Put the pencil in the bin like me.
You do it. Break time. Hand me the red token. Hand me the red token like this.
You do it. Good job. Touch the pink card. Touch the pink card like me.
You do it. Touch the pink card like this.
That’s touch the pink card. Put the pencil in the bucket. Put the pencil in the bin like me.
You do it. Put the pencil in the bin like this.
That’s put the pencil in the bin. Break time. Put both pencils in the bin. Nice work. Touch the yellow card. Like me. You do it. Touch the yellow card like this. Hand me the black chip. Break time. BRIAN: So as you can see,
during this treatment session there are three different sorts of
consequences being delivered. If compliance occurs,
praise is delivered. If problem behavior that is
self-injurious behavior occurs, the therapist simply continues the prompting
sequence, not allowing escape. And if the new escape
response occurs, that is touching the break card, Jennifer gets a break. TRAVIS: Hand me the red chip. Good work. Touch pink. Break time. Put all three pencils in the bin. Put all three pencils in
the bin like me. You do it. Put all three pencils
in the bin like that. Hand me the black chip. Break time. [MUSIC PLAYING] CLOSED CAPTIONING PROVIDED BY

Leave a Reply

Your email address will not be published. Required fields are marked *