Teaching Children with Autism
I am an information junkie. I have more information about teaching children with autism than I know what to do with. Perhaps I will work my way through some of it, wharing various items with you. For the most part, I have no idea where they came from – apologies to those authors that I can’t thank. Furthermore, most of this technical stuff is well beyond the capacity of my little brain. J
Mon Apr 30, 2007 7:57 am (PST)
I appreciate and welcome your cogent, well thought response. We are all students in one way or another and it is exchanges such as these that help us learn. So, here’s my take…
Looking through the microscope for a second, we can clearly see that qualitiative and quantitative stimulus changes are the “actions” responsible for subsequent changes in behavior. Since I do a lot of training with non-behavior analytic professionals and families, rather than sticking to stimulus change to mean the host of dimensions that it can often encompass, I delineate between stimulus changes in the antecedent realm (e.g. those procedures designed to change maladaptive behavior) and those changes in the realm of consequences, which include, as you have eloquently pointed out, changes in a given environment produced by access to a reinforcing event (which of course have implications on subsequent antecedent situations, but let’s not complicate things too much here, for sake of relevance to our point).
In the case of socially mediated reinforcement contingencies, the stimulus changes produced by the behavior are the mechanisms underlying what we would refer to as the ‘consequence’ piece of the positive or negative reinforcement contingency. In these cases, this change is one of the “products” that we can measure often in addition to observing any changes in behavior over time. However, with “automatic” reinforcement contingencies, this stimulus change may not be directly observable or, in many cases, measureable. In reality all we can do is hypothesize functional variables based on changes in rates of behavior under controlled environmental conditions. As the article that Regina so kindly posted suggests, more research is needed in this area to determine exactly what variables are responsible for behavioral maintenance. In this case, any discussion of stimulus change and its effects on behavior become the “explanatory fictions” to which you refer. I mean this with no offense to anyone reading our discussion, but this is exactly what prevents Sensory Integration Training / Therapy from being accepted as an evidence-based practice. It’s going to take quite a while for imaging technology and neuroscience to make the leap to applied practice to the extent that imaging data will become the methods to measure dependent variables sufficiently to conduct effective research.
In many cases, the client with autism lacks the verbal repertiore necessary to convey specific changes in his/her internal environment before or after engaging in stereotypy. So, without the imaging technology, will never know exactly what effects the stereotypy or items in the environment associated with these response patterns, have in terms of reinforcing value. In contrast, most people that run, surf, etc. can provide verbal reports of what reinforcers are in effect and what internal stimulus changes are occurring. Neuroimaging studies have clearly demonstrated the release and inhibited reuptake of Serotonin during continuous aerobic exercise (aka “runner’s high) and the release of other endogenous opiates during instances of SIB (documented in studies with subjects during treatment programs to quit drug abuse) have served to maintain these behaviors most durably. As it’s outside of my field of expertise, I’m sure that you could research the popular meds that have aggression and/or SIB suppressing properties and find that many possess actions that act inversely to the contingencies I’ve described above. Dr. Travis Thompson has volumes of information on these subjects and presents data during his talks when touring the country.
These findings have implications for stereotypy in individuals with autism in that the behavior is hypothesized to alter the internal environment through mechanisms that (1) have no direct contact with the external environment or (2) are enhanced through manipulation of some feature of the external environment. Please note that #2 is not to be confused with social mediation, as reinforcer acquisition is not dependent upon the external environment, but potentially enhanced by its introduction into the response chain. This is the basis for my point about the plate in Kelton’s original example acting as a transitive MO rather than an Sd. But clearly, more research is needed here.
So, let’s take this a step further and address more of your points. You are right in that many other variables may enter the equation as potentially affecting the quality / quantity of reinforcement, event to the extent of creating additional reinforcement contingencies. In the running example, we must not forget that people that run religiously, do so because the act of running itself produces the ‘internal’ reinforcers to which I referred above, this being supported by the sources of evidence cited above as well. However, the other contingencies to which you refer, also affect running as a behavior, including the related health benefits and potentially the social benefits of runnining with a partner. I’m sure similar variables are in effect for surfing.
In your newspaper example (which was very good BTW), the newspaper report does act as a transitive MO as the behavior of reading the paper is increased temporarily due to the value of surfing being increased (potentially) if the report predicts conditions very favorable for surfing. Note that all of the autoclitics you used in your textual behavior (i.e. all of the “er”s you added to your ‘adjectives’ in discussing your newspaper example) are reflective of reinforcers of greater magnitude (i.e. value).
However, I disagree with your analysis that the changes in surfing conditions are responsible for reinforcing the behavior of surfing. If that were the case, the surfer would not surf if the newspaper reported conditions were anything less than extremely favorable in your example. Again, the favorable conditions signal a potential increase in the value of the reinforcers produced by surfing, which exist independently, but are obviously affected by the conditions themselves. As an analogy, runners run in all weather conditions and from personal experience, I enjoy running in low humidity and moderate temps much more than running in August in New Jersey, but I still run in August anyway. For lack of a better term, the “foremost” reinforcer is directly produced by the behvior of running or surfing (those internal changes again), with collateral changes in the quality or quantitity of other salient stimuli affecting behavior, reinforcer magnitude, and stimulus-changes when applicable.
I hope this adds to the discussion, rather than detracting or mucking it up. It’s very difficult to discuss concepts this technical without adopting a highly specific verbal repertiore. I realize that in many cases, readers of this forum may not have equal learning history with regards to basic behavioral principles. However, I continue to believe that this discussion has great pedogogical value for all. Thanks for making some wonderful reinforcers available to me….
As a final thought, perhaps hypothesizing what would be necessary to extinuish the behaviors of surfing or running in these examples would help with conceptualizing these ideas…
As always, I welcome any further discussion.
Ronald A. Petrucelli, M.
Board Certified Behavior Analyst – #1031400
Senior Behavior Analyst
Allegro School & Programs
—– Original Message —-
From: jclintmurphy <email@example.com>
Sent: Saturday, April 28, 2007 9:01:17 AM
Subject: [VerbalBehavior] Re: automatic reinforcement discussion
Ok here are the points that I have identified that I am getting stuck with.
1. Ron- “However, the underlying assumption (and all its baggage) behind automatic reinforcement is that the act of engaging in any behavior belonging to the relevant response class selects the behavior to continue.”
Clint- It’s been my understanding that the behavior itself is not responsible for the increase in future probability. My understanding is that the stimulus change produced by the behavior is responsible for the change in some quantitative dimension of the behavior (usually increase in future probability if the stimulus change functions as a reinforcer).
“Just as in Jack Michael’s classic screwdriver example, the plate does not make automatic reinforcement more available; it probably makes it more valuable in the sense that adding it to any behavior of flapping, spinning, waving, etc. renders those behaviors more potent as reinforcers”
Clint- Here again, behaviors are referred to as reinforcers. It seems to me that certain types of stimulus change are more available in the presence of a plate.
“Is the act of running on a treadmillreinforcin g to someone that really likes to jog? – yes probably,but would a run down a road that paralells beautiful mountain or ocean scenery be more reinforcing – more likely, yes. Do the modifications of the external variables change the availability of running or its function as an “automatic reinforcer”? Not necessarily”
Clint- my understanding is that the act of running is not the reinforcer, but rather its product on the environment. Temporary physiological changes (for lack of better words bc I am somewhat clueless in this area)- sweating, increase heart rate, adrenaline, maybe also an effect on private verbal behavior e.g “now I can eat a cheeseburger later and not feel bad” etc. All kinds of other stimuli are differentially available in the presence of a mountain road they may be responsible for increased mountain road running- smells, animals, sights etc. Access to running is not more available but all kinds of other reinforcers are.
“If the behavior is truly automatically reinforced, then the plate is just one stimulus in an idiosyncratic chain that the person with autism has a tendancy toward following most likely due to the disability’s symptoms of overselectivity toward environmental stimuli, strong adherence to routines and a somewhat temporally distant reinforcement contingency
Isn’t this a kind of explanatory fiction? It seems similar to saying Johnny eats dirt because he has Pica. Explaining a behavior by labeling it?
“Bottom line, my theory here is that these idiosyncratic patterns, probably for one of the reasons I cite in point #2 above, can fall into a category of ‘superstitious behavior’ as described by Skinner, not because the plate has nothing to do with the actual reinforcement contingency but because it is not the only way to access the desired outcome. We just have to teach the kid to do something else.
Clint- this is an interesting point. I think this may be true A LOT of the time. But sometimes it may not be true. Some objects are more likely, when manipulated a certain way, to produce some specific stimulus change in my experience. As an example I have an old raggedy t-shirt that is my favorite one bc it is slightly softer than others and other people just don’t get it.
Clint-This is how I have been conceptualizing this entire issue. I will illustrate with a new analogy kinda similar to the mountain one.
I have a friend named Dave that likes to surf. He is always talking about it, and checking surf reports and looking at weather forecasts. Let’s say the forecast come in one day and says 10 ft
swells out of the southeast, slight wind out of the west, waves are clean and glossy. This represents very good conditions for Dave in a number of ways. Now if I was reading you correctly, your analysis would say that this information (surf report) has increased made the
behavior of surfing more reinforcing for him and if you looked at his behavior, he is clearly more motivated to surf bc he is getting his gear together and canceling work and heading to the beach etc. This is how I have been looking at it- Although surfing is equally available during great waves condition and without (he in theory can go surfing on 1 ft waves every day), there are some types of stimulus change differentially available during great waves condition-
longer rides, faster drop-ins, crazier wipe outs, getting tubed, maybe adrenaline etc, maybe more people watching etc. It is these changes that reinforce surfing and behaviors leading up, not the behavior of surfing itself.
Transitive MO’s- There is definitely motivational variables at work here to. The surf report increases the reinforcer effectiveness and increases frequency of behavior that results in getting to surf – the surf report increases the value of ACCESS to surfing. Things that
become more valuable may include- escape from work, a surfboard, wetsuit, and a ride to the beach among other things.
Anyways, maybe I am completely missing your point! Or maybe you’re analysis is offered from a different level, not sure. I felt compelled to post my thoughts so maybe you can clarify where we are departing in our views. From the posts I think I share a similar conceptual viewpoint with Kelton. And also from current and previous posts it is quite clear that you are very fluent so I consider this a learning opportunity.
Thanks for your time,
— In VerbalBehavior@ yahoogroups. com, “ronpetrucelli”
<ronpetrucelli@ …> wrote:
> For some reason Kelton was unable to respond to my post (in response
> to his, etc.) so he sent this to me via personal e-mail. I think it
> continues to extend what I believe to be a productive discussion of a
> current conceptual and applied issue in behavior analysis. To his
> post below I respond as such:
> The reinforcement responsible for selecting something as specific as
> a jelly bean probably has nothing to do with the generalized
> motivation of being ‘hungry’. Did the behavior of eating jelly beans
> evolve over time or did you eat one and find it delicious, thus
> selecting the behavior for future occurrence? What I have found is
> that in the case of these ‘automatically reinforced’ behaviors in
> autistic kids, the relevant response class often comprises multiple
> topographies, although the specific response that is most frequently
> selected (for many reasons) often becomes the ‘ritual’. I think this
> is where the plate comes in. Yes the plate makes spinning better, but
> classifying it as a reinforcer is topography based and cannot account
> for the direct reinforcement- behavior link. If, however, you can
> spin the plate and the child finds that reinforcing, you can make
> that statement more accurately according to current classifications,
> but you’re not dealing with automatic reinforcement any longer, but
> socially mediated positive reinforcement contingencies. I can see
> how a highly idiosyncratic preference for a specific item is common
> between both the jelly bean and plate examples. However, the
> underlying assumption (and all its baggage) behind automatic
> reinforcement is that the act of engaging in any behavior belonging
> to the relevant response class selects the behavior to continue. As
> you have stated, the act of chewing & swallowing things other than
> the jelly bean may not be as reinforcing as the act of chewing and
> swallowing the jelly bean itself. Adding the jelly bean makes the
> behavior highly specific and directly connected to variables located
> in the external environment. In the case of the autistic child
> (theoretically at least), the child does not have to engage in plate
> spinning to be automatically reinforced, nor does the sight of a
> plate necessarily have to acquire stimulus control over any or all of
> the responses in the response class. We should be able to produce the
> same effect by replacing the plate with other objects. Taking your
> logic in a prior post one step further, if this is the case, then we
> can successfully classify behavior based on function rather than
> topography. Just as in Jack Michael’s classic screwdriver example,
> the plate does not make automatic reinforcement more available; it
> probably makes it more valuable in the sense that adding it to any
> behavior of flapping, spinning, waving, etc. renders those behaviors
> more potent as reinforcers. Transitive MO yes, SD, possibly but not
> The specific properties of an item / activity that enables it to
> produce the behavioral changes associated with reinforcement are not
> necessarily as important as the terminal behaviors associated with
> securing it in the applied arena. One may find reinforcement on many
> levels associated with playing on the playground, but the terminal
> behaviors associated with accessing it are rooted in the external
> environment. These are the behaviors that we typically strive to
> change from maladaptive to prosocial. Taking this one step further,
> think about how we teach kids with autism to play with other kids…
> Once we get past these ‘socially mediated’ reinforcement
> contingencies, we can then explore your points about qualitative
> properties of reinforcement. Is the act of running on a treadmill
> reinforcing to someone that really likes to jog? – yes probably, but
> would a run down a road that paralells beautiful mountain or ocean
> scenery be more reinforcing – more likely, yes. Do the modifications
> of the external variables change the availability of running or its
> function as an “automatic reinforcer”? Not necessarily. However, the
> qualitative properties associated with the delivery of the reinforcer
> have changed and will likely shape the person’s behavior in the
> future. More bang for the buck. That’s why I like Jelly Bellies
> better than the cheap stuff:)
> I think the logic that you are following in your molecular
> examination of the nature of the specific reinforcing properties may
> indeed be sound – it makes sense to me. However, toward the goal of
> not putting the cart before the horse, it may be easier to consider
> that we typically treat approach or access behaviors as the socially
> and functionally relevant terminal responses in applied situations.
> Interestingly enough, these underlying processes that affect
> preference with regards to automatic reinforcement or self-
> stimulation are the stuff that debates between OTs and behavior
> analysts are made. I welcome your response to a most illuminating
> Kelton’s message 4/23/07: “I attempted to respond to your post in the
> verbal behavior forum, but
I received an odd web site when I clicked “send”, so I don’t have
> confidence that it made it through. Here is a copy of my response.
> I’d love to get your thoughts on it. Here it is:
> Nice post Ron! Keep ’em coming! I’d like you to humor me a bit more.
> OK, so with regard to your comment here:
> “In the case of playing with GI Joe, plate spinning or hand waving,
> access to items in the environment seems to function more in the realm
> of motivational variables than posessing discriminative properties.
> The items make the behavior / reinforcer more valuable (perhaps
> increasing the reinforcing magnitude of the response?), not more
> This is good stuff, and I’m trying to get a grip on it.
> Here is my counter point to the comment above:
> If your example holds true, then let’s say that I have an appetite for
> jelly beans (analogous to a child who has an “appetite” for
> flickering/spinning things).
> When I see a jelly bean, I then go through the responses involved in
> jelly-bean-consumpt ion which culminate in chewing and swallowing. The
> taste receptors (among others) are probably critical in detecting the
> reinforcing properties of the stimulus.
All of this occurs in the same way as the child who sees the plate and
> goes through the process which culminates in the behaviors that
> produce spinning plates. Presumably the “light receptors” (e.g.
> eyes!) are critical in detecting the reinforcing property of the
> stimulus. Do you see where I’m going with this?
> In both cases the MO is in effect with or without the jelly bean (or
> plate) present. The child and I both are in a state of “wanting” for
> our respective jollies. The items must be visible for the response to
> occur. The actual items act as an Sd of sorts…not an MO.
> Oh sure, I can “chew and swallow” with our without the jelly bean, but
> this lacks the critical “taste” component. Similarly, the child can
> flick or flap his fingers in an attempt to produce the same visual
> effect, but it often just can’t compete with the spinning plate
> effect. We can physically go through the motions at any time, but it
> will lack the real reinforcer in the absence of the jelly bean/plate.
> Your example is very interesting. It’s always interesting how the
> standard formulation of “MO-A-B-C” seems so simple and yet quickly
> degrades into these brain twisters.
> So I’m not sure if I can fully agree with your comment, but it
> certainly serves as food for thought. Perhaps you can straighten me
> out a bit, I could very easily be wrong.