Saturday, July 26, 2008

Much Better MSN Feingold Article

After a highly credulous Feingold article last summer, MSN manages to get someone who knows what they are talking about. Our favorite online news-source medical expert….Dr. Rob does a reasonable job providing factual information about food sensitivities and ADHD.

For the casual reader, the Feingold diet was an elimination diet that gained some popularity in the 70s and 80s as a treatment for hyperactivity and ADHD. There was some early research that indicated it might be useful, but by the mid 90s it was clear that the larger, better controlled studies were not showing an effect.

As always none of this deterred the believers, as the comments from last year show. But that is another story for another day. Thank you MSN!

Sunday, July 06, 2008

Criticism and Behavior Analysis

Well here we are again, in what is quickly turning into a very broad disagreement. Christschool's latest response "But this is Sparta!" is now out. My reply will be in point-counterpoint format.

"My Rebuttal: I never stated that there hasn't been group studies in which non Lovaas behaviorists have done. What I stated, which is contained in the quote you used was that the 47% myth has "never been replicated independent of Lovaas' small group of behaviorists"."

Oh I see; but I notice that a little bit of verbal dexterity is required to make this work. There has been a replication that produced 48% by a Lovaas student and there have been other non-Lovaas students who produced statistically significant research in this regard. However, there has certainly never been a non-Lovaas student who produced the famous 47%. Well then, I cheerfully cede the point.

"Thus, using Interverbal's own self selected reference we can see that there has been no replication of Lovaas' 1987 47% myth according to Sallow, Graupner (2005)."

To support this, Christschool sites the following from Sallow and Graupner (2005):

"There have now been several reports of partial replication without using aversives"

Of course, this is not what Sallows and Graupner argued at all. By defintion we can only produce a partial or systematic replication, because a portion of the indepedent variable (aversives) are absent. Sallows & Graupner are clear and up front about this.

However, a systematic replication is still a replication. This type of replication either extends the population parameters of the original study (external validity) or allows us to analyze changes in the independent variable. Sallows & Graupner do both. It is a different sort of replication, but it is a replication regardless.

"Based on my background as a commercial bankerHYPERLINK "", I would make an educated guess that 800 employees would translate into about a revenue number of nearly $30 - $ 50 million per year ( I have no proof of this number, but like Interverbal states when he says behaviorists are qualified to discount possible PTSD complications from ABA because behaviorists work with lots of autistics and are qualified to make this determination, I too wish to employ my expert background in commercial banking to assert that revenue number)."

I do not take issue with Christschool's argument, but I wish he would not misrepresent mine either. I do not argue that behavior analysts are qualified to discount PTSD, unless of course they have credentials and experience in the area. Some behavior analysts do have expertise in PTSD. One of my undergraduate mentors was an expert and researcher in PTSD. However, I argue that we see no evidence of it. If we see no evidence of it and no science exists to the contrary why would we study it?

I analyzed this in my last post using a rhetorical device; namely, displacing certain words words into Christschool's argument to create a new argument that accurately reflects another view of autism that both Christschool and I oppose. The point I made was that Christschool's logic was poor in this regard.

"For instance, and I don't mean to pick on Interverbal here, but when I stated that vaccines have caused autism in children (based on information supplied to me by 2 children who had been diagnosed with autism as a result of undergoing vaccines as children) and that there aren't any studies that have been done to look into possible complications from vaccines such as autism, his response wasn't what I would expect from a curious and science based person."

I think this was a poor argument when it had the words ABA and PTSD and I think it is still a poor argument now.

"Afterall, DAN! has scienceness studies too and all produced by DANNITES!"

No, it does not have applied science. This is a false comparison. The biom advocates have some basic lab science, but not applied science. Although, there have been some limited studies of both the GF/CF diet and Mb12 shots. Both studies indicated no effect.

"Similarly, researchers outside Lovaas' circle can't produce the 47% figure either, only those "certified" Lovaas replication sites can produce the 47%, oops, I mean the 48% indistinguishable criteria."

A callous argument to the point of error. A change from 47 to 48% will be within the band of error. Moreover, despite the verbal dexterity used to frame this argument, there is existing research with statistically significant results which was not produced by Lovaas' students.

"I purposely did not refer to Sallows in my previous post, "Skinner Confuses Science with Terminology", because he is just too easy to discredit on circumstantial grounds alone"

Only if we employ a fallacy.

"Is this what passes for ethics in Behaviorism? Is this what passes for science in Behaviorism? Lies, damn lies and emotive marketing. No different than DAN! and just as scienceness."

Individuals make their own choices. The WEAP site does not reflect the way I or others discuss ABA. Nor does this tell us anything about behavior analysts' ability to accept criticism.

"Interverbal, I stand by my statement. No replication has ever occurred outside one of Lovaas' inner circle of behaviorists and I disagree with you. This is Sparta!"

I acknowledge that with the specific wording you are right. Frankly, I think the point is worthless? My concern is that a replication exists at all, not who did it. Although, I notice once again that others also achieved statistical significance who were not in Lovaas' group.

As to Sparta, I think I have demonstrated that behavior analysts can and do accept criticism when it is shown to have merit. You have not cogently argued this point. Instead you give an example of one center advertising in a way we would both disagree with.

Christschool, if you had argued that "sometimes behavior analysts incorrectly quote ABA science or autism science", then we would not be having this debate. We would be in total agreement. Instead you argued broadly that behavior analysis, is unscientific, a point you have massively failed to support. The best evidence you have for your original argument, that behavior analysis is unscientific (in other words that this is Sparta) are the dancing Spartan heads on your blog.

Saturday, July 05, 2008

Not Sparta

The following post is offered as a reply to Christschool's recent post "Skinner Confuses Science with Terminology". - Noam Chomsky". My responses to Christschool will be in a point by point basis.

"It's been my experience that behaviorists are quick to recoil and become defensive with inquiry that challenges their "science". From my observation, behaviorist's reaction to skepticism is very similar to the alternative/biomed advocates."

I think it first necessary to draw a distinction between the science of behavior analysis and the philosophy of behaviorism. These are not quite the same thing. It is possible to practice the techniques of behavior analysis and not philosophically be a behaviorist. I would argue this happens regularly in clinical psychology.

Some may ask; doesn't it seem strange that there is both a science and a related philosophy? But I would argue that this is not strange at all. I myself belong to several philosophies which complement and direct my inclination to science, but are not part of science itself. These include the philosophies of skepticism, empiricism, determinism, and materialism.

But to return to the point, yes, some behavior analysts can be very quick to dismiss criticism. Our approach can be dogmatic. I would even argue that as a field, behavior analysts have been worse than average at accepting criticism from those outside the field. One of the ways I have selected to help counter this trend is to operate a blog, which at least on occasion, delves into behavior analytic issues.

However, it would be faulty logic, a non-sequitur in fact, to assume that because behavior analysts are on average worse at accepting criticism from the outside, that the science of behavior analysis is a pseudo-science.

"Both groups cite as evidence studies, anecdotes and faux terminology to support their positions."

The idea that it is faux terminology is the idea one would need to first prove. This is the fallacy of begging the question. As to the anecdotes, he is correct, even though this is a person specific error.

"The behaviorists use
Lovaas' 1987 study and his unique 47% indistinguishable criteria HYPERLINK ""(which has never been replicated independent of Lovaas' small group of behaviorists) to sell their "services".""

This is simply not true. There have been any number ABA autism group studies that were not conducted by Lovaas' immediate students. An even basic review of literature will show this to be the case. Also, Sallows & Graupner (2005) state:"We found that 48% of all children showed rapid learning, achieved average posttreatment scores, and at age 7, were succeeding in regular education classrooms. These results are consistent with those reported by Lovaas and colleagues (Lovaas, 1987; McEachin, Smith, & Lovaas, 1993)."

One can argue that Sallows & Graupner (2005) also had unresolved threats to the validity of their study (and I would agree) but one can not argue that they produced different results from Lovaas (1987).

Furthermore, it is the ad hominem fallacy, to argue that one is wrong and/or dishonest because of whom one is. The fact that some of the research has originated by Lovaas' students should not bias us in interpreting the results. I do not argue that Christschool is wrong for the following reason, but I will point out that the group I usually associate with this fallacy are the vaccine etiology of autism advocates. This seems like an appropriate comparison for the sake of accuracy, since Christschool states that his latest article compares behaviorism and the alternative/biomed advocates.

"However, they conveniently leave out from their marketing materials the fact that the children in that study were never randomized and were subject to physical aversives. Is that honest or is that protecting the dogma of behaviorism?"

Fair question, behavior analysts like any human service providership must make sure the public and specifically consumers have enough information to make an accurate judgment. This has not always been done correctly in my view. However, Christschool had best be careful to acknowledge that this is individual specific. It is not nearly true of all behavior analysts.

"Do scientists leave out very relevant facts (no randomization and the use of physical aversives) in discussing their results with prospective customers?"

Yes, sometimes they do. I don't agree with this practice, but this failure is hardly limited to behavior analysts. One of the most common (and most valid) criticisms we hear from the anti-vaccine movement, is the lack of correct sharing of information by pediatricians and vaccine manufacturers. Nor is it even specific to the sciences in the human service fields. This is a valid criticism of behaviorism (in some cases), but to be consistent, Christschool would have to be willing to apply this criticism broadly across many fields of science. I don't want a problem that exists in the broader subfields of science to be attached merely to behavior analysis.

"If challenged, both groups simply recoil like a viper"

Yes, some behavior analysts and advocates can react this way to criticism. Not all do. Some of my formative experiences in behavior analysis were learning the polite (or mostly so) give and take during debate about behavior analysis with course instructors and professors. So, when faced down by a glaring 19 year old angry about the accuracy of the Matching Law, what do you suppose the much older and well known professor did? Tell him to be quiet? Tell him he just didn't get it? Mark down his grade in participation points? No, he invited me to have an in depth discussion after class. The professor and I spent two hours going over the issue. During which time I was treated with patience and politeness.

That issue has still never been resolved to date. But that professor provided a wonderful example of how to approach disagreements. How many venerable professors in any field could be so patient and selfless with their time? Now the point of my anecdote is not that a dogmatic approach never occurs, but that here is at least one example were it did not. I would argue that there are others..... Many others.

"ready to strike rather than acknowledging that perhaps they could be wrong."

But doesn't this depend on the quality of the criticism? And why criticism may always be necessary to guard against dogma, not all criticism has merit. It is not the job of scientists to acknowledge any criticism, but to scour it for merit and see if it contains it. Not all criticism has this. There is some criticism so bad, it merits serious criticism in and of itself. That is the catch in science or scientific criticism. You can easily end up looking like a fool. And if you are not willing to take this risk, then you can not really do science of scientific criticism. And this is true for any advocacy movement as well.

"For instance, and I don't mean to pick on Interverbal here, but when I stated that ABA has caused PTSD in autistics (based on information supplied to me by 2 autistics who had been diagnosed with PTSD as a result of undergoing ABA as children) and that there aren't any studies that have been done to look into possible complications from ABA such as PTSD, his response wasn't what I would expect from a curious and science based person.

Me:"The field hasn't done any studies that I know of that look at PTSD in those that went through ABA

"Interverbal:"Nor should they. There is no real suggestion of an ABA-PTSD connection. The suggestion there could be, was made and continues to be echoed mostly by those in the psychodynamic paradigm. There are lots of genuinely good criticisms of behavior analysis and ABA in autism specifically. A possible PTSD connection isn't one of them. This is the type of comment that behavior analysts laugh off and correctly so.""

Christschool asks me to accept as evidence second hand information I have no way of checking. I can't possibly weigh this information against what I know about science, because I can't possibly access this information at this time. I do not accept this argument for behavior analysis and I would not accept it at any other time either. I am happy when I see that I have made an error to acknowledge it in full and move on. However, there is no error here. I completely reject Christschool's criticism in this regard.

Here is a rhetorical exercise that helps show this point:

"For instance, and I don't mean to pick on Interverbal here, but when I stated that vaccines have caused autism in children (based on information supplied to me by 2 children who had been diagnosed with autism as a result of undergoing vaccines as children) and that there aren't any studies that have been done to look into possible complications from vaccines such as autism, his response wasn't what I would expect from a curious and science based person."
"Based on Interverbal's representation of behaviorist's views, there is no need to accumulate empirical data in order to draw a scientific conclusion."

This was not my argument. My argument was there is insufficient evidence to justify a study in the first place. Also, behavior analysts will laugh this off and rightly so. Not because it is a criticism, but because it is a very poor criticism. Many behavior analysts in the field have worked with hundreds of children and adults with autism. None of the ones I have spoken to see evidence of this. Even I, a novice of behavior analyst must have worked with and collected data on over 100 people by now. Nowhere in my experience and in the experience of the people I have spoken to, was there a hint that PTSD was or could be an issue. Where is the reason I should take this criticism seriously?

"One would think that those interested in human behavior would be curious about how their "therapy" might affect the emotional well being of autistics?"

Well, I am "one" and here is what I would think. I think the whole ABA = PTSD is a residue left over from the psychodynamic tradition in autism (and they still say this). They argue that autism is caused by some psychological trauma, leaving the child with autism emotionally more vulnerable to PTSD. I think that this criticism has been adopted by certain parties in the self-advocacy/ally movement in the absence of clear thought on the issue. I think it is telling the some of the more scientifically oriented critics of ABA in this movement do not offer this specific criticism.

"Noam Chomskey is a professor of linguistics at M.I.T. and perhaps one of the most engaging and prolific intellectuals of the 20th century. In 1971, The New York Review of Books published his essay
"The Case Against B.F. Skinner"HYPERLINK "". Chomsky makes the case that Skinner's theory of Verbal Behavior isn't really science but a sort of secular dogma (dogma is my interpretation)."

I am afraid I must argue against an engaging intellectual. However, first I will note that Christschool confuses the linked Chomsky article and Chomsky's earlier analysis of Skinner's 1957 book "Verbal Behavior". As to the earlier analysis: Some have argued that Chomsky in just 50 pages or so demolishes both Verbal Behavior and behavior analysis in general. I think this is a very fair analysis of these works. Provided one has actually read neither.

Skinner, created a classical book of theory, based on arm chair analysis and on observation of his daughters. This technique has a long history. The developmental psychologist Jean Piaget did just the same in his detailed and excellent "The Origins of Intelligence in Children". Chomsky himself does his fair share of arm chair analysis, as do I.

I won't try to break down Chomsky's criticisms here, some of which are quite sophisticated and good. And some of which have nothing to do with science and a great deal to do with wrangling over terminology (sound familiar?) and musings on just how important free will is.

However, if Chomsky's main criticism is a lack of science to back up Skinner, then I would agree this was a very fair criticism in the 70s. But, I respectfully submit to the reader that it is not 1967 anymore. Both the study of verbal behavior and behavior analysis has matured a great deal since the 60s and 70s and both of Chomsky's critiques. I myself was involved as a research assistant in a major review of all existing research in verbal behavior. In our review, we learned some ideas are well supported and some are not.

Verbal behavior being an interest of mine, I had a topic a few years back that fascinated me. It was stimulus-stimulus pairing to produce speech in non-verbal children. It made crystal clear sense theoretically. But, as more research came out, it did not support it. As pseudo-scientists shouldn't we have cried some intellectual variant of "This is Sparta!" and brushed off the results? Where was my... where was our.... arrogant and dogmatic dismissal of evidence?
Criticism must be able to be given, in the absence of which dogma may creep.

However, I have offered my arguments as to why a number of Christschool's criticisms were without merit. I now leave it to the reader to proceed from here.

Friday, July 04, 2008

Peer Review in the Hub

Christschool and I seem to be getting in our fair share of disagreements these days.Recently he critiqued and old article on punishment that I had done. I responded here, with a rebuttal and some criticisms of my own. Our main bone of contention was the definition of torture. In response to the article and to our subsequent discussion, Christschool retiled his article; it now begins in part with "Interverbal defends Matthew Israel". A hand grenade tossed into a debate, if ever there was one.

Also, in the comments here on my rebuttal, the topic began to shift from whether physical aversives = torture, to the problems of behavior analysis. This has led to a new article from Christschool that addresses some of the broader problems he sees in behavior analysis. It also features quotes from myself to help prove the point, surprise, surprise.

I think, this sort of issue frightens some of our fellow bloggers. They might worry that here are two Hub bloggers having a serious disagreement. Some might wish us to drop the issue and focus on shared points of advocacy. However, I truly think this would be the wrong approach. Any view can become dogmatic. I would argue that if we fail to guard against this, dogma may creep in.

I would also propose that what makes us different here at the hub, is not just a more positive view of autism (something unbelievable to other parties), but that we allow ourselves to dissent if we see the need. I directly contrast this with other options such as the Age of Autism Blog.Don't let Christschool's and my disagreement frighten you merely because we disagree. It is a very serious debate, but it may help both us and our readers better understand the issues at hand.

It is also healthy for the Hub itself. The process of peer review which the hub provides, although casual, is also excellent. Here are statisticians, and geneticists, and professionals, and self-advocates, and parents all come together. But peer review only functions when it is free. And to be free, dissent must be able to be given in a fearless manner.

Some, readers may wonder why I bother to write such a disclaimer (I have already done so before after all), but I would mention that the Hub occasionally picks up new members and also there are always readers who are new to the hub. And even for the older members, it may serve as a healthy reminder. And I hope the reader keeps this in mind as s/he reads Christschool and myself.My reply to Christschool's latest article will follow in a day or so.

Wednesday, July 02, 2008

So.. Lincoln or Darwin?

I just finished reading an outstanding comparative piece, looking
at Lincoln and Darwin. It asks who was the more influential man.
Despite being an inherently absurd question and leading to the inevitable
peeing contest in the comments, the article is simply just excellent. It is also
mostly accurate.

Coming from a scientific and skeptical viewpoint, it is excellent to see Darwin's
life presented correctly in a major media source. Also, the comments are great deal
of fun. Put on a hard hat and grab some popcorn and have at it!

Tuesday, July 01, 2008

A Reply to Christschool

Back in November in 2006, I wrote a somewhat controversial article detailing a portion of the history of behavior analysis and aversives. Some people liked it and some people didn't like it one bit. However, no one has challenged its factual basis. Until now.....

Christschool is the maker of excellent and well known autism advocacy YouTube videos. He is also a fellow Autism Hub member. Very recently he wrote an article criticizing my post for inaccuracy. Specifically, he writes:

"Contrary to Interverbal's analysis that Professor Skinner was "strongly against punishment" and the "behavior analytic community abandoned Skinner's argument", B.F. Skinner was not only an advocate of extreme aversives, but he was literally the architect of the aversives that led to the death of Mr. Milletich.To be Continued...."

To this end he cites the following, culled from this article. "Professor Skinner said: ''I don't like punishment at all. But some people are temporarily out of the reach of positive reinforcement, and a small amount of punishment may help bring them within reach of therapy.'' He also defended Mr. Israel, saying, ''The critics who call what he's doing torture don't know what they're talking about."

I have significant problems with all of this. Beginning with Christschool's title. His title reads:

"Critics of Mr. Israel don't know what they are talking about." - B.F. Skinner"

This quote appears nowhere in the New York Times article. It seems to be based on the following Skinner quote; ''The critics who call what he's doing torture don't know what they're talking about". Christschool's quote is inaccurate and misleading. By adding a few words and taking a few more out, he has radically changed the meaning of Skinner's comment.

Moreover, Skinner is right. It is not torture. The word "torture" means something specific. What certain parties did in terms of using physical aversives in the past, or still do now, might be disagreeable and I would certainly argue against it on ethical lines, but it is not torture. This sort of argument is like accusing parents who choose to use EDTA of poisoning their children. We might disagree with it, we might argue that it is unsafe, but it is not poisoning. This would be a very poor argument and some readers, even if they dislike aversives, are likely to recognize it for what it is, a fallacious appeal to emotion.

Moving on, Skinner consistently wrote concerning his stance against punishment. I argued this in my previous post and very carefully referenced my assertions. I feel no obligation to do so again. Christschool may argue that by backing aversives in unusual circumstances, for a brief period of time, and to a small degree, Skinner should no longer be considered to be considered "strongly against" aversives. Fine, but.... in such a case, I too can no longer be considered strongly anti-aversive. This is news to me, and probably news to a number of other folks both behaviorist and non-behaviorist alike. If the case arises (and indeed it has) of a choice between life-and-death and aversives, I would choose the aversives. In my ethical analysis, death is certainly the greater evil compared to limited and context specific delivery of physical pain.

Christschool, also evidently took issue with my statement that the wider behavior analytic community rejected Skinner's arguments. This is bizarre; I can only guess that Christschool misunderstood what my words meant. Skinner certainly did argue early in his career that punishment is weak. The behavior analytic community did not take this to heart.

I rather carefully detailed the journal articles starting in the 1950s when punishment based research for humans began to appear with some regularity. The existence of punishment focused research, the green light by peer reviewers, and the lack of dissent in letters to the editor, is a clear indicator of a general lack of agreement with Skinner in this regard. In fact, in an extremely telling moment, one of the first letters to the editor where dissent is provided concerning aversives; the letter writer actually congratulates the researcher for abandoning the Skinnerian dogma that punishment is ineffective.

Finally, Christschool writes, "B.F. Skinner was not only an advocate of extreme aversives, but he was literally the architect of the aversives that led to the death of Mr. Milletich."

This would be most unusual. There is no journal record to suggest that Skinner researched these techniques, they are not made reference to in any of his major works for a certainty, and in none of his minor works either from what I can tell. Nor did he employ them with his daughters as they themselves have made very clear. I can only hope that he will provide some sort of evidence for this, in part-two of his series. Although I confess, I am openly skeptical of his claim.

Advocacy, no matter how worthy the cause, needs to be factually accurate. No real service is done in its absence.