ABA and Misrepresentation
Every few years the issue of whether ABA is quackery or not kicks up on the Hub. It seems to be cyclical. It is a debate I have been a part of before and seems that I have the opportunity to do so again. Recently some discussion along these lines has popped up again. I believe the dissenter’s criticism can be summarized as one or more of the following:
1. ABA is best summarized as quackery.
2. ABA does not have affirmative research.
3. ABA destroys or reduces a different autistic style of learning.
4. ABA is emotionally traumatic to autistics.
5. ABA intrinsically leads to physical or emotional abuse.
Criticism can be good, it is certainly necessary that criticism be given, but not all criticism has merit. It is our job (that means all of us) to scour what is said for merit. Because the five points listed above are troubling, I believe they deserve to be addressed in a serious and frank manner. I have selected to do so on my blog because it is the most powerful and open advocacy tool I posses and I believe the issue merits it.
My responses:
1. ABA is best summarized as quackery.
Quackery is best defined not by possessing an outlandish theory, but by failing to allow research to inform theory. In quackery the rules are set, the theory is inviolate. The more things change the less quackery does. However, behavior analysis does change in accordance with research. The theory is not inviolate. It adapts. For example as an upper level undergraduate I was impressed and quite taken with some research that suggested that stimulus-stimulus pairing might help non-verbal students increase their vocalizations. It made crystal clear sense theoretically. There was an article or two that showed an effect. So, I busily engaged myself drawing up research that could help replicate or extend the initial research. However, new research appeared by time I began to work on this issue. The research showed little to no effect via this technique.
I was severely disappointed. I had worked hard in the initial stages or writing and now my idea was shot all to pieces. That is the way it goes in science. If you are not willing to look stupid you can’t do science and I certainly felt foolish at the time. But it wasn’t just myself….. others who were interested in the area also moved on when the new research emerged. The field adapted in response to the research.
But there are other examples as well; evidence shows problems with the extinction burst and humans. All of this is discrepant with theory. Further, a major review was done several years past concerning all existing applied research in the area of verbal behavior. The authors found some ideas were well supported, others not so well at all. These are just a few exmaples.
If one searches, one can probably find an anti-scientific or quackish remark from one or more behavior analysts somewhere. Probably even from Dr. Skinner somewhere. I see this as irrelevant however. The best description of the general case of behavior analysis is a field of study that changes in response to research findings. That rules out quackery.
2. ABA does not have affirmative research.
First I want to distinguish between two types of research. Group design and single case design. Both, if designed correctly can answer a question in a powerful and legitimate way. However, it depends on the question and upon the specific design. There is no such thing as a one size fits all research design.
In autism and ABA there have been a number of research examples using group design. Some of these studies have received a fair bit or criticism (of varying merit) some of which has come from other practitioners of ABA. I accept some of the criticism of certain studies as correct, but not so for others.
However, the majority of the work concerning applied research of ABA in autism comes not from group studies, but from single case design. For specific techniques the evidence record is excellent and very broad (well over 1000 studies). One can correctly assess the efficacy of ABA without assessing the single case design. Moreover some techniques researched in ABA are shared concepts with the cognitive paradigm. Mutual exclusivity comes to mind. The research here has even been a cooperative effort for certain research examples. Moreover there are shared research endeavors between speech pathologists and behavior analysts. Again, the research record for certain techniques is very strong.
3. ABA destroys or reduces a different autistic style of learning.
I admit the above idea is possible, but I don’t think so. Nor does it have anything like an adequate proof at this time. But one can make predictions based on the theory and see how well the theory holds up. For example based on this theory we might predict that ABA reduced or destroys an autistic child’s ability to acquire high level splinter or savant skills. But I myself know several examples of children who participated in ABA who acquired savant skills while they were still in ABA or just after. So, already we know from anecdote alone that ABA and the acquisition of savant skills are not exclusive.
However, this does not answer the question of whether ABA interferes to a lesser degree with the acquisition of these skills. The way to test this would be to assess a large group of students in ABA against a large group in an alternative learning situation that is as unlike ABA as possible. A matched pairs design could be used with the matching to be done along the lines of the existence of certain materials in the home and school environment. The numbers children with savant skills could then be compared.
This would be a good first step in proving the above theory. But this burden falls entirely on the shoulders of the advocates of the theory. And of this time, it is merely theory.
4. ABA is emotionally traumatic to autistics.
I have no doubt that this is the case for some autistics. I know of none, but I am sure it is the case. Just like it is the case for certain non-autistic persons going to a mainstream school. However, I believe it is incorrect to generalize from a specific case to the general case in both instances. In my experience it is not true that autistic children in contemporary ABA problems cry more than any other program with children of comparable age. Nor do they seem less happy to from what I can tell. Nor is there any data to suggest that ABA causes PTSD.
5. ABA intrinsically leads to physical or emotional abuse.
This is rubbish pure and simple. There is not a shred of data to support this. I know certain behavior analysts who are incredibly conscientious about this issue. I remember being taught by a well known behavior analyst to avoid even mildly aversive procedures like time-out like the plague and how to “go to a knee” or even sit down to avoid being intimidating to a student if we had to talk about a behavior problem.
I believe this to be relatively close to the general contemporary case. I know there was physically abusive behavior committed by early behavior analysts in the 60s-90s. I feel no inclination to answer for these which I have clearly stated I am against, than I do about physically abusive behavior committed by non- behaviorists during that same time span or even today.
1. ABA is best summarized as quackery.
2. ABA does not have affirmative research.
3. ABA destroys or reduces a different autistic style of learning.
4. ABA is emotionally traumatic to autistics.
5. ABA intrinsically leads to physical or emotional abuse.
Criticism can be good, it is certainly necessary that criticism be given, but not all criticism has merit. It is our job (that means all of us) to scour what is said for merit. Because the five points listed above are troubling, I believe they deserve to be addressed in a serious and frank manner. I have selected to do so on my blog because it is the most powerful and open advocacy tool I posses and I believe the issue merits it.
My responses:
1. ABA is best summarized as quackery.
Quackery is best defined not by possessing an outlandish theory, but by failing to allow research to inform theory. In quackery the rules are set, the theory is inviolate. The more things change the less quackery does. However, behavior analysis does change in accordance with research. The theory is not inviolate. It adapts. For example as an upper level undergraduate I was impressed and quite taken with some research that suggested that stimulus-stimulus pairing might help non-verbal students increase their vocalizations. It made crystal clear sense theoretically. There was an article or two that showed an effect. So, I busily engaged myself drawing up research that could help replicate or extend the initial research. However, new research appeared by time I began to work on this issue. The research showed little to no effect via this technique.
I was severely disappointed. I had worked hard in the initial stages or writing and now my idea was shot all to pieces. That is the way it goes in science. If you are not willing to look stupid you can’t do science and I certainly felt foolish at the time. But it wasn’t just myself….. others who were interested in the area also moved on when the new research emerged. The field adapted in response to the research.
But there are other examples as well; evidence shows problems with the extinction burst and humans. All of this is discrepant with theory. Further, a major review was done several years past concerning all existing applied research in the area of verbal behavior. The authors found some ideas were well supported, others not so well at all. These are just a few exmaples.
If one searches, one can probably find an anti-scientific or quackish remark from one or more behavior analysts somewhere. Probably even from Dr. Skinner somewhere. I see this as irrelevant however. The best description of the general case of behavior analysis is a field of study that changes in response to research findings. That rules out quackery.
2. ABA does not have affirmative research.
First I want to distinguish between two types of research. Group design and single case design. Both, if designed correctly can answer a question in a powerful and legitimate way. However, it depends on the question and upon the specific design. There is no such thing as a one size fits all research design.
In autism and ABA there have been a number of research examples using group design. Some of these studies have received a fair bit or criticism (of varying merit) some of which has come from other practitioners of ABA. I accept some of the criticism of certain studies as correct, but not so for others.
However, the majority of the work concerning applied research of ABA in autism comes not from group studies, but from single case design. For specific techniques the evidence record is excellent and very broad (well over 1000 studies). One can correctly assess the efficacy of ABA without assessing the single case design. Moreover some techniques researched in ABA are shared concepts with the cognitive paradigm. Mutual exclusivity comes to mind. The research here has even been a cooperative effort for certain research examples. Moreover there are shared research endeavors between speech pathologists and behavior analysts. Again, the research record for certain techniques is very strong.
3. ABA destroys or reduces a different autistic style of learning.
I admit the above idea is possible, but I don’t think so. Nor does it have anything like an adequate proof at this time. But one can make predictions based on the theory and see how well the theory holds up. For example based on this theory we might predict that ABA reduced or destroys an autistic child’s ability to acquire high level splinter or savant skills. But I myself know several examples of children who participated in ABA who acquired savant skills while they were still in ABA or just after. So, already we know from anecdote alone that ABA and the acquisition of savant skills are not exclusive.
However, this does not answer the question of whether ABA interferes to a lesser degree with the acquisition of these skills. The way to test this would be to assess a large group of students in ABA against a large group in an alternative learning situation that is as unlike ABA as possible. A matched pairs design could be used with the matching to be done along the lines of the existence of certain materials in the home and school environment. The numbers children with savant skills could then be compared.
This would be a good first step in proving the above theory. But this burden falls entirely on the shoulders of the advocates of the theory. And of this time, it is merely theory.
4. ABA is emotionally traumatic to autistics.
I have no doubt that this is the case for some autistics. I know of none, but I am sure it is the case. Just like it is the case for certain non-autistic persons going to a mainstream school. However, I believe it is incorrect to generalize from a specific case to the general case in both instances. In my experience it is not true that autistic children in contemporary ABA problems cry more than any other program with children of comparable age. Nor do they seem less happy to from what I can tell. Nor is there any data to suggest that ABA causes PTSD.
5. ABA intrinsically leads to physical or emotional abuse.
This is rubbish pure and simple. There is not a shred of data to support this. I know certain behavior analysts who are incredibly conscientious about this issue. I remember being taught by a well known behavior analyst to avoid even mildly aversive procedures like time-out like the plague and how to “go to a knee” or even sit down to avoid being intimidating to a student if we had to talk about a behavior problem.
I believe this to be relatively close to the general contemporary case. I know there was physically abusive behavior committed by early behavior analysts in the 60s-90s. I feel no inclination to answer for these which I have clearly stated I am against, than I do about physically abusive behavior committed by non- behaviorists during that same time span or even today.
posted by Interverbal at 11:20 PM
37 comments
