Teaching to the Test
Many of you will have experience with "No Child Left Behind", and with standardized testing as a measure of how well a school serves its students. The argument is that if you test schoolchildren and reward schools that do the best, schools will increase the quality of their teaching. The problem is that there's an increasing tendency to focus on test performance instead of learning--"teaching to the test"--that may yield an increase in a school's funding even though the teaching is actually substandard. Kids whose teachers focus on the test material may be able to get high scores on standardized tests easily, but fail at applying their knowledge or connecting it to other kinds of information.
Okay, let's leave that behind and talk about something different. Let's talk about ABA--excuse me while I drag out the soapbox yet again. The basic premise of ABA is that you have really well-defined goals; then you break those goals down into tiny little parts; then you specifically teach each part by rewarding correct attempts and either punishing or ignoring incorrect ones. In many clinical trials, ABA has been found to be superior to wait-list control groups in reducing the symptoms of autism.
Yes, that's right, I admitted it: Children who undergo ABA will score lower on tests of autism severity. Standardized tests of autism severity.
ABA, more than any other method, focuses specifically on symptoms. If you want to reduce a child's stimming, you focus specifically on that, with a written-down plan, well-defined rewards, well-defined consequences for making specific movements of the type you want to extinguish. If you want to teach a child to make eye contact, you focus specifically on that, reward him for looking at your eyes; punish him for looking away. If you want to teach him to greet someone "appropriately" with "Hi, how are you?" rather than "Did you know there are a hundred and seventy-five billion galaxies in the observable universe?" then you specifically reward him for that behavior. Sticker charts, candy, loss of privileges, even in some cases physically painful aversives all serve to corral the child into a particular, pre-defined pattern of behavior.
ABA focuses, more than any other method, on the specific items you would see on a standardized scale of autism severity.
Are we seeing the problem here yet?
There are young children who can easily spit out the answer to "5+9=14", but don't particularly know what addition means, or how to apply the idea to "if you have nine cookies and Mother gives you five more". There are older children who know that Egyptians used hieroglyphics to write, but don't connect the idea to the similar written language of Chinese pictograms. High-schoolers may know how to use the quadratic formula; but they have no idea how what they're doing connects to the concept of square roots.
Autistic children who go through ABA will be taught, specifically, to mask the signs of autism. They may greet you with a look in the eye, a smile, and "Hi, how are you?", but will they know how to read the expression in your eyes? Will they know that "Hi, how are you?" is not just what you are supposed to say when somebody greets you, but what its function is as an acknowledgement of the other person's presence and an invitation to social interaction? Will they know that a smile can mean not just "happy" but "I don't intend to be hostile"; and will they know that a smile can lie? Will they know that "Heya, little lady!" is a greeting, too, or will they be confused?
Just like teaching to standardized tests, ABA is very, very weak on teaching basic concepts; teaching the reasons behind things; teaching things in such a way that they can be generalized and applied to other things. ABA, being a standardized technique that works basically the same way for every child once the target areas are identified, doesn't take into account the child's personal strengths; nor does it attempt to help him deal with weaknesses except by brute-forcing the memorization of coping techniques that may or may not be helpful. (Counting to ten does not work for everyone who gets angry, for example.) ABA doesn't take into account that some autistic traits are benign or even extremely beneficial--or that these traits are different from child to child--and thus loses any benefit that can be gained from them as well as causing additional problems when the child loses access to important, though odd-looking, coping skills. And, worst of all, it is totally insensitive to the desires, thoughts, and interests of the child, unless using them as motivations to elicit desired behavior.
ABA can be useful for some things. The method of breaking things down into small parts and teaching them specifically works for things that don't have to be generalized to anything else, that don't need to go past rote memory, and that don't interfere with a child's personality. ABA to teach a child to dress himself, for example, makes sense (there will still be some problems generalizing to different kinds of clothing); ABA to teach him how to write a letter to his grandmother makes no sense at all.
But ABA works, at least according to the measures by which it's tested--measures that are sensitive only to specific symptoms of autism, not to whether the child is happy or whether he is really learning, or whether he is using his gifts, or whether he communicates what is really on his mind rather than what he has been prompted to say.
By the tests, ABA is superior. Long-term outcomes, in my opinion, will tell a different story. ABA may make you perfect for an institutional environment--it sure teaches you to do what you're told!--but does it do anything to increase your ability to adapt, to learn, to reach your potential? I don't think so.
Okay, let's leave that behind and talk about something different. Let's talk about ABA--excuse me while I drag out the soapbox yet again. The basic premise of ABA is that you have really well-defined goals; then you break those goals down into tiny little parts; then you specifically teach each part by rewarding correct attempts and either punishing or ignoring incorrect ones. In many clinical trials, ABA has been found to be superior to wait-list control groups in reducing the symptoms of autism.
Yes, that's right, I admitted it: Children who undergo ABA will score lower on tests of autism severity. Standardized tests of autism severity.
ABA, more than any other method, focuses specifically on symptoms. If you want to reduce a child's stimming, you focus specifically on that, with a written-down plan, well-defined rewards, well-defined consequences for making specific movements of the type you want to extinguish. If you want to teach a child to make eye contact, you focus specifically on that, reward him for looking at your eyes; punish him for looking away. If you want to teach him to greet someone "appropriately" with "Hi, how are you?" rather than "Did you know there are a hundred and seventy-five billion galaxies in the observable universe?" then you specifically reward him for that behavior. Sticker charts, candy, loss of privileges, even in some cases physically painful aversives all serve to corral the child into a particular, pre-defined pattern of behavior.
ABA focuses, more than any other method, on the specific items you would see on a standardized scale of autism severity.
Are we seeing the problem here yet?
There are young children who can easily spit out the answer to "5+9=14", but don't particularly know what addition means, or how to apply the idea to "if you have nine cookies and Mother gives you five more". There are older children who know that Egyptians used hieroglyphics to write, but don't connect the idea to the similar written language of Chinese pictograms. High-schoolers may know how to use the quadratic formula; but they have no idea how what they're doing connects to the concept of square roots.
Autistic children who go through ABA will be taught, specifically, to mask the signs of autism. They may greet you with a look in the eye, a smile, and "Hi, how are you?", but will they know how to read the expression in your eyes? Will they know that "Hi, how are you?" is not just what you are supposed to say when somebody greets you, but what its function is as an acknowledgement of the other person's presence and an invitation to social interaction? Will they know that a smile can mean not just "happy" but "I don't intend to be hostile"; and will they know that a smile can lie? Will they know that "Heya, little lady!" is a greeting, too, or will they be confused?
Just like teaching to standardized tests, ABA is very, very weak on teaching basic concepts; teaching the reasons behind things; teaching things in such a way that they can be generalized and applied to other things. ABA, being a standardized technique that works basically the same way for every child once the target areas are identified, doesn't take into account the child's personal strengths; nor does it attempt to help him deal with weaknesses except by brute-forcing the memorization of coping techniques that may or may not be helpful. (Counting to ten does not work for everyone who gets angry, for example.) ABA doesn't take into account that some autistic traits are benign or even extremely beneficial--or that these traits are different from child to child--and thus loses any benefit that can be gained from them as well as causing additional problems when the child loses access to important, though odd-looking, coping skills. And, worst of all, it is totally insensitive to the desires, thoughts, and interests of the child, unless using them as motivations to elicit desired behavior.
ABA can be useful for some things. The method of breaking things down into small parts and teaching them specifically works for things that don't have to be generalized to anything else, that don't need to go past rote memory, and that don't interfere with a child's personality. ABA to teach a child to dress himself, for example, makes sense (there will still be some problems generalizing to different kinds of clothing); ABA to teach him how to write a letter to his grandmother makes no sense at all.
But ABA works, at least according to the measures by which it's tested--measures that are sensitive only to specific symptoms of autism, not to whether the child is happy or whether he is really learning, or whether he is using his gifts, or whether he communicates what is really on his mind rather than what he has been prompted to say.
By the tests, ABA is superior. Long-term outcomes, in my opinion, will tell a different story. ABA may make you perfect for an institutional environment--it sure teaches you to do what you're told!--but does it do anything to increase your ability to adapt, to learn, to reach your potential? I don't think so.