Brain Science Is Not Education Science

The Metropolitan section of the New York Times last Sunday (April 15, 2012) presented on its first page an intriguing headline: “Making Education Brain Science”. The article, by Jenny Anderson, described a Manhattan private school, the Blue School, which is said to have been founded with the idea of incorporating scientific research about child development into the classroom, and is now “a kind of national laboratory for integrating cognitive neuroscience and cutting-edge educational theory into curriculum”.

What does this mean in practice? Anderson mentions teaching young children about the role of the amygdala in out-of-control behavior, and building in periods of reflection that are thought to facilitate executive functions.  Instruction about ways to  describe emotional experiences is also part of the curriculum. One kindergarten child, while “drawing his emotions”, described a battle between happy and angry feelings and said, “The happy fights angry, but angry gets blocked by the force field and can’t get out”; he also explained that happiness can escape through the mouth, but anger is blocked and turns into sadness--   presumably statements based on the instruction he had received and on the opinions of the school’s adviser, Daniel Siegel, co-author of The whole-brain child. Siegel advises against multi-tasking on the grounds that concentration is better at making synaptic connections, among other things.

But brain science, while a fascinating and worthwhile study in itself, is not directly relevant to educational methods. It may offer hints as to ways learning takes place, but those hints need to be tested empirically before we claim that they facilitate learning. Education science is a study of its own and carries out—or should carry out—randomized controlled trials of pedagogical methods, giving evidence about outcomes of different educational techniques.

Brain science, as it exists today, is in fact barely relevant to education science. To jump from one to the other is the error of critical thinking called the non sequitur; effective educational methods do not necessarily follow the suggestions we can derive from neuroscience. Believing that we can go straight from the first to the second is an example of what Howick, in his 2011 book The philosophy of evidence-based medicine, called the “pathophysiologic rationale”--   the belief that we understand how a process works, and what things can go wrong with it, so we can automatically decide how to solve the problem, without recourse to tedious outcome studies. This erroneous approach has been applied to educational decisions with increasing frequency since the 1997 White House Conference on Early Childhood Development and Learning: What New Research on the Brain Tells Us About Our Youngest Children. Although leaders like Kathy Hirsch-Pasek have argued strongly against the assumption that neuroscience automatically tells how education should proceed, the neuroscience fad continues to be embraced by the education community.

Presumably those who tell young children about the role of the amygdala also make suggestions about self-regulation or even guide young children to practice useful techniques (like counting to ten before speaking). If we see those children improving in self-regulation and therefore in educational success, to what do we attribute the improvement? Is it knowledge about the amygdala? Is it adult guidance toward techniques of self-control? Is it simply maturation?  Only genuine education science in the form of careful outcome studies can answer these questions. (However, I’m willing to bet that knowing the word “amygdala” has nothing to do with it.)

What, by the way, is happening with the kindergarten child who believes that happiness can get out of your mouth but anger gets blocked by a “force field”? Did he think this up for himself, or has the Blue School curriculum informed him that he needs catharsis of negative feelings--  a concept still accepted among psychoanalysts, but rejected by empirical work? If this is being taught, the Blue School has veered far from its stated dependence on neuroscience. The assumption that brain science defines pedagogy is bad enough, but to present outmoded psychological concepts as “education science” is a real mistake.

The Blue School seems to have important goals for children, including a shaping of their emotional lives toward positive achievement and an acknowledgment of the social aspects of learning as outlined over the years by Vygotsky, Barbara Rogoff, and many others. Genuine education science would test the importance of those goals and the effectiveness of the methods used to reach them. But you can’t create an education science by waving flags with the word “brain” embroidered on them.

Treating Autism: Some Better Options

Not long ago I expressed concern about the use of “holding therapy” for autistic children and referenced a blog that had discussed this and come to a more positive conclusion. When supporters of “holding therapy” or other complementary-and-alternative treatments hear of criticisms like mine, they often respond by asking, “Well, what would you do, then? What do you think are effective treatments, if you think this one is ineffective?”

Fortunately, I have some answers to this, in part because of the presentations by Dana Lombardi and Steve Glazier at the April meeting of the Delaware Valley Group of the World Association for Infant Mental Health. Dana discussed Pivotal Response Treatment (PRT) and Steve talked about a treatment developed by the late Stanley Greenspan, Developmental, Individual-difference, Relationship-based therapy (DIR, or DIR/Floortime). Although these treatments come out of different theoretical backgrounds, and although they have differing levels of evidentiary support, they have certain things in common--- particularly their focus on relationships and on the satisfaction of successful communication with other people.

Pivotal Response Treatment (PRT) was derived from Applied Behavior Analysis, the well-known treatment for autism that involves operant conditioning techniques and reinforcement for desirable behavior. Robert and Lynn Koegel of the University of California at Santa Barbara developed PRT as a way to enlist each child’s special motivations and pursue treatment in naturalistic rather than lab-based ways. This means that rather than choosing a behavior to work toward without reference to a given child’s interests, the PRT therapist looks for things the child likes to do and works around those. Then, rather than reinforcing behavior with a generic reinforce like an M & M, the therapist makes the reinforcement relate to the child’s interest and activities. For example, if the child likes to play with balls, the reinforcer for trying to say “ball” is to be given the ball, not to be given food. Some important goals for children in PRT are that they initiate activities more often and that they respond more quickly to events. These and similar changes are expected to make the child’s emotional life more positive. Therapists try to keep children’s interest and motivation high by using different toys or other items and changing tasks within a single activity. They are concerned with fostering goal directed communications--  not just saying a word, but saying a word that will get another person to do something you want. PRT therapists also work to help children manage themselves, by helping them learn to do things like relax their bodies. All of these efforts are intended to work toward giving autistic children the skills to interact with other people, to use language for play and learning, and thus to have some constructive alternatives to frustration tantrums.

PRT is a highly plausible approach whose principles agree with long-accepted evidence about the nature of learning. Its emphasis on pleasurable involvement of parent and child in the treatment is also congruent with much that has been shown about early development.

The PRT model has been supported by research. For example, a paper by N. Nefdt, R. Koegel, G.Singer, and M. Gerber (The use of a self-directed learning program to provide introductory training in Pivotal Response Treatment to parents of children with autism. Journal of Positive Interventions, 12, 23-32) reported a randomized controlled trial in which a treatment group of parents was helped to learn principles of PRT, and the parents and their children’s later behavior were assessed. The treatment group was compared to a control group of families who were on a waiting list for the treatment. PRT was followed by significant increases in language use, as compared to the wait-list group. The advantage of the randomized approach, of course, was that it allowed the researchers to account for any improvement that might have occurred in the natural course of development, without treatment, and enabled them to focus on changes that could be attributed to PRT itself.

The DIR/Floortime model is derived from a combination of psychoanalytic principles and some of the suggestions about cognitive development made by the Swiss theorist Jean Piaget. Three ideas (which I am about to take from Steve Glazier’s handout) are foundational for DIR. One is that “language, cognition (including math and quantity concepts), as well as emotional and social skills are all learned through interactive relationships which involve affective exchanges”. This principle implies that treatment of autistic children should not emphasize specific skills, but that pleasurable and playful social exchanges foster the kinds of development that will later make those skills possible. A second idea behind DIR is that there are individual differences in motor and sensory processing that may make it easier or harder for children to participate in social interaction, and that identifying influential differences can help identify the path to communication with a given child. Third, DIR is based on the idea of a predictable series of developmental stages, each based on earlier experiences; working with a given child requires knowing where he or she is developmentally--   for example, whether the child is able to self-regulate well enough to take an interest in the world without being overwhelmed by it.

Floortime, the primary method of treatment used by DIR practitioners, involves having an adult engage in play with the child for 20 to 30 minutes, on the floor, using toys and objects. The adult’s job is to watch what the child seems to do (even if it’s just looking at an object) and to follow the child’s lead--   perhaps by looking at the same object and then reaching out to touch it gently. Back-and-forth communications are repeated, with the goal of having longer and longer chains of communications, with more and more verbalization and emotional expression. (This is described in more detail at www.icdl.org). Following the child’s lead is a way of engaging with the child’s own motivations, and parents who learn these techniques may be able to escape the “learned helplessness” that has developed in the course of frustrating interactions with their autistic children.

DIR is highly plausible in its emphasis on using the child’s motivation to shape the treatment. It can be considered as using principles of reinforcement, too, as the child’s actions lead to predicable consequences (I use this word to mean “outcomes”, not punishment!) and therefore contain the pleasure of mastery over the world.

Some small pilot studies using randomized controlled trials have supported DIR. For example, a 2011 paper by K. Pajareya and K. Nopmaneejumruslers, “A pilot randomized controlled trial of DIR/Floortime parent training intervention for pre-school children with autistic spectrum disorders” (Autism,15, 583) concluded that training parents to use Floortime resulted in significant improvements on several scales (although language does not seem to have been a focus here). In addition, D.M. Casenhiser, S.G. Shanker, & J. Stieben reported in “Learning through interaction in children with autism: Preliminary data from a social-communication-based intervention” (Autism, 2011) that in a randomized controlled trial DIR helped language development more in children with higher IQs.

Both Floortime/DIR and PRT depend on the training of parents and the parents’ ability to incorporate treatment into the natural family setting. ( Curiously, proponents of holding therapy often complain that in conventional treatments like these, the therapists “blame” parents and interfere between parent and child. ) They also encourage engagement of the parent with the child’s interests and wishes and in this way are opposed to the authoritarian bias of holding therapy and related unconventional treatments. Floortime and PRT are plausible, evidence-based treatments for autism; holding therapy is neither of these. Need I say more?

Misunderstanding Autism: Mistakes Have Consequences

The web site http://newautismtherapy.com is interesting--  in a disturbing sort of way! The site offers a discounted (financially, I mean) self-designed home therapy that seems to be based on a variety of sensory stimulation techniques. This is an idea very common among occupational therapists, that you can design a “sensory diet” that will counteract developmental or other difficulties.  The concept goes back to Condillac at the time of the French Revolution and was picked up by other historical figures like Itard, the man who worked with the “Wild Boy of Aveyron”, as well as more recent people. Of course, it’s never been shown to be effective by systematic investigation, but it’s got plenty of history behind it--  nothing “new” about it..

There are other problems associated with newautismtherapy. One is a link to a blog that shows some remarkable assumptions leading to remarkable misunderstandings. At http://newautismtherapy.com/autismblog/holding-therapy/, the writer, Tony Rowe, refers to having had a “love-hate” relationship with holding therapy, but now having tried it as an adult and thinking the outcome was good. He quotes the European holding therapist Jirina Prekop as stating that the treatment, which involves physical restraint that may be fought hard by the subject, to the point of blood and bruises, is “To hold firmly with love someone who is in deep soul crises, until he or she has cried out all the rage and sorrow and feels freer and more contented”.

In quoting Prekop, and elsewhere, Tony Rowe shows a commitment to a belief about emotion that dates back a long time, was assumed to be true by Sigmund Freud and the ethologists, but is probably incorrect. This is the idea that unexpressed emotion acts hydraulically; it does not dissipate, but builds up pressure that eventually will wreak havoc. To get rid of the pressure, according to this viewpoint, it’s necessary to cause the emotion to be fully expressed, and the expression may be far outside the context of whatever was the original cause of the feeling. “Vent” the emotion, and the person will be cured of the problem, with symptoms of the internal pressure also disappearing. Holding therapy is based on that set of beliefs. Of course, there is a little problem--  that does not appear to be the way emotion works. On the contrary, expressing emotion may intensify it rather than dissipate it. Tell off someone who has offended you, and you’re likely to feel even angrier, as if your own action has proven to you how hateful that person really is. No systematic investigation has supported the belief in catharsis, or “drainage” of negative emotions. And nobody even thinks that acting very happy would impair your experience of happiness, which it should do if expression of emotion could dissipate feelings.

Rowe’s blog runs into further trouble with http://newautismtherapy.com/autismblog/what-causes-autism/.  This post suggests that autism is caused by some experience that triggers a “fight, flight, fright or freeze” reaction, resulting in an overstimulation or shutdown of some senses (see first para, above). Decades ago, people like the former art historian Bruno Bettelheim pushed the idea of a traumatic experience as the cause of autism. However, no systematic investigation of autism has revealed that this is the case, unless you include evidence about prenatal exposure to disease as showing psychological trauma.  On the contrary, as you can read in the New York Times this morning (4/5/12), systematic work is showing genetic factors, including mutations, as likely causes of autism. Because thousands of different genetic events could be responsible for interference with development of communication and social interactions, this won’t be understood instantly, but the process of understanding is under way.

Rowe’s belief about the cause of autism leads rather obviously to the suggestion of holding therapy and sensory stimulation as a treatment. However, that kind of thinking is what is called the “pathophysiologic rationale” (see Howick, J. [2011], The philosophy of evidence-based medicine: A philosophical inquiry. Blackwell-Wiley). This rationale involves thinking you know what the mechanism of a problem is, and deducing a treatment that “should” work on that basis--  but believing you can skip the work that would show the treatment is effective. Unfortunately, that’s the step that can’t be skipped, and it’s wrong factually and morally to imply that it can. Making that mistake causes families in need of help to avoid conventional treatments in favor of alternatives that should not be expected to be helpful.

When you add into this mix the newautismtherapy suggestion that you can design and implement your own autism treatment--- and that you have to buy the directions rather than finding them in a published book or journal article--  the result appears to be the makings of a good old-fashioned jug of snake oil. Don’t take it internally!

Causes, Effects, Statistics, and SIDS; or, Pop in a Pacifier?

Many readers will have come across the idea that the reported increase in autism has as much to do with changed diagnostic categories and changed reporting procedures as it does with any actual changes in this serious mental health disorder. Autism is not the only problem for which changes in definitions and reporting have had an effect on nationwide statistics. Until Kempe and his colleagues formulated their descriptions of child abuse in the early 1960s, little information was available, and injuries that we today would think begged for investigation were recorded as accidental. Do the same issues apply to Sudden Infant Death Syndrome? This is more than an abstract question, because the answer is needed to help us understand what child care methods should be recommended to new parents. (N.B. I do not claim that the answer will be found below; I’m just planning to discuss some factors that are relevant.) Because there are probably multiple causes for SIDS, it’s important to have an accurate measure that will let us know whether a specific factor actually makes a difference.

As parents and many non-parents will know, the official position of the U.S. government is that SIDS rates have been and will continue to be reduced when parents follow the advice given by the Back to Sleep program (www.nichd.nih/gov/sids/). This parent education program concentrates on baby sleeping position as the primary factor in SIDS, with the prone (tummy down) position rejected in favor of the supine (on back) sleeping position. Beginning in the 1990s, this program was diametrically opposed to the earlier conventional wisdom, which was that babies should lie in the prone position because that would make it easier for them to clear their airways if they should spit up or vomit. It was a hard sell to parents, because they were inclined to do what they were accustomed to and what was advised by their own parents and other experienced caregivers. Nevertheless, more and more (although not all) of parents, doctors, and nurses have bought into the Back to Sleep program.

Proponents of Back to Sleep have pointed out that the SIDS rate has dropped since the program began, and this is certainly true.  However, it is also true that the rate had been on the way down before the program began, and its trajectory has not changed much. It is also true that ascertaining that an infant death was caused by SIDS is not a very simple matter. There are alternative categories for deaths of this type. They may be classed as “accidental suffocation and strangulation in bed” or as “cause/unknown/unspecified”. (“Murder” is another possibility, but that’s a different story.)  As SIDS rates declined around the beginning of the present century, the rates of deaths in those other categories actually increased (see Shapiro-Mendoza, Tomashek, Anderson, & Wingo, American Journal of Epidemiology, 2006, Vol. 173, pp.  762-769).

More research in this area has yielded information about the complexities of SIDS and underlined the fact that there is no single factor at work here. For example, Li et al (American Journal of Epidemiology, 2003, pp. 446-455) interviewed a sample of mothers whose infants had died suddenly and unexpectedly, and found that one factor appeared to be being put to sleep in an unaccustomed position. Babies who had been used to sleeping in the supine position were more likely to die when placed in prone than those who were accustomed to lie in the prone position, suggesting that parents need to be in agreement with each other about sleeping position, as well as in agreement with child care staff or grandparents who may decide to use the prone position that they find more familiar.

Further research by the same authors (Li et al., British Medical Journal, 2006, issue 7532, pp. 18-21) brought up a new and potentially important factor in SIDS causation and prevention. This was the use of pacifiers (or “dummies”) --  a practice much hated and condemned by many parents and amateur advisers. Li et al interviewed mothers of babies who had died from SIDS and compared them to mothers of healthy babies who were matched on measures like age and socioeconomic status. They found that although the risk of SIDS was significantly greater for babies who slept in the prone position and did not use pacifiers, there was no difference in risk of SIDS for the pacifier-using babies, no matter which position they slept in.

In spite of the concerns raised about changes in the SIDS rate, and in spite of the demonstrated effect of factors other than position, we continue to have a public health emphasis on the supine position as the best intervention for SIDS prevention. But isn’t it time that other factors were emphasized as well? How about a Pop in a Pacifier program--  or would that be too shocking for people who hate to see the pacifier habit begin?  How about a Talk to Your Child Care Helpers program--  highly relevant, when so many babies in the U.S. have multiple caregivers? It’s not even certain that sleeping position is the major factor in SIDS, so why put all the public health eggs in this one basket?