ABSTRACT. Facilitated communication was regailed as a major breakthrough in 1990 in communicating with people who have autism. The claim was that 90% of children suffering from nonverbal autism could communicate with the aid of a keyboard and a facilitator guiding their hands to make the keystrokes. Scientific examination of the claims made by facilitated communication proponents showed a different story. When facilitated communication was performed under controlled conditions, the majority of studies showed no effect. Those few controlled studies which did show benefit suffered from severe methodological problems. It has been found that the authorship can not be attributed to the autistic subject. Rather, it is the facilitator unconsciously doing the typing via the ideomotor effect. The illusion of subject authorship is very strong, as is made clear by the belief in its efficacy by both its practitioners and parents of autistic children making use of facilitated communication. Coupling this with a parent’s desire to interact with their autistic child makes it even more so. But facilitators have on many occasions accused a parent of sexual abuse through their subject, and at times a child’s testimony via facilitated communication has been accepted as evidence despite its failure to pass the Frye test. Facilitated communication has been discredited in the scientific community, yet it continues to be applied. The harm it causes is clear. It has destroyed lives with false allegations of sexual abuse and may prevent the special needs of autistic and sufferers of similar nonverbal disorders from being met on the basis of what a facilitator says via the subject.
Happy New Year to all! It’s been a while. My wife and I spent New Year’s Eve at a local bar. My head was on a swivel trying to follow two hockey games (the Canada-USA world junior game and the Flames vs. the Oilers traditional NYE game which I would have killed for tickets to see) all the while playing that satellite trivia game NTN. My neck is sore.
I pulled out a bunch of papers I had previously printed on fascilitated communication. I got interested in this after viewing a news clip about an unfortunate man in a vegetative state all-of-a-sudden able to type on a keyboard. Do watch, as it is very interesting. I can only imagine how his parents felt when all of a sudden this unfortunate man could type messages out. But then, he could only do it assisted by a fully conscious person guiding his hand. Later I remembered the case of an autistic child who was found able to express full thoughts on a computer keyboard if not verbally. Of course, this required someone to guide their hand. Hmmmmm.
Facilitated communication began in 1990 with the claim by Douglas Biklen of Syracuse University that alleged that communication was possible with 90% of nonverbal autistic individuals through assisted typing1. The credulous look at fascillitated communication and take it for what it is at face value, like those who promote out-of-body experiences that I took care of in a previous post. I’ve got little patience for credulity. Sure, they may be right, but only by accident. Some of us prefer to ask the question: What does science have to say?
In short, under controlled experimental conditions, fascilitated communication (hereafter referred to as ‘FC’) has unequivically been determined to be unsubstantiated garbage. And controlled experimental conditions are the only way to determine claims like this. (Indeed, I would say that they are the only way to evaluate any claim, and those claims which can not be established via controlled experiments should always be provisionally rejected.)
In a review of the literature by Mostert2 looking at 29 studies of FC between 1995 and 2001, 19 controlled studies refuted FC claims, six controlled studies supported FC claims and the remainder were uncontrolled studies supporting FC claims. The six controlled studies supporting FC claims can be divided into two groups: the first assert unspecific and undefined findings in support of FC (showing no difference between FC and other communication techniques, the assertion of unspecified accuracy at follow-up when no such accuracy appeared in the initial study for instance) and those reporting substantive findings of FC claims.
Purported Controlled Studies in Support of FC – The Case of Cardinal, Hanson and Wakeham
Let’s take a closer look at a study demonstrating a positive effect and why experimental design can be tricky. In a study by Cardinal, Hanson and Wakeham3 the hypothesis to be tested was that rudimentary information can be transmitted to a naive facilitator. Their subjects (numbering 43) exhibited a number of disabling conditions (autism, cerebral palsy and developmental delay) and were described as severely handicapped. Data was taken at a number of different schools in a classroom setting. All subjects had made use of FC prior to the study.
To test the hypothesis in question, single, randomly selected words were shown to each subject by a ‘recorder’ (a teacher who had worked with the subject prior to the study and recored the keystrokes later entered aided by the facilitator) in the absence of the facilitator (which is how the facilitator is made ‘naive’…). The facilitator then entered the room and the subject typed via facilitation and the letters recorded. Using this procedure, 3800 trials were performed. The word set used consisted of 100 nouns and verbs containing no more than six letters and taken from the school curriculum. The facilitator had open access to the word set.
Data from several baseline conditions were obtained. In the first baseline condition, trials were performed as above, but without facilitation (though the facilitator was present). As one would expect, low levels of accuracy were obtained (that is, the keystrokes made bore little resemblence to the word shown to the subject). The facilitated condition was then tested. That is, the word was shown to the subject by the recorder (in the absence of the facilitator) and the word entered on the keyboard aided by the facilitator. In this condition, a high accuracy was obtained. A second baseline condition was then obtained in the manner of the first. Again, a low accuracy was the result.
The conclusion of the study was that some FC can pass accurate information, verifying the hypothesis. But, there’s a problem. Have you spotted it? I spotted it before I read Mostert’s take on it, but then I do experimental design for a living (though in a very different discipline, but experimental science is experimental science). The recorder had access to the word shown to the subject.
This was a silly mistake on the part of the authors. While the facilitator was effectively blinded (good), the recorder was not (really, REALLY bad…). The recorders are in the position where they can strongly influence (even if unconsciously) experimental outcomes. Even worse, the keystrokes were reported to the recorder, rather than observed directly by the recorder. The recorder could easily provide unconscious cues as to the accuracy of the keystroke back to the facilitator, providing a feedback mechanism a la “Clever Hans”.
Nor did I like the fact that the facilitators had access to the word set used. This is just a bizarrely unnecessary bias introduction. Combine the knowledge of the word set with unconscious recorder feedback and what you get is a refining mechanism. That is, the accuracy improves rapidly as the typing is going on in a kind of ‘twenty questions’ fashion.
The other studies showing a positive effect of FC are similarly amateurishly flawed. Remove these biases using proper controls and the result is that no effect is observed. That is, the claims of FC are falsified by the overwhelming number of studies demonstrating no effect.
So how does FC work? In a manner similar to a Ouija board, in fact. Kezuka4 measured the force applied by the facilitator with a strain guage and showed that it was the facilitator which was communicating, not the subject. The process, as it is in the case for the Ouija board, is guided by what is known as the ideomotor action. Penn and Teller very effectively demonstrated how ideomotor action works with the Ouija board and that the sense of an external guidance is purely an illusion on their excellent television show Penn & Teller’s Bullshit!5.
Many might ask, “So what if facilitated communication is falsified? What’s the harm in giving families hope that their autistic child is actually a real person in a bad body?” As Kezuka wrote4, “Imagine that one day a nonverbal autistic child suddenly starts to type messages, such as “I am not retarded,” using a computer keyboard while being touched by an assistant. Facilitated communication (FC) appears to create this miracle around the world.” Indeed.
Facilitated Communication and the Law
Well, as with so many things taken to an extreme, real harm can result even setting aside that false hope can itself be massively damaging. What happens when the family figures out that it’s all an illusion? I couldn’t imagine the despair that such a realization would result in, no matter that the facilitator his/herself truly believed they were doing good.
But real harm has happened on several occasions. While in scientific circles FC has been soundly rejected, in the legal arena testimony obtained via FC has been accepted in cases of alleged child abuse. As per Kezuka, imagine one day after the autistic child wrote “I am not retarded” begins writing ‘Daddy abused me.'” In 1994, four dozen (!) cases of sexual abuse (along with not one case of physical abuse, supicious in itself due to its statistical unlikelihood) were reported6. Departments of Social Services must investigate all alleged cases of sexual abuse, to be sure, but when the only available evidence is from a discredited source, such investigations should end quickly with the conclusion in the negative. Yet this is not the case. A case here in Canada has been tried where the mother of a 12-year-old communicating via FC was accused of murdering her daughter on the basis that the child expressed a wish to die through the facilitator7. The girl was starved to death. If lives destroyed by false allegations of abuse and even death is not harm, I don’t know what is. How such testimony makes it into evidence, despite the failure of FC claims to pass the Frye test (to pass the Frye test, testimony must be via a scientifically accepted method) is beyond me.
People want to believe FC works. As in the video above, even people who should know better can be fooled. And dispite it’s having being discredited, Biklen continues to promote FC and is a member of the Facilitated Communication Institute at Syrcuse University, which offers training in FC.
I have nothing but sympathy for parents who will understandably grasp at anything which will allow them to believe their children are not autistic. Reading testimony from parents making use of FC is heart-wrenching, even more so knowing that they have been duped. We see the same grasping at straws from many anti-vaccination proponents who blame their child’s autism on falsified claims that vaccinations are harmful. As with the Ouija board, the illusion FC gives of a child suffering from a nonverbal disorder such as severe autism can be incredibly powerful. I have no doubt that even the faciltator also believes that their hand is being guided by the subject. And there will be those who will say that I am not being ‘open-minded’. To them I say, “Bullshit!” I have presented what science has said on the matter and rejecting claims on the basis of overwhelming evidence to the contrary is not being ‘closed-minded’. Should studies supporting FC which pass examination of their methodologies turn up, I will re-evaluate my posiiton. That’s how open-mindedness works. It is not maintaining something is possible in the face of a mountain of evidence against it. For those who take this position, that *splat!* is the sound of your brains spilling out.
Reality is reality, and running from it – while easier – is not the way to go. FC may very well keep proper care from being used in those who quite clearly have special needs.
1. Biklen D. Communication Unbound: Autism and praxis. Harv Educ Rev 60:291 (1990)
2. Mostert MP. Facilitated communication since 1995: A review of published studies. J Autism Dev Disord 31:287-313 (2001)
3. Cardinal DN, Hanson D and Wakeham J. Investigation of authorship in facilitated communication. Mental Retardation 34:231-42 (1996)
4. Kezuka E. The role of touch in facilitated communication. J Autism Dev Disord 27:571-93 (1997)
5. Penn & Teller’s Bullshit!, season 1, episode 12 (2003)
6. Margolin KN. How shall facilitated communication be judged? Facilitated communication and the legal system. In Shane H (ed) Facilitated Communication: The clinical and social phenomenon. pp. 227-58, Singular Press, San Diego, CA (1994)
7. Konstantareas MM. Allegations of sexual abuse by nonverbal autistic people via facilitated communication: testing of validity. Child Abuse Neglect 22:1027-41 (1998)