Thursday, February 19, 2009

24/7 Speech monitor?

I want to discuss Jerry Maguire's idea of a 24/7 speech monitor to evaluate speech.

We need to objectively measure the change in stuttering due to a treatment in order to evaluate the treatment. For this purpose, we can ask the stutterer himself or people close to him, but that is not objective. It might be useful for dramatic changes. So we need to record them and analyse the speech using an independent person.
But here we have big big big big big problems that lead to apparent success:
  • If you know that you are better recorded, you are on your best behaviour and of course try to use speech techniques learned. 
  • You are recorded once every so many weeks or months, which is 0.001% of all your speaking time. You have good and bad days and fluctuate a lot. 
  • Coming back to a clinical environment makes you more fluent because you have associated fluent speech with this environment. 
  • You might be good in reading but not on phone calls. So you need to be recorded across a variety of situations.
Virtually all outcome studies have these weaknesses, and that is one of the the reasons why so many therapies claim good results, even though you never quite meet all those people and if you meet them, they are not that much cured.

What to do about it?
One possibility is secret recording, e.g. phone them up under a false pretence. I know of a concrete example of this happening after the official recording, and suddenly the treatment was less successful. So whenever you see a treatment result divide it by 2!!!! I am not joking.

Another way is, as Jerry suggested, to give us a speech rate monitor, like a heart rate monitor, which we would carry along 24/7 and which would measure our fluency. Let me analyse his idea.

The Pros:
We cover all speaking situations over a long period. So instead of 0.001% we record all, and we record across a range of different situations. You can argue that it is a secret recording but in a fake sense. They know about it but they tend to forget.

The Cons or Obstacles:
  • You still know that you are better monitored and this will help you be more fluent. You might forget after a while but not for long.
  • Someone needs to sit down 24/7, listen to your beautiful speech, and count all stuttering incidents! Wow. I do not want to be the person who does this. It is unrealistic, so you would need to take random sample, but then you might miss this one big stuttering event. Unlike heart rate monitors, there is no automatic recording and analysis. It needs to be done by hand.
  • Not sure all speaking situations should be recorded. Maybe bedroom and bathroom should be excluded! ;-)
  • Resources intensive: imagine you have a trial of 100 patients and assume it takes one week to analyse 6 months of day per patient for one researcher, so you need 100 weeks to do it.
  • Recording speech is very memory intensive.
My conclusion:
I guess you can do it with the necessary resources and clever use of random samples, but I am convinced you need to analyse by ear unlike heart monitors.


Greg said...

His idea of the 24/7 monitor is really the only way to study stuttering. Every other methodology, including the ones that I publish, are fatally flawed.

The problem w/ the 24/7 monitor idea is that it suffers from the same fatal confounds as everything else:

1. Stuttering cannot be empirically measured.
2. We can't even define what stuttering acts are.
3. We can't even define what the stuttering phenomenon is.

So until we can define the pathology, all attempts at measuring it will be flawed and of questionable validity.

Sorry to rain on anyone's parade, but this is just the way it is...

Greg said...

Oh--sorry, I forgot to sign the above post...


Anonymous said...

greg, what have you published and where do you publish? you described the problem very well, any solutions? i thought the handbook of stuttering by bloodstein (how old is he?)defined stuttering and associated behaviors.

Greg said...

There has yet to be a good definition of the stuttering pathology itself. You should look into it, if you ever have the time. They're all based on the concept that stuttering is a speech behavior, but such a notion is wrong. There's tons of research demonstrating the existence of 'sub-perceptual stuttering'; not only that, the idea that stuttering *must* be a speech disorder is speechist, to say the least. (Ask any well-connected deaf person...)

No, I've not published a ton. I'm young, and I spent tons of time working in deafness which didn't really promote active research. However, I've got 4 currently in press at the moment, so maybe I'll catch up.

In any event, Bloodstein said a lot of things; that doesn't mean that he's right. According to him, we all stutter because we incorrectly diagnosed ourselves as stutters--and if we can just distract ourselves from this diagnosis, then we'd be fluent people.

The behavioralists (or fluency-shapers, speech-motor people) make a similar mistake to that of the psychological perspective. Both of these camps have mistaken the symptoms for stuttering as the cause, and then they try to define the pathology based on (misunderstood) symptoms. So the psychological perspective cites stuttering essentially as a personal character flaw (i.e., anxiety). No, they got it wrong. Anxiety is the result of stuttering, not the cause of it. The speech-motor people defined stuttering as a behavioral disorder; no, the act of stuttered speech is what the body produces to try to "fix" or compensate for stuttering (occurring on the neural level.)

I actually wrote a post about it today on my blog.

Hope this helps; And stop being anonymous.


Anonymous said...

but all your publications are in
what is the impact factor?

it is a little can we trust you and believe what you are saying...

some people choose to be anonymous for "safety reasons"

Anonymous said...

greg, by the way. I just saw that you are the co-editor of the journal. I guess it is easier to publish when you are the editor. right?

Greg said...

Hi Anonymous. Growing tired of your cowardice. No, I don't count any JSTAR articles; that's just what I do in my spare time. It's a public service to help PWS and school-based SLPs.

Check out the upcoming JoCD, PMS, JMSLP and PFD issues. If you knew anything about SLP, you'd know the journal's acronyms.

Bottom line--feel free to think whatever you want; it'll make no difference in reality. If you like the status quo, then be my guest. It's done great things for the stuttering population thus far...

Pamm said...

Oh, I definitely like the robust dialogue going on here.
Differing opinions are the spice of life, gentlemen!

Anonymous said...

Greg -

I am not 1st anonymous,but I have grown tired of the bullshi* research and therapy in the field of stuttering by you and your fellow posers... er... "researchers".

As to cowardice, you seem to throw a lot of bullshi*... er ... critisizing... at dead researchers
on your blog. Don't seem all that brave to me.

Stop your whining... Your arrogance in your ignorance is really annoying...

Anonymous said...


You wrote;
"So the psychological perspective cites stuttering essentially as a personal character flaw (i.e., anxiety). No, they got it wrong. Anxiety is the result of stuttering, not the cause of it."

In my case, anxiety definitely plays a big part in the severity of my stutter. Anxiety of stuttering, and also general anxiety. Here are a couple of examples:

Years ago when I attended my first speech therapy session at a local hospital, I was required sit in front of 3 SLPs and answer questions about myself - presumably to measure the severity of my stutter. I could hardly get a word out and I was a complete mess. The anxiety of being in unfamiliar surroundings with unfamiliar people made my stutter much worse than usual. I went back there one week later; this time the surroundings were more familiar, and the SLPs didn't seem quite so scary (one of them was very sexy), so my stutter was not so severe. Of course the SLPs tried to take credit for my dramatic "incredible improvement".

Another example:
A few years ago, I was going through some very distressing personal issues that were making me anxious. I was losing sleep, and could not think straight. During that period (about 3 months), my speech was terrible. But once the issues were resolved, it was like a great load had been lifted from my shoulders, and I felt much more relaxed. My speech was back to normal ... still stuttering, but not so severely. I've seen this with other stutterers ... one guy was quite amusing to me; I could always tell when he was having trouble with his girlfriend because his stutter became much more severe.

These kinds of situations convinced me that anxiety plays a huge part in stuttering - anxiety makes the amygdala more active, and this may result in an inhibitory effect on my speech centre. So I've tried to simplify my life, do meditation, don't worry so much about the little things, and just become a more mellow and relaxed person. I still stutter, but not so severely.

I acknowledge that not all stutterers have the same experience, and this leads me to believe that stuttering may have various different causes. Even though 2 people may stutter the same way, their pathologies may be very different ... just like 2 white mice might be indistinguishable to even an expert eye, but DNA analysis shows that they are really different species.


Greg said...

Hi George--you're right, anxiety does play a big role in stuttering severity. Never for a second would I suggest otherwise. It's just not the cause--which is what past theories have suggested. (That stuttering is a psychological disorder, related to anxiety.)

I'm a very big believer in desensitization as a foundation in stuttering therapy.

Good luck,

And as for the new anonymous above, yeah--I'm tired of poor research and therapy as well. And, admittedly, mine will suffer from much of the same flaws as all the others, but it's a necessary byproduct of the academic system. Publish to keep your job. In any event, the reason why I spend so much time online, finding and mentoring other PWS is precisely due to this reality: we really don't know what were talking about when it comes to stuttering. (Myself included.)

I'm easy to find; feel free to enlighten me relative to my ignorant arrogance; anyone that knows me knows that I appreciate constructive feedback.