Saturday, November 27, 2010

On non-verbal verbal programs. and nusing simulators

Continuing the discussion,  let's look at the ultimate non-verbal learning skill -- language skills.

There are two competing models here of how humans operate, and they have very different implications. This is the MBA / Zen training argument from the last 2 posts carried to extremes.

One model (or myth) is that humans are really just sort of computing machines.  You put "facts" and "rules" into the mental hopper, and, if they don't fall out again,  you're done.  Stuff has been learned, people can pass tests,  success!

I've watched this model fail repeatedly with foreign language instruction.   Well meaning educators teach vocabulary and grammar rules to students,  who stuff it into their short-term mental hoppers,  and can regurgitate it on exams.   Here's the funny thing though -- if you take one of those students of, say, French, and put them in Paris, you'll find that they are unable to communicate in French. 

Why is that?  Well, it turns out that the human brain CANNOT do on-the-fly algebra and compute, using rules of grammar,  what needs to be said in the time available to say it, even if the student "knows" the rules of grammar 100%.  
 
This kind of "knowing" -- having stuff stored in stacks and heaps in the memorization hopper -- does not automatically equate to or turn into ability to USE the information.

In computer terms -- we cannot use knowledge in real-time by running what's called an "interpreter" program on it. It simply takes too long.  Human brains are pretty lousy at algebra, and even if they do it well, they do it very slowly.   Like real world computers that need speed, we can't operate from raw "code" or "programs" , but need instead to compile the rules into a totally different pre-processed, pre-loaded shape for "the program" to actually be of any real-time value.

In other words, we need practice, and very specific practice -- we need to develop context-sensitive recall of information and preferably entire mindsets and phrases or sentences that work in that situation, and maybe, if we're writing a paper,  AFTER we recall it, we can use the rules of grammar as a post-hoc check on whether what we recalled rapidly (without computation) agrees with what we get when we re-compute what it "should be."

That, of course, is the whole basis for a totally different approach to language acquisition, namely the Berlitz method,    which manages to teach people to speak the same way children learn -- amazingly, totally devoid of any conscious or explicit awareness of rules of grammar.  

So we have a paradox in our school systems, in that first our children learn to speak English, completely without the aid of any rules of grammar at all,  and then, later,  with considerable effort and sometimes marginal success,   we attempt to teach them "rules of grammar."

Frankly, I would rate as a total failure any school that successfully taught my children "rules of grammar" and "vocabulary" and produced as output children that were unable to speak.   I would rate as a success any school that resulted in my children able to speak fluently in a language, even if they had no idea what "rules of grammar" were implicit in what they were saying, and didn't know a gerund from a noun from a pluperfect subjunctive.

We all learn language DESPITE the rules, not by using the rules.  We all know that it is correct to say "a big, red ball" and it is just plain WRONG to say "a red, big ball",  even though we'd be hard pressed to say WHY one makes sense and the other is just eerily wrong.

What I hear the Japanese saying regarding an MBA education is that stuffing facts and theories and logical models and other "MBA curriculum" type stuff into students brains does not equate to making them "better managers".    Even if facts and rules are "correct" (and some are questionable),  there is essentially no benefit to having the rules made explicit and articulated in words.

Any company would prefer to have a manager who did the right thing, but couldn't explain WHY it was the right thing, to a manager who could tell you later what the right thing would have been to do, but didn't do it and didn't realize that was the "applicable rule" when the situation went by.

The same is true of nursing education.  I'd rather have a nurse who did the right thing, but couldn't explain why, than one who did the wrong thing, but later could tell you exactly why it was wrong, based on "Bennigton's Model of the Seven Gotchas", or God knows what.

There is an implicit assumption here that, if only the schools can cram symbolic knowledge into the brain's hopper and confirm it can be replayed on a test,  that "learning" has taken place.    Of course,  without "practice" they do know perfectly well that if you put the nurse into a practical situation and the patient goes into cardiac arrest,   odds are essentially zero that this symbolic form of knowledge will come to mind, spring into action and guide what the nurse does next. 

Yet, this always seems to come as a surprise.

So, now comes onto the stage "virtual reality simulators" where it is possible for the nursing students to be put into a realistic scenario,  where certain behaviors would be correct, and let them ATTEMPT to compute, in real time, which of the zillion facts they learned in school might be relevant, while the "patient" expires.   After they have failed in that manner,   we can all go in the other room and "debrief",  ie,   with lots of time and no pressure we can have the luxury of computing ("recalling") what the RIGHT thing would HAVE BEEN to do in that situation.      According to normal practice in using simulation in nursing,   at this point we are done.

Astounding.   The nursing student has practiced slow recall, and discovered a fact (we could have told them in advance) that their symbolic knowledge store doesn't actually work in practice.   Then,  we can post-compute what should have occurred there, and the student can IMAGINE what it WOULD HAVE BEEN LIKE if they had done the correct thing.

Here's what's crazy -- in pilot training, the next thing to occur would be to REPEAT that very same simulated drill,   in which case maybe 50% of the students would "get it right", despite having just gone over it in debrief.    Then we REPEAT AGAIN the very same drill, and maybe 90% of the students "get it right".    Then we repeat again and again, until 100% of the students have "mastered" the situation->response linkage.     Then, we wait a few days and do this again, and discover that many of these linkages have evaporated, and performance is back to 70% correct.  So we drill again until it gets back to 100%.   And repeat after another few days.  Etc.   After a long enough time, with sufficient repetitions at intervals,   the linkage will become permanent and the reflex will become automatic, a new reflex.  

But be clear about two things.   One,  the reflex occurs first, and the mental confirmation ("grammar") that it is correct occurs afterwards, not vice versa.   This is a different kind of "learning."
Second,  this kind of learning comes from actually DOING the desired response when faced with the triggering situation,  which requires MULTIPLE sessions of EXACTLY THE SAME SCENARIO.

So, a student pilot, for example,  will take up a plane and spend hours trying, over and over, to execute a left 90 degree turn without losing altitude,  or a particular type of landing.

Because real-life simulators in nursing schools are so expensive,   there is neither time nor inclination for the students to do the "same drill" twice, and it is NOT done.  It's done once, there is a "debrief" which is still reported by the students as "very valuable" in terms of "learning",  and they stop there.

I have to conclude that nursing schools are misunderstanding the role of simulation in pilot training, if they think they are doing "the same thing."

Furthermore, I think it is just critical that VIRTUAL REALITY simulators be available to nursing students so that they CAN afford to go back to that situation and PRACTICE "doing it correctly", with only themselves or other students checking that they in fact did finally "get it right".

In fact, the order should be reversed.  FIRST students should watch a video of what it is they are about to learn.    THEN, students should watch themselves DOING The correct thing in virtual reality, where their behavior has been fully "scripted" and all they have to do is go along for the ride.    THEN the students should practice doing the right thing and scoring themselves with some sort of checklist after watching the instant replay of their own actions.   THEN they should put this situation into their list of "quiz me" flash-card type file, which the simulator now has permission to throw at them unexpectedly from time to time.   And THEN, when they are scoring 95 to 100% of the time correctly,  they should go do this on the REAL manikins in the REAL "simulation lab" --the one that costs $1 million to put in place and is only open M-F 9-5.

Frankly,  the virtual practice is probably worth more in that sequence than the "real simulator" confirmation that they got it right.  If you had to pick only one or the other, I'd go with the virtual simulator repeated self-assessed drill than a one-pass-and-debrief with a "real"simulator.

No comments: