by: Michael Pryor, President, Click-Air
I was sitting in the right seat of a Cessna 172 recently, gazing into the distance while my student worked through his pre-taxi checklist, and I noticed that a beautiful white Cessna 210 on short-final had not extended its landing gear. We were on a ground control frequency, so I keyed the microphone and transmitted an urgent call.
"AIRCRAFT ON FINAL HAS ITS GEAR UP!" The aircraft descended below my line of sight behind a row of low T-hangers. I watched to see if the message made it from the ground controller to the lead controller. Moments later I saw the 210 climbing out and dropping his gear and heard, "Thank you," from the ground controller (I didn't get a 'thank you' from the 210 but he owes me.)
"Feedback"
Every action has its result. If you fly an airplane at the ground without the wheels down, you will have certain expensive results. The feedback to you will begin with the sound of metal grinding along the pavement. One of my jobs as a flight instructor is to shape a behavior called "putting your landing gear down." I must make sure that my students learn to put their wheels down without ever having seen first-hand what happens if you don't. So I buffer their experience.
A beginning flight student is, at least in the cockpit environment, essentially, autistic: unable to communicate-at least on the radios; unaware of many things; and, ultimately, incapable of keeping himself alive for any length of time. So I gradually guide him through series of a progressive tasks, letting him experience the results of his actions, making sure that the reinforcement (both from me and from success in the chosen task) is frequent enough. Too little reinforcement and they say goodbye to flight training and buy a jet ski. Too much reinforcement is a problem, too. How much is too much? A good question. It seems to me that students can only deal with a certain amount of stimuli, even good stimuli, before they go bleary-eyed. And negative reinforcement must be avoided altogether if possible (more on that later).
Natural Feedback as Reinforcement
At a certain point in the training, it becomes important to let the flight student have the full landing experience, good or bad, right up to the point of being unsafe. So, I fix faulty landings for students-grab the yoke or step on a rudder-until they learn enough to be past the point of hurting the plane. Then I leave them alone to give it their best try and let the aircraft teach them.
The aircraft gives immediate feedback for every action you take. Without prejudice it will obediently reflect your every input. Good landings are very reinforcing. Poor landings just bounce us around or scrape a little rubber off the tires; embarrassing for the student but not fatal. (A student's perfect landing, incidentally, is reinforcing for the instructor. It actually may have been accidental, but it makes me, the teacher, feel that I am accomplishing something!)
The Skill-Cycle
A noticeable increase in skill and confidence seems to come after the instructor finally gets out of the airplane and starts letting the student fly solo. The student is now wholly responsible for every action taken, or not taken, and experiences all the natural feedback available (an infinite amount, really) without the instructor's input. A pilot's actions and the resulting feedback from those actions, along with all the other information available in the cockpit environment, form what I call the skill-cycle. My goal as an instructor is to guide the student along until the feedback from his actions, and all the natural feedback available, is all the information he needs to survive ( like raising a baby isn't it?).
'Clever' Students
Some students require very little effort to teach. They seem to remember everything presented to them. They come to lessons prepared. For me, they are generally ego builders (translation: they learned to land after a only a few tries.) The challenge with them is to keep up with their rate of skill acquisition!
Along the way these students quickly learn to tune in to my cues. For instance, I use hand cues to speed up radio calls at first, pointing at the things that need to be mentioned in each call. This helps to eliminate the beginner's typical awkward pause between each phrase. "Uh, Tower? Uh, this is Cessna 243 approaching uhh, requesting....uh..."
Eventually, I fade my hand cues, and the professional treatment by Air Traffic Control becomes the real reward for making a good call.
Now I have realized that I unconsciously also cue other actions which I am hoping the student will start to accomplish. For instance, if, during a landing, they pass the point at which they should be reducing power to descend and land, I begin to twitch and to look back at the runway disappearing. The student translates this cue to mean, "I should reduce power and then the instructor will say 'Good.'" But what have I taught? To wait until the instructor twitches!
Clever Hans
This dynamic reminds me of the horse named Hans who, apparently, could do math. His handler would give him a math problem and Hans would tap his hoof for the correct number of times. Everyone was amazed. But Han's cue to stop hoofing was the human reactions, such as the sound of everyone sucking in their breath as he hoofed the right number. Clever, eh?
So I have to discipline myself to sit still, and offer no cues for that particular thing, throughout numerous landings perhaps, until the student does it from his own initiative. Then I can reinforce the higher behavior that I am looking for, in this case the decision-making behavior, selecting that response as being more desirable than just the ability to reduce power obediently.
The Personal Challenge
Therein lies the difficulty in shaping behaviors through positive reinforcement. I must be able to let numerous less-than-optimal actions slide by, resisting the natural tendency to bark or shame ("So...are you going to stay up here all day?) The key is in understanding the subtle power of reinforcing the right stuff at exactly the right time. Reinforcement can and should be intermittent. Choosing escalating levels of the 'right stuff' to reinforce guarantees that the reinforcement doesn't become boringly predictable. More importantly, the reinforcement's proximity to the behavior offered is critical. It seems that after the first second, the value of a reinforcer as information drops off considerably. Late reinforcers are at the root of novice trainer frustrations; the animal isn't sure what it did right. What happens has to happen in the plane and during the event; like the airplane itself, my feedback has to be immediate.
Climb, Descend, FURBISH!
While I know that I can become impatient, and as frustrated at times as any flight instructor, I make sure to reward the actions that I want to see, however miniscule they might seem ( a note to flight instructors: when you see the "ball" centered: praise!) And if my student also is gracious about his instructor's shortcomings, I reward that!
But ultimately, somewhere along the way, the instructor must fade away, sign off all the student's papers, and be out on the ramp with the next dreamer, saying "This is the yoke, this is the throttle and those are the rudder pedals."
I just need to remember that what they are hearing is:
"This is the kredspez, this is the blemhob and those are the dortskad pedals".
Post new comment