Teaching, no grades

[From Bill Powers (961010.1715 MDT)]

>By collecting the right data within a well-modeled
>learning process, you shouldn't need a test that is
>detectably separate from the learning process itself.
>Instead, a student would have a continually (measured in
>seconds, or less, not weeks, months, or years) updated file
>that includes estimates of the student's ability to control
>any given subset of a full constellation/network of perceptions
>relevant to being educated.

Wow. That sounds like the exact answer to the grading problem.

It's interesting that learning effects are a problem in psychological
research. If you allow the subject to learn during a test, you aren't
measuring the knowledge or skill with which the subject came into the test.
So what do you do? Try to disguise the object of the test, not giving away
the answers or dropping hints from which the subject might learn something.
In other words, you consciously avoid doing the things that might aid
learning. Not a great habit for a teacher to get into.

I had a physics professor who gave his classes quizzes at least once a week.
He explained that they would be graded, but that the grades would not be
recorded or used for any purpose except by the students themselves. They
were, in effect, checklists. If you got 100% on a test, it meant you had
learned everything important he had wanted to teach that week. If you got a
lower grade, he expected you to go back and figure out what you missed, so
you were always confident that you knew the stuff up to that point. A very
stress-free course. I'll bet that guy would have liked PCT.

Best,

Bill P.

[Martin Taylor 961015 12:10]

Bill Powers (961010.1715 MDT)

I had a physics professor who gave his classes quizzes at least once a week.
He explained that they would be graded, but that the grades would not be
recorded or used for any purpose except by the students themselves. They
were, in effect, checklists. If you got 100% on a test, it meant you had
learned everything important he had wanted to teach that week.

I had a metallurgy professor who used the exact opposite approach, with
the same result. He gave out the final exam questions at the start of the
course, and said that everyone would get an A regardless of what they wrote.
(Unfortunately after a month or so he said that the authorities had told
him he couldn't give all A's, and he would have to grade the exams. But
the questions would stand). There were three questions, of which two had
to be selected for answer. I remember one: "Discuss the metallurgy of steel."
They all required the subject to be understood if the answer was to make sense.
I found that course to be the best and deepest in my whole four years
of engineering physics. And for many years after that, I _did_ understand
the metallurgy of steel (all forgotten now, unfortunately).

But you and I are both talking about a time of life when the student has
had enough experience to be able to judge what is likely to be useful to
learn, and under those circumstances, to control for grades instead of for
understanding could lead to conflict (just as could controlling simultaneously
for matching cursor and target position _and_ velocity, or for keeping the
car in the middle of the lane and keeping the steering wheel centred).

Martin