Re: Using the Normal Distribution to Grade on a Curve


bm459@xxxxxxx wrote:
Reef Fish wrote:

bm459@xxxxxxx wrote:
DarkProtoman wrote:

David Winsemius wrote:
"DarkProtoman" <Protoman2050@xxxxxxxxx> wrote in
news:1155586101.575352.159710@xxxxxxxxxxxxxxxxxxxxxxxxxxxx:

DarkProtoman should print a copy of this cost and give it to his
chairman, and invite him to give his response. :-)


RF

First of all I agree that a test where a student scores 100 or very
close is not a particularly good test unless you happen to have a lot
of validation behind it. In this case we have an assistant prof who
can not have such validation unless he is using some standardized test
developed by someone else. In that case the grade scores should have
been given with the test questions for various levels of achievment.

As I recall the courses I took in both undergrad and grad school where
I really learned a lot had tests where the class high score might be a
60 or at most a 70. Of course the class low score might be a 14. That
makes it easy to measure achievment.

In the case of the grades given you have two choices. Particularly as
the results are not anything like a normal curve. One, as I stated, is
you have a bunch who are learning nothing and are not going to be
prepared for higher level courses, assuming that this course has
content of use in higher courses and is not simply a filler like PE or
how to use the library. The other is you have a class with a couple of
over achievers who are getting abnormally high scores. There simply is
no way to tell which is which from the data given.

If you can not tell which case you are dealing with there is no fair
way to judge anyones performance on the tests. If you can not judge
performance you can not assign grades fairly.

Objecting to grades does not solve anyones problem. Pass fail is still
a grading system. So is giving everyone a pass for paying fees. In
either case you are simply removing any ability for understanding
achievment. But I will be the first to agree that for some courses
pass for everyone is an adequate measure as the course is meaningless
to future life.

I also realize that many tests can be gamed. I recall a philosophy
course I took as an undergrad for example. After review of the first
test of the quarter it was clear that gaming the test and not doing any
of the assigned reading was the way to get a grade in the class. So
for the second test I read about half of the boring assigned reading.
I got a high B on that test so knew I had the right idea. For the rest
of the quarter I did not open a book and aced the final with ease. I
learned just about zero in the course. But it filled a boring liberal
arts requirement and got me credits so I could graduate. The time I
saved I drank extra beer and studied extra in the real courses. Not
entirely a waste I would say. That was not the only course I did that
with. Even think there are some valuable life leasons in there. Like
spend your time doing well at what really counts and do what you must
to get by on stuff that does not count.

We would likely like each other if we were neighbors RF. You sound
damned near as mean as I am.

What does "gaming the test" mean? I've never heard of that term before.
I want to know if my undergrads are doing it.

.

Relevant Pages

  • Re: Using the Normal Distribution to Grade on a Curve
    ... been given with the test questions for various levels of achievment. ... I really learned a lot had tests where the class high score might be a ... In the case of the grades given you have two choices. ... for the second test I read about half of the boring assigned reading. ...
    (sci.stat.math)
  • Re: Using the Normal Distribution to Grade on a Curve
    ... First of all I agree that a test where a student scores 100 or very ... been given with the test questions for various levels of achievment. ... In the case of the grades given you have two choices. ... What does "gaming the test" mean? ...
    (sci.stat.math)
  • Re: While studying CS
    ... The best-paying job that I know that a C.S. undergrad got was a 6 figure ... Note the emphasis on grad level courses. ... Your grades, of course, are important, but not critical. ...
    (comp.programming)
  • Re: Environmentalists
    ... though I made good grades as an undergrad and manage to scrape by with mostly B's in grad school, it remains to be seen if I can put any of this stuff to use yet. ...
    (rec.bicycles.tech)