On 12 Jan 2000 11:38:06 -0800, [EMAIL PROTECTED] (EAKIN MARK E)
wrote:
>
> As an example in class, I was considering using the letter grade
> distribution in class as an example of an approximate multinomial
> distribution. How close do you feel this approximation is?
- If you look at one set of numbers, it might look multinomial, hey,
it is "numbers" into "categories." But when I think about it, I
can't think of how I would define a Population that one particular
class would be a random sampling *of*, and still follow multinomial.
It seems to me that the teachers try to define the approximate counts
to fall into categories, and those categories are ordered; the
observed results are more consistent (I bet) in the letter-grade
AVERAGE that is produced, than you would get from multinomial
sampling.
How close is the approximation? -- Well, for what purpose? For most
comparisons, it would probably be more fruitful to look at the
Averages.
--
Rich Ulrich, [EMAIL PROTECTED]
http://www.pitt.edu/~wpilib/index.html
