Paul Bernhardt wrote:
>
> I appear to have an unusually small undergraduate stats class this next
> semester.
>
> Query for the stats teachers out there: if you had a small class, about
> 7 to 10 students, are there any activities or approaches that you would
> love to do but haven't done because class size makes it unwieldy.
> Suggestions are desired since I have this opportunity.
If you don't normally have access to a one-computer-per-person
classroom, you can get a good approximation with a class this size using
a couple laptops. (I used *one* computer with five students - with 7 to
10 you'd need two or three.)
Make sure there's a seat for you to join them!
This allows - as well as a technology-intensive approach to standard
procedures - heavy use of simulation to settle basic questions of
robustness, sample size, etc. that would not be possible at an
introductory level otherwise.
With 7 to 10 students you could, for instance, get one student to study
the behaviour of the t test for uniformly distributed data, another to
examine another distribution, and so on - not forgetting the Cauchy
distribution!
Also: With 10 students you should be able to be sure enough of whether
students are doing their own work to put much less stress on the final
examination, or even to drop it or make it optional.
-Robert Dawson
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
