I think these are good points. Probably it is too early to be leaping into a statistical study without some prior practical exploration, and an initial qualitative study is what is needed to develop a hunch into a hypothesis. Student feedback would seem to be a good starting point (and was in fact the seed of the idea in the first place). Discovering the right direction is better than going a mile in the wrong one.
Daniel
On 2/12/06, Matt Jadud <[EMAIL PROTECTED]> wrote:
It seems to me that an A/B study theorizes ahead of any data you might
collect; you're already assuming that life drawing will have an effect,
and that effect will manifest itself as a change in the student's grade
or performance in the course. But if your only measure for its effect is
student performance on the course, then you might be missing all kinds
of interesting things. For example, you might begin by running the class
and having someone come in and interview the students a few times
through the semester. Or, perhaps have someone observe the class, and
couple that with transcriptions of a microphone you wear around the
classroom. Or, perhaps you might encourage the students to keep a
journal, and do the same yourself. Or perhaps, or perhaps... the design
of such a qualitative inquiry does require some thought.
Rich sources of data collected early on might become goldmines for
refining any exploration you want to do later. Also, these kinds of
explorations don't require you to postulate a theory about how life
drawing + programming works, but instead allows you to collect data that
might suggest a plausible starting point for your theorizing.
I guess I'm saying that if you are proposing to change the fundamental
nature of your classroom, an A/B study run over the next 10 years may
just produce crude, statistical noise. On the other hand, a semester of
qualitative work that explores student responses to an instructional
method that will likely challenge them (and you) in significant ways
could be a very potent source of data, and help refine any notions of
(long-running) qualitative studies you'd like to do.
M
Lindsay Marshall wrote:
>
>>I'm also considering the question of how to measure the success of such a
>>trial. One possibility I considered, as the proposed course is in
>>preparation for the Sun Certification exam, is to split the group into two,
>>and have half of them leave the room for an "ordinary" (inactive) break
>>whilst the other half are doing life-drawing; then collect average exam
>>scores for each sub-group. However the trial would have to be repeated many
>>times to get a large enough sample to make this result meaningful. On the
>>plus side - assuming we are limited to 10 for the life-drawing - this
>>approach could double the full class size to 20 students...
>
>
> I agree that you would have to do this often to iron out any variation. (There would also be the issue of avoiding self-selection - non-drawers in the control class, drawers in the life class) The other, perhpas theoretical, problem with this is the ethical issue : you ought/have to make sure you have everyone's signed agreement up front or else you end up with the possibility of one group doing particularly well (for whatever reason) and someone in the other group claiming they were disadvantaged and then sueing. Trainging them for an assessment that is adminstered by someone else would clear a few other issues though.
>
> (And all you ppigers testing things on your first year students are also on a slippery ethical slope...)
>
> L.
>
> ----------------------------------------------------------------------
> PPIG Discuss List (discuss@ppig.org)
> Discuss admin: http://limitlessmail.net/mailman/listinfo/discuss
> Announce admin: http://limitlessmail.net/mailman/listinfo/announce
> PPIG Discuss archive: http://www.mail-archive.com/discuss%40ppig.org/
