Re: [Phys-l] Inquiry

[Brian Whatcott <betwys1@sbcglobal.net>]

What an excellent note!
It cuts to the heart of the teacher's position.
If one is unsure which of two methods is more effective, then testing 
differential results seems like an excellent way to getting some clue as 
to the merits.
That's what physicians and surgeons are forced to do, apparently.

Focusing on that word "testing" gives me the first idea. One needs some 
objective test, and the one that seems to have some scientific support 
is this one that tests for gain.   So the idea of a random selection is 
not so good as a matched groups approach.

So I start by selecting two groups with results showing a similar mean 
and deviation in this "pre-" test. I might consider explaining to the 
students in each group about this plan (though this is sure to color 
results with their perceptions of the superior method.)

Then after a test period (a month? a semester? a year?) I offer the post 
test
and look for a significant difference between the groups.

It's not double-blind, it's not single-blind, but it's a scientific 
approach (more than 80% of your colleagues can claim, would you think?)

Brian W

Philip Keller wrote:
> So let's say that I begin with the belief that my AP physics students would be better prepared for college if they had some experience reading their texts and figuring things out for themselves outside of class.  That is a conjecture.  And now I want to do some research to confirm or refute this.  But I have some questions and concerns about how to do this.
>
> I could randomly divide my students into two groups.  That's about as far as I get without trouble.  Because now I want one group to read on their own, the other not.  But I am still the teacher, responsible for presenting the curriculum.  Do I not teach the non-reading other group this material?  I have to teach it to them if I plan on testing both groups.  I can't test one group and not the other -- kids talk to each other, and so do their parents.  I think I'd get a few emails about this.
>
> Well, let's say I figure a way past this. For a few sections in the text, I find a way to get half of these students to work through it on their own, while I teach it to the other half.  Now how do I test my original conjecture?  It seems that I will have to follow their college careers, measure their success in learning science in college and then tease out the contribution of my experiment.  I don't know how to do this.
>
> I  don't thing it would help to give the FCI/FMCE before and after to both groups.  My conjecture was not about those tests.  I don't know if there is research that shows that performance on those tests correlates with success at learning physics and engineering in college and beyond.  I suppose their must be.  Otherwise, it's just conjecture too.
>