[plt-scheme] Extending

On 8/23/07, Shriram Krishnamurthi <sk at cs.brown.edu> wrote:
> I presume you mean "instrument".

It's been a long day. :)

> What were your findings re. BlueJ?

It was difficult to draw good statistical results from the data
collected regarding their compilation behavior because the data was
partial---opportunistic, if you will---and only collected when
students were using public PCs on campus. This meant in-lab behavior
and any programming they did outside of class, on-campus. From what I
could see, I  had a complete picture of some students programming
throughout the term, and for others, very little. It isn't an easy
observation to carry out, and it is rich/complex data to work with.

That said, there were clear qualitative differences. At the least, I
believe that visualization of these behaviors can have diagnostic
value in formative contexts in the classroom, and possible other
applications. As soon as I'm a bit more settled (i.e. I have moved
into my apartment, etc.), I'll be diving into some of that data again.

I think I might have an opportunity this term to pilot the capture of
some of the same kind of data in Scheme. I suspect, from some things I
saw at IUB, that there will be some similarities of behavior, but I'd
rather have some real data in my hands to start assessing that.

Cheers,
M