TITLE: Another Way Our Students Are Different From Us
AUTHOR: Eugene Wallingford
DATE: December 18, 2006 4:00 PM
DESC:
-----
BODY:
On my CS 1 final exam, I asked the students to suggest
a way that we could hide a string in a sound, where a
`Sound` object is encoded as an array of sound
samples in the range [-32768..32767]. Here is the last
part of the question:
*
Under your scheme, how many characters could you hide
in a song that is 3:05 long?
*

My caricature of the typical answer is...
*
It depends.
*

Actually, most of the students were a bit more specific:
*
It depends on the sampling rate.
*

Of course it does. I was looking for an answer of this
sort:
*
(185 * sampling rate) / 100
*

... for a scheme that stored one character every 100 sound
samples. It never occurred to me that most students would
get as far as "It depends on the sampling rate." and just
stop. When they realized that they couldn't write down as
answer such as "42", they must have figured they were done
thinking about the problem. I've been doing computer science
so long, and enjoying math for much longer, that I naturally
wanted to write a formula to express the result. I guess I
assumed the students would want to do the same! This is yet
another example of
how our students are different from us.
Fortunately, nearly all of them came up with a scheme for
hiding a text that would work. Some of their schemes would
degrade the sound we hear considerably, but that wasn't the
point of the question. My goal was to see whether they
could think about our representations at that level. In
this, they succeeded well enough.
Well, I guess that depends.
-----