[Edu-sig] Significant drop in CS interest in high schools

[kirby urner]

On Sun, Aug 30, 2009 at 8:54 AM, Edward Cherlin<echerlin at gmail.com> wrote:

<< SNIP >>

> I am working on how to teach CS ideas in third grade using tools such
> as Etoys Smalltalk, UCBLogo, and Turtle Art, all of which are packaged
> in Sugar for the OLPC XO and other Linuces. Etoys and UCBLogo are
> available for numerous platforms, and Turtle Art is written in Python,
> making it easy to port. We already have more than 40 years experience
> teaching programming in elementary schools with Logo and Smalltalk.
>

I'm glad you say "CS ideas" and not "CS" as I think schools are wide
open to innovative curriculum writing they're able to somehow shoehorn
into the pipelines they've already built.  Schools have a math
pipeline, have for a long time, so lets just feed our CS content
through that and branch to departments later, in college, where CS is
its own department.

That being said, of course this or that elite academy might have CS as
an elective.  I'm not for wagging my finger at these folks, as if I
don't appreciate a quality Quaker school when I see one (Haverford is
into Python).

It's just that the politics in Oregon (and so probably other places)
are such that it's easier to enroll math teachers in the idea of more
merit pay in exchange for skills-building around phasing OO into
everyday math, than it is to railroad CS as any kind of "mandatory"
subject in high school.

Resistance to CS as "required" (like math is, at least three years of
it) comes from "programming" being perceived as just one more
profession, and therefore getting too big a footprint i.e. not even
doctors or lawyers get their own track.

"Math", on the other hand (or "maths" as some say) is supposed to
cover all numeracy skills, even alpha-numeracy skills, relevant to
propagating a culture, i.e. is sufficiently an "umbrella term" to
permit such as "digital math" (aka discrete, concrete) wherein
"executable math notations" (e.g. Mathematica, Python, J...) get used
(i.e. we move beyond calculators and beyond flatland (purely planar
geometry) in one fell swoop).

So if we're stuffing the math stocking with CS goodies, how do we
integrate with what's already there?

I've suggested some of the most obvious bridges:

(a) the idea of "a function" is common to both, even though New Math
(SMSG) spun "function" in contrast to "relation".

That's actually a thread about "side effects" if you think about it,
i.e. the same inputs should always generate the same outputs unless
you've got some randomizer as a global, hitting from the side as it
were, but then why not pass that explicitly --- anyway, talk about
"wild cards" fits here.

(b) the idea of "math objects" with Polyhedra the unifying bridge
between "OO talk" (about objects in general, turtles in particular)
and "things with color, shape, other attributes" (i.e. objects like
furniture, machines, bodies...).

When it comes to (a), I have all the usual abstract algebra for
anchoring games with sets of elements and operations (groups, rings
and fields -- easy to get to when the math is concrete).  The only
change with OO is the operations are "inside" the numbers, not
"imposed from without".

When it comes to (b), I follow the approach in 'The Book of Numbers'
by Conway and Guy, leverage such tools as On-Line Encyclopedia of
Integer Sequences (e.g. 1, 12, 42, 92...).

In giving just and (a) and a (b) here, I'm not presuming we don't have
a (c), (d) and so on, plus I could say a lot more about both (a) and
(b), but most of that's already published and helping other gnu math
teachers hone their game.

I've found very limited receptivity to these innovations outside of
some Portland-based pilots we're working on, but then it's not my job
to boss math teachers.  They'll figure out what to do in some
naturally selective adaptive process.

I've found greater receptivity overseas, and, given my international
school background, am looking forward to field testing these
innovations in these more cosmopolitan settings.  We have a teacher
training program going, with Free School offerings for students
willing to be involved in these pilots.  I've been blogging more of
the details.  Linus Pauling House is a hub (isepp.org -- the
organization I represented at Pycon this year, where I delivered a 3
hour workshop with Holdenweb.com).

However, in saying "international school" I'm not presuming I need to
hop a jet to help with these pilots.  We have a lot of international
families working at Intel around here.

Many see the logic in what we're trying to do and want to encourage
some level of success, if for no other reason than we're helping their
bottom line (lots of need for computers) -- but then there *are* other
reasons.

Kirby