[Beowulf] about clusters in high schools

[Robert G. Brown]

On Fri, 27 Jan 2006, Jim Lux wrote:

> Entire state of CA: 35,135 lowest, 56,444 avg, 53,804 BA+60
> Even districts with very low average family incomes don't have salaries that 
> are a lot lower.  (The low income has other effects.. more is spent on 
> supplemental programs for remediation).

Ummm, it isn't really fair to use CA as an example of what is "typical".
My brother owns a trailer on a 1 acre lot in Santa Rosa, for example,
that is a ramshackle piece of aged out crap that wouldn't be terribly
welcome at any trailer community in NC.  Its cash value is maybe $1
million (obviously, mostly for the lot itself).  I know people in CA who
have sold houses there and come to NC and bought three houses with the
proceeds, literally.  If my house in NC were in CA in a similar
community near a similar size city, it would be worth several million
dollars (which I assure it it is NOT in NC:-).

Super-inflated property values add a certain baseline minimum to the
cost of living there, even with rent control and the like (as in my
brother cannot charge anything like the $7K/month or so that would be
"fair market" rent for a ramshackle trailer).  The CA economy in the
highly desirable living/affluent zones (like e.g.  the New York City
economy) "works", but it is somewhat distorted compared to 90% of the
rest of the country.

Apparently NC's average salary is $43K, compared to $47K national
average (available various places online).  However, entrance level is
much lower and while it isn't by any means an absolute rule, "computer
instructors" are a lot more likely to be young than old, and I'll wager
that only a very small percentage of them have actual Bachelor's degrees
in computer science.  Compare this to:

   http://www.computerworld.com/departments/surveys/skills

and you can see that a "help desk operator" makes more, on average, than
a high school teacher in NC.  All the "real" computer jobs in NC make
maybe $15-20K (on average) more a year than teaching high school, and
let people start making this kind of money far younger than they would
in teaching school.

So OK, you got me, 4-5x was a bit hyperbolic.  The main point, though,
is not -- there is and will likely continue to be a large and immediate
marginal benefit to anyone who leaves a job teaching computer skills in
high school and enters almost ANY sort of real IT position in industry
or (even) higher academe (which pays less than industry but employs a
lot of IT people, making the average a bit deceptive).  For young people
just starting out in teaching the marginal difference of $10-20K can be
a significant fraction of their original salary, and its MARGINAL
benefit (factoring in insurance and retirement, taxes, social security,
and the fixed costs to the family of mortgage and food) can be the
difference between not so genteel poverty and some degree of comfort or
the money required to start a family.  I know several people who used to
be teachers but who literally couldn't afford to stay in the profession,
at least if they wanted to remain married.

Consequently the people who remain in teaching seem to be a mix of the
idealistic and dedicated and the personality disordered and borderline
incompetent.  Fortunately the majority are in the former category but
there are more than enough of the latter individuals to make life
miserable for parents, students, and school administrators alike (same
as any other profession, I suppose, but the impact is greater in schools
where it is often not too easy to fire them).

In many high schools there are very dedicated teachers of computing who
would LIKE to improve their skills, they would LIKE to teach more
challenging courses but they are simply trapped.  They don't HAVE the
skills, there are relatively few ways for them to get them at any cost,
and where there ARE programs there isn't anybody willing to pay the
cost. Even CC teaching programs are fairly few and far between and they
have their own staffing problems.

The interesting thing is that I think that everybody -- governments,
school boards, parents, students, teachers -- understands that computing
is one of the most important single things for an individual to master
in school these days.  Computers are also expensive little suckers, they
become obsolete in 3 years and decrepitly obsolete in 5 years, and the
BEST INTENTIONED TEACHERS IN THE WORLD lag the mastery of the technology
by more like 10.  The kids being taught, who have grown UP with
computers, often understand them better than most of the teachers in
their school, sometimes including the ones nominally responsible for
teaching them about computing.  This creates a huge resource problem.

You can't solve a school's computer problem by a single huge bolus of
money.  To do this properly, one needs to FUND computing with ROLLOVER
REPLACEMENT of every computer the school owns, including those provided
for all the teachers, every five years at the latest.  One has to FUND
systems management for an entire LAN.  One has to FUND software.  One
has to FUND teachers to teach computing, in the teeth of a 30% salary
differential available for the asking to anyone competent and willing to
work for IBM, or SAS, or Pfizer, or GE, or countless smaller companies.
All as recurring costs, line items in the budget, with a budgeting
process that is highly resistant to change and that often still
(somehow) views computers as glorified typewriters, things to buy once
and use until it physically starts to come apart.

Duke has taken YEARS to come to terms with this, even with a lot of
people who fully understood the problem (both sides of it), and is still
reeling from the expense of it all.  Secondary schools often don't even
clearly comprehend the economic problem involved, or do comprehend it
but are checkmated by the availability of economic resources, or get by
on the basis of constantly seeking grants or donations to buy band-aids
every X years to stop the worst of the bleed.  Grants or donations from
non-public sources often come with strings attached of their own...

>>  don't have the LEVEL of
>> experience in networking and supercomputing to be able to support a
>> beowulf program meaningfully.  That's what the Wake Tech thing showed
>> me.  Teachers from schools would need e.g. community colleges LIKE WTCC
>> with programs where they could teach the teachers, before there will
>> ever BE any teachers that can teach this as a course or part of a
>> meaningful high school experience with some continuity.
>
> This is true in many areas.  My mother (a retired teacher,principal,assistant 
> supt) is involved in a program to improve math education at all levels, but 
> especially elementary, and it's a lot of work. We find this peculiarly 
> ironic, since she was a sociology major in school, and now probably has an 
> Erdos number of 3, since she's been working with a bunch of mathemeticians. 
> Collective bargaining for teachers has had a lot of unintended side effects, 
> and one of them is that it makes it hard to do things not directly tied to 
> the curriculum and core objectives.  Union shop stewards and politically 
> active members (a small, but vocal percentage) take a dim view of workers 
> doing more hours than contractually obligated, especially if they are 
> uncompensated, which means that "enrichment" kinds of things (training how to 
> build beowulves, for instance) have to come out of paid time (in some form), 
> and there's precious little extra available.
>
>> The long term solution to the problem, perhaps, is to do what Doug and I
>> and many others have been working on for years -- create a sufficiently
>> robust and strong set of web based resources, including the PEOPLE (many
>> of whom are on this list) who are able and willing to act as mentors, as
>> teachers of teachers, as supporters of CC programs to formally train
>> teachers -- that one can bootstrap the process, where any bright student
>> CAN build a beowulf at a school, where a faculty person working with
>> them CAN learn about linux and supercomputing in the process, where
>> there IS a chance that a program can be born out of the experience.
>
> But.. that just makes it less expensive overall, but doesn't address the core 
> problem, which is that you need to have an "enduring program" of HPC (or 
> robotics, or couture, or auto repair), and to do that, you need to have MONEY 
> on an enduring basis at the district level.  Doesn't take a lot of money, but 
> it has to be there.  Very, very few schools will do away with a program that 
> has any money associated with it, even for what seems to be de-minimis 
> amounts, especially if it has a multi-year commitment.  That money could pay 
> for a few hundred dollars as a stipend for a teacher (who would then be 
> compensated, however trivially, for their extra work, getting out from the 
> union problem).
>
> So, get that endowment fund going.

Interesting solution!  Of course this produces a sort of awkward
plutocracy as far as setting curricula is concerned, where someone or
some corporation (ahem, ahem) with lots of money (ahem) and an agenda
(ahem) can endow a fund with strings attached and arrange to shape those
young minds along the paths they select.  Or they can achieve the same
goal by donating equipment or software instead of money and avoid having
to pay the taxes that might have ultimately funded the schools directly
(but without the influence).  Beyond the corporate, often as not
individual initiatives along these lines tend to push still more
resources into schools in affluent communities while neglecting poor
ones just because the poor ones don't have the wealthy donors, because
parents have a hard enough getting involved where their own kids' future
and opportunities are at stake, let alone kids in poor rural communities
Down East (in NC, that would be the eastern part of the state, where
there are whole counties that are very poor indeed).

Consider: the Gates foundation is currently dumping huge amounts of
money into building "small high schools" as a solution to the various
ills of the secondary schooling system in the US.  To its credit, much
of the money IS going to poor communities, this is not an example of the
rich getting richer, it is doubtless a noble and philanthropic
initiative.  However, it is also an excellent example of the distortion
of public policy according to the whim of one very wealthy foundation.
>From what I can discover, there is little data to support the assertion
that "small high schools are better than large ones" by any widely
accepted metric.  Indeed, it has been pointed out -- quietly, since who
wants to count the teeth of a gift horse? -- that smaller high schools
are often relatively resource-starved, as they often lack facilities
found in larger ones or cannot achieve class sizes that justify the
maintenance of specialized programs.

This is easy to understand in terms of basic economics -- it is a lot
easier to (say) not build specialized facilities for some group of
students (say, advanced ones, or musically inclined ones, or athletic
ones) if there are only five or ten students at a small school that
might qualify instead of twenty or thirty at a larger one.  There is an
economy of scale that makes it worthwhile to operate many programs only
when a school reaches a large enough size, and one might THINK that time
and the "natural selection" of spending scarce money mostly wisely have
conspired to create schools in the sweet spot of this multidimensional
optimization problem.  One might THINK that communities have the right
to determine what school size is best for their own community's needs.
Not according to the Gates foundation.

Of course this might be wrong.  Maybe small schools ARE better in some
nonlinear way overall, or maybe forcing many states into the model of
supporting smaller schools by virtue of giving them money to build the
initial physical facilities will cause something good to emerge.  Either
way, it is an excellent example of a very, very small group of people
with lots and lots of money and their own particular vision (right or
wrong) creating national-scale public policy out of whole cloth with
nothing like a democratic process involved.  When you've got $29 billion
dollars in foundation endowment money, who needs silly things like a
school board or cost benefit analysis?  They'll take whatever we deign
to give them and be grateful for it, and if it is wrong nobody will
figure it out for twenty or thirty years anyway, and by then we'll all
be dead. (Yes, this is probably overly cynical, but funding "small high
schools" as some sort of predetermined optimum instead of funding
"optimally sized high schools" smacks of high-handed control and
preconceived solutions instead of the exercise of human judgement, or
maybe of the desire to fund a LOT of small schools instead of a lot
FEWER larger ones, as if the number of schools itself was the metric
being optimized).

It is left as an exercise for the studio audience to meditate upon the
loss of freedom and danger to democracy that this sort of process
represents.  It is the primary reason that I knee-jerk oppose
Microsoft's IT monopoly, whose pernicious political and social influence
is matched only by entire industrial groups such as "drug companies" or
"oil companies", not by any particular company in those industrial
groups. Not only does concentrating that much money in the hands of a
very few, very rich individuals create a huge distortion of e.g. the
political landscape, but the fact that the underlying business is the
management of INFORMATION creates whole new levels of threat to our
personal freedoms.

Schools are ALREADY routinely "bribed" into de facto becoming factories
for the next generation of dutiful proles by virtue of e.g. donating a
whole pile of carefully branded hardware, by donating otherwise
overwhelmingly expensive software (or more often, adjusting the price to
what in a sane non-monopolistic universe would be fair market value).
Having been there with my hands out on behalf of MY kids local school, I
promise that schools are usually pitifully eager to be so bribed -- it's
often the only way they can afford to DO anything because IT is just
plain expensive.

Alternatively, one can push at the national or state level to have a
particular curriculum emphasized as a matter of national or state
policy.  One can get involved with overhauling the school system
politically, trying to get public high schools (which actually spend
MORE MONEY per student than all but the MOST expensive private schools
in the community charge for tuition, at least around here) to run
efficiently and still offer a challenging curriculum.  This is MUCH more
difficult and creates resource problems of its own, but in the long run
is both more economical and more democratic.

Democracy and economics ultimately both favor the penguin...  I think.

    rgb

-- 
Robert G. Brown	                       http://www.phy.duke.edu/~rgb/
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb at phy.duke.edu