Re: Creating smarter Belgian riders

[Robert Chung]

Ryan Cousineau wrote:

> If I sound like a big fan of charter schools, it's because I am.
> If their only purpose is to self-select for parents who actively
> want their kids to get a better-than-average education, that
> might be enough.

Not if those schools did a poorer job.
http://www.ncspe.org/publications_files/OP111.pdf

 

[Bill C]

Robert Chung wrote:
> Ryan Cousineau wrote:
>
> > If I sound like a big fan of charter schools, it's because I am.
> > If their only purpose is to self-select for parents who actively
> > want their kids to get a better-than-average education, that
> > might be enough.
>
> Not if those schools did a poorer job.
> http://www.ncspe.org/publications_files/OP111.pdf

Thanks for putting that up. Our youngest son is going to be part of
the ongoing math study when he starts at the Performing Arts Charter
school his sister is at. They use a two track approach to teaching
math. One traditional, and one built around applied math and problem
solving. This seems to work well for them, but they are putting out
very few hard science/math types. They put about 90% of their kids in
college and I'd guess that 80% of those are in BA programs not BS
courses. Needless to say that the Verbal/language scores are a lot
higher than the math, but that's because that's what these kids are
into and it's emphasised.
They also have a 2 hour longer school day so they can work in all the
Performance and Arts courses, along with having to work shows and
events outside typical school hours during the year also.
Gives a lot of kids who would do poorly in a normal HS environment a
place to really thrive. If they want to do the Arts stuff they have to
do the academics and it works. They make kids redo work until they
understand it. Not just pass a test but they have to demonstrate
understanding of the lesson to pass. This works because older students
are required to do monitored mentoring and tutoring etc, to graduate,
along with staff being incredibly committed to the kids. Seems to work
really well overall.
Bill C

 

[Robert Chung]

Bill C wrote:

> Gives a lot of kids who would do poorly in a normal HS environment a
> place to really thrive. If they want to do the Arts stuff they have to
> do the academics and it works. They make kids redo work until they
> understand it. Not just pass a test but they have to demonstrate
> understanding of the lesson to pass. This works because older students
> are required to do monitored mentoring and tutoring etc, to graduate,
> along with staff being incredibly committed to the kids. Seems to work
> really well overall.

http://207.70.82.73/ra/special/bonus/grantmakers.ram

 

[Tom Kunich]

Robert, do you REALLY believe that bunk?

 

[Robert Chung]

Tom Kunich wrote:
> Robert, do you REALLY believe that bunk?

I believe that there are several facets to school quality, and that one of
them is to see how much its students improve.

 

[Tom Kunich]

Building a case by claiming that they're "controlling" for unknowns
such as what a student would do somewhere else doesn't strike you as
being just the slightest bit, er, un-discerning?

 

[Robert Chung]

Tom Kunich wrote:
> Building a case by claiming that they're "controlling" for unknowns
> such as what a student would do somewhere else doesn't strike you as
> being just the slightest bit, er, un-discerning?

Oh, that's what you meant. It's certainly not perfect, but we don't
usually get a chance to run natural experiments in real life so
observational controls are often the best we can do. This happens to be
the reason why almost all of the really interesting statistical
innovations over the last couple of decades have come in areas where we
can't run randomized experiments. And, in a deeper sense, trying to figure
out how a student would do somewhere else (or in a different teaching
system) is exactly the question we should be trying to answer before we
pass new policy.

That said, there's actually a pretty interesting natural experiment going
on in charter school evaluation in the Chicago charter schools lottery:
there were too many applicants, so they lotteried the slots. They're
tracking the performance both of the kids who got in and the ones who
didn't to see whether there's a difference in their improvement.

 

[Michael Press]

In article <[email protected]>,
"Robert Chung" <[email protected]> wrote:

> Tom Kunich wrote:
> > Building a case by claiming that they're "controlling" for unknowns
> > such as what a student would do somewhere else doesn't strike you as
> > being just the slightest bit, er, un-discerning?
>
> Oh, that's what you meant. It's certainly not perfect, but we don't
> usually get a chance to run natural experiments in real life so
> observational controls are often the best we can do. This happens to be
> the reason why almost all of the really interesting statistical
> innovations over the last couple of decades have come in areas where we
> can't run randomized experiments. And, in a deeper sense, trying to figure
> out how a student would do somewhere else (or in a different teaching
> system) is exactly the question we should be trying to answer before we
> pass new policy.

We are conducting experiments. And as the subject line
suggests, the experiments are failures. Or perhaps the
experiments successfully tell us that the methods are
failures.

> That said, there's actually a pretty interesting natural experiment going
> on in charter school evaluation in the Chicago charter schools lottery:
> there were too many applicants, so they lotteried the slots. They're
> tracking the performance both of the kids who got in and the ones who
> didn't to see whether there's a difference in their improvement.

A few hours reading, writing, and doing sums is enough for
anyone under the age of twelve; then home to do their
chores. Whoops! Nobody home. I know what to do. State
supported day-care.

--
Michael Press

 

[Robert Chung]

Michael Press wrote:

> We are conducting experiments. And as the subject line
> suggests, the experiments are failures. Or perhaps the
> experiments successfully tell us that the methods are
> failures.

Successful schools are all alike. Every unsuccessful school is
unsuccessful in its own way.