May 14, 2010

A Bridge Too Far

Over at Bridging Differences, Deb Meiers writes:

That was an amazing and surprising find re. Milwaukee charters. I thought that at the very least they'd get the advantage of being in a more diverse (integrated) setting with more middle-class kids and that being chosen (even by lottery) would produce a kind of halo effect. Why it didn't is what should baffle the media. But it doesn't.
I comment:

Or perhaps, your implicit assumption that diverse (integrated) settings with more middle class kids confers an educational advantage which leads to improved student performance is invalid.


The assumption rests on shaky empirical support in the first place. So, one would think that this additional piece of potentially-negative evidence might lead an un-biased thinker to question her underlying assumptions. Why it doesn't baffles me.

PS: This might be one of the best examples of irony I've ever seen in an education blog.

PPS: It's also a good example of why we never make any progress in education. Policy thinkers become so wedded to their pet assumptions and will bend over backwards to discount contrary evidence. Classic Confirmation Bias.

3 comments:

Dick Schutz said...

PPPS. A train wreck of two "Classic Confirmation Biases" here. Neither school integration nor vouchers improves student performance. Why anyone would think that either or both would "should baffle the media."

KDeRosa said...

Are you sure about that, Dick?

Disclaimer: I haven't looked at the underlying data.

Dick Schutz said...

Both "integration" and "vouchers" are political/social/economically motivated initiatives. Nothing wrong with that. What's wrong is to think for a minute that they will "improve student performance."

With
--the great variability withing classes, within schools, within, districts, and within states

--the lack of attention to any aspect of the instruction that's actually delivered

--prevailing comparative research methodology in education

--the instructional insensitivity of measures of "student performance

there is no chance under the sun that either of the two can be shown to have any reliable effect.

It really doesn't require any empirical research to arrive at this conclusion. A thought experiment should suffice. It took decades to dispel the integration notion (and some people still didn't get the message). Vouchers (masked as Charters or Scholarships)just haven't produced the successes that are being promoted. Even the poster-school examples do not stand up to close scrutiny.

For all the "gee whiz" Charter Schools one can find equivalent "gee whiz" public schools. But the further away you get from the unique local condition and the unique personnel involved, the less replicable.

The jargon is "they don't scale." Actually, there was nothing "there." That finding has been replicated over and over since the days of Follow Through and most recently in the 13-year "Study of Instructional Improvement:"

http://www.cpre.org/images/stories/cpre_pdfs/sii%20final%20report_web%20file.pdf

we argue that design-based school improvement tends to work best—not when the process encourages local educators to invent instructional and organizational solutions to the practical problems of teaching and learning that they face—but rather when it helps teachers learn how to use a well-specified set of practices through extensive supports

Educational policy makers and policy wannabees are ignoring solid research and chasing bad research. Not smart.