the Washington consensus & teacher quality

Now that the Supreme Court has ruled against race-based student assignment plans (within parameters set by Justice Kennedy, I gather), I wish to heck we could get our policy wonks & columnists to stop writing race-based columns and policy briefs.

The enforced focus on race leads directly to wrong assumptions (white schools: good; black schools: bad) and, I'm convinced, to increased racism (black parents/character/culture: bad; white parents/character/culture: good). That's not the intention, obviously; policy wonks and education columnists are trying to narrow or close the achievement gap, not ratify it.

But all too often the continual association of "badness" with "blackness" and "Hispanicness" does the opposite.

It's not that policy wonks and columnists should be color-blind. What's good about NCLB - one of the things - is that it forces affluent school districts like mine to disaggregate their data, to tell taxpayers & parents how well their disadvantaged kids are doing.

The problem with making race the focus is that you factor out curriculum, pedagogy, teacher training, and even education case law and accountability as essential components of analysis. Once you do that, you limit your ability to understand what's going on inside our schools.

I don't understand what's going on inside our schools, either, but I've come to feel that there's a great deal more - or less - than the Washington consensus suspects or even imagines.


the consensus:

Reformers in the nation’s capital agreed on ....the nature of the education problem....How did the NCLB advocates understand the problem they intended to solve? First and foremost, they were concerned about the nation’s “achievement gap”--primarily the disparity between the performance of white and Asian students on the one hand, and African-American and Latino students on the other. In 2000, the typical African-American 12th grader was reading and performing math at approximately the same level as the average white 8th grader. Leaders of both parties declared this to be unacceptable, a violation of equal opportunity, and a threat to America’s future competitiveness (Thernstrom and Thernstrom, 2003).

I'm convinced that "the nature of the education problem" isn't the achievement gap.

The achievement gap is a scandal; it needs to be narrowed and/or closed and/or reversed, depending upon the child.

But the achievement gap isn't the problem. It's the symptom of the problem, or so I will believe until sound value-added analysis shows me I'm wrong.

The essential reason for the 4-year achievement gap, or one of the essential reasons, is that middle and upper-middle class white children are more "bulletproof" than disadvantaged children for various reasons, no doubt including parents who can "assist with homework," (scroll down for CA Math Frameworks) parents who can hire tutors, parents who build word and world knowledge in their children as a matter of routine, black and Hispanic children fearful of acting white, and Lord knows what all.

In light of all this, my guess is that paying "effective teachers" lots more money to teach in urban schools won't narrow the gap appreciably, though I assume it would help.

This is an example of the "perception gap" between people with kids in the schools and analysts without kids (currently) in the schools. Pundits and policy analysts read the new report on teacher equity from Tennessee (pdf file; eduwonk link) as proof that teacher equity is a major problem and, thus, a major solution, if we could just take all the good teachers away from the white kids and give them to the black kids.

Sometimes you have to wonder: are analysts thinking about the real world at all? If urban parents didn't like busing, just how receptive are urban parents of any race or ethnicity going to be to a top-down scheme to strip their children's schools of the best teachers?

While the TN study does show that low poverty/low minority schools have better teachers than high poverty/high minority schools, I suspect the figures are far less dramatic than members of the general public would expect:

most effective teachers:

High poverty/high minority schools: 17.6% of teaching staff is "most effective"
Low poverty/low minority schools: 21.3% of teaching staff is "most effective"

18% versus 21%

It's possible that a 3-point difference is much larger than it appears "on paper," of course. And the report does say that the "most effective" teachers in high poverty schools aren't as effective as the "most effective" teachers in low poverty schools.

But this passage makes me wonder:

A teacher effect score below zero indicates that the average student in the teacher’s class made less growth than the statewide average, while a teacher effect score above zero indicates that the average student in the teacher’s class made more growth than the statewide average.

[snip]

If a teacher’s effect score was below zero, and one standard error above the score was still negative, the teacher was categorized as “least effective.”

If a teacher’s effect score was positive, and one standard error below the score was still positive, the teacher was categorized as “most effective”.

As usual, I wish I'd been able to take a statistics course by now..... but lacking the requisite background knowledge, my question is: if we're comparing "most effective" teachers in high poverty schools to "most effective" teachers in low poverty schools on the criteria of their students beating the statewide average ---- aren't we potentially saying that the "most effective" teachers in the high poverty schools are actually better than the "most effective" teachers in the low poverty schools?

Wouldn't it be more of an achievement to have your high-poverty students beating the statewide average than to have your low-poverty students beating the statewide average?

Or have I got the logic completely wrong? [see Doug's answer, below]

The figures on "least effective" teachers are also much closer than we non-pundits would assume:

High poverty/high minority schools: 16% of the teaching staff is "least effective"
Low poverty/low minority schools: 23.8% of the teaching staff is "least effective"

I suspect most middle and upper-middle class parents would be horrified to learn that students in classes taught by 16% of the teachers in their schools are ending up performing more than 1 standard deviation below the statewide average.

That 16% figure, btw, jibes with the figure Ed was always cited when he headed the California History Social Science Project. Administrators and teachers universally told him that 15% of teachers were terrible.

Here's another interesting passage:

Although many of the beginning teachers in high poverty/high minority schools are among the state’s most effective, many of them do not stay in these schools or lose their effectiveness over time.

There are more "most effective" beginning teachers in high poverty/high minority schools (roughly 17%) than in low poverty/low minority schools (roughly 12%).

The big disparity happens down the line, in the number of expert teachers (using the novice, professional, expert classification) teaching in high poverty/high minority versus low poverty/low minority schools.

When you look at teachers teaching from 11 to 20 years (and remember: expertise takes 10 years to develop, research on teacher effectiveness notwithstanding) the gap is immense:

High poverty/high minority schools: 16% of teachers teaching 11-20 years in high poverty/high minority schools are "most effective"
Low poverty/low minority schools: 27% of teachers teaching 11-20 years in low poverty/low minority schools are "most effective"

So.....looking at this, and I realize it's dangerous to make comparisons across time.... what I see happening is that high poverty schools start out with 17.6% of their novice teachers being "most effective"; 10 years later that figure stands at 16%.

Low poverty/low minority schools in TN start out with 12% "most effective" novices; 10 years later that figure is up to 27%.



trouble in River City

This is bad news, because the baby boom teachers are retiring and being replaced by novices.

I've been reliably told that 10 years ago my district didn't hire novice teachers; we didn't even interview novice teachers.

Now all of our hires have no more than 5 years teaching experience as far as I've been able to determine. (This may not be the case for SPED teachers.)

People tell me this is happening all over the country. Given the pension situation in many communities, I assume that's the case.

Low poverty/low minority schools are systematically lowering teacher quality through hiring practices.

If the Tennessee figures hold true for other schools, my own school district may be looking at a reduction of a teacher quality from 27% of teachers being "most effective" to as low as 12%.


update from Doug:

"Or have I got the logic completely wrong?"

Not completely. 8-)

First, by the description, the students skills aren't being compared to the average, their progress is being compared to the average. On the face of things, this seems reasonably fair. A good teacher should be able to create more progress in his students than a poor teacher. This seems to be a measure of that.

Second, the dividing line is not teachers whose progressed more than a standard deviation below the average, but students whose progress was more than one standard error below the average. With a large population of teachers, this would seem to be a very small difference.

In fact, this would result in any teacher whose classes learned detectably better than average being rated as "most effective". Talk about your low expectations!

What this says to me is that the difference between the top 20% of teachers and the mean is barely detectable. In other words, this would imply that there is little benefit to most teachers, since more than 60% of teachers are statistically indistinguishable.

Perhaps the top 10% (or 5%, or 1%) is actually capable of making a significant difference in learning, but that can't be determined from what you report.

I find that both surprising and discouraging. All I can hope is that either I made a mistake in analysis or the survey has some problem.