Wednesday, September 29, 2010

Desperately Seeking Logic

Yesterday in Old Bridge Governor Christie laid out the planks of his education reform platform. Here are his six initiatives, although he didn't neglect forget to leaven the proceedings with his customary aplomb, including these bon mots (courtesy of PolitickerNJ):

“Tenure is the sclerosis that coats the veins of our school system.”

“I don’t bash teachers. I bash their stubborn, self-interested union. That’s who I bash.”

“We are paying a king’s fortune for an education system that isn’t giving our children the royal treatment.”

Here’s his proposals; the last two would require legislative approval.
  • Spend $10 million on our data system NJ SMART ( by the way, see today’s NJ Spotlight on its gaping holes and how its incompleteness was a major deficit in our Race To The Top quest) so that we can tie student achievement to teacher effectiveness.
  • Create a Task Force on Teacher Effectiveness to decide how to link teacher pay and student achievement (originally planned as an unwieldy 36 members, now down to 9).
  • Create alternate route programs for principals.
  • Create rank of “master teacher” and “master principal,” with added incentives of higher pay, more professional development, and, potentially, leadership of their own charter schools.
  • Restructure teacher compensation to eliminate seniority-based raises.
  • Base tenure on merit.

The NJEA responded immediately with a press release citing the results of this past summer’s study by the Economic Policy Institute (EPI) called “Problems with Use of Student Test Scores to Evaluate Teachers.” President Barbara Keshishian added,
As Gov. Christie sets out to pursue an education agenda that has significant implications for the future of New Jersey’s entire public education system, New Jersey residents and policymakers would be wise to ask whether it is an agenda based on sound educational practice, or simply another attack on New Jersey’s excellent public schools.
(New Jersey's "excellent public schools?" When will she retire that old bromide? Jeez. Tell that to the kids in Newark and Camden and Trenton and Paterson.)

One salient point: EPI’s study was funded by NEA. Its Board of Directors includes Presidents of the International Association Of Machinists and Allied Workers, Service Employees International Union, Communication Workers of America, United Steelworkers of America, United Auto Workers, American Federation of State, County, and Municipal Employees, Workers United, Service Employees International Union, AFL-CIO, American Federation of Teachers. (On July 30th EPI published another study arguing that NJ’s public employees are underpaid.)

The anti-value-added model evalution cadre – those who oppose using student academic growth to evaluate teacher effectiveness – is strong, boasting such luminaries as Diane Ravitch, Linda Darling-Hammond, and, more locally, Bruce Baker at Rutgers. But the logic is specious. Data-driven teaching evaluations aren’t perfect. Therefore, let’s rely on the current system, which is almost universally acknowledged to be devoid of all meaning.

Remember “The Widget Effect,” the 2009 study from The New Teacher Project, which examined teacher evaluations in four states -- Arkansas, Colorado, Illinois and Ohio – and determined that “all teachers are rated good or great:”
In districts that use binary evaluation ratings (generally “satisfactory” or “unsatisfactory”), more than 99 percent of teachers receive the satisfactory rating. Districts that use a broader range of rating options do little better; in these districts, 94 percent of teachers receive one of the top two ratings and less than 1 percent are rated unsatisfactory.

Excellence goes unrecognized
When all teachers are rated good or great, those who are truly exceptional cannot be formally identified. Fifty-nine percent of teachers and 63 percent of administrators say their district is not doing enough to identify, compensate, promote and retain the most effective teachers.
Our current system of evaluating teachers stinks, say teachers, administrators, education experts, and policy analysts. A proposed alternative – value-added models – is better, but not perfect. Therefore, let’s keep the one that stinks. If logic were an item in a standardized test, opponents of value-added models would fail.


schoolfinance101 said...

It is not specious to point out that VAM is too problematic to use, even if current eval. inadequate.

Your knowledge of the statistics & modeling issues in VAM are sorely lacking:

What is specious is to imply this false dichotomy. current system no good, therefore VA models the answer. Sophomoric logic.

Laurie Goodman said...

I don't think Ravitch, et al are saying VAM is wrong so we must keep what we're currently doing. Those aren't the only two choices. Pretty sure everyone agrees that improvement is mandatory...but the VAM method is too's being promoted as a magic bullet where "the numbers don't lie." Well, actually, sometimes they do.

NJ Left Behind said...

Actually, Bruce, the NJEA leadership's logic is sophomoric. No one's arguing that VAM is perfect. But NJEA is fighting for our badly flawed system that privileges teacher job security over student achievement? Then again, that's a union's job, right?

Not using student growth to gauge teacher effectiveness doesn't work. Relying completely on student growth to gauge teacher effectiveness doesn't work either. The key is to develop a system that incorporates various factors -- including student growth.

schoolfinance101 said...

The problem for me, as a statistician, is that the numbers are so problematic that it's hard to figure any really good way to integrate them into the system. it's not a simple process of identifying the magic share of teacher evaluation that should be tied to the scores. I'm nolonger convinced that it's any sizable share. That doesn't mean that changes to teacher evaluation can't or shouldn't occur. It's still a false dichotomy.

Saying that "not using VAM doesn't work" does not mean that "using VAM does work" or that "VAM can be used reasonably." I'm increasingly skeptical in this regard, having previously taken the technocratic view on this.

And I'm not in this to defend NJEA. And the opposition to this use of VAM doesn't primarily come from NJEA. It comes from the National Research Council, National Council on Measurement in Education, American Educational Research Association and others. Even those informed scholars who favor finding some way to integrate these measures are increasingly cautious in their approach - Nothing close to the current proposal (50% share).

So this isn't some union-backed opposition to what everyone else thinks is right and necessary. It's not just the EPI brief, which happens to be signed by some exceptional scholars of public policy including Helen Ladd. In fact, at this point, as far as I can tell, it's really only political hacks and pundits who seem to be favoring a major role for VAM in teacher effectiveness ratings. Most informed scholars have sought out much more reasonable ground.

kallikak said...

The only 'sure thing' about this proposal is that it will gut existing salary guides and---depending on the height of master teacher and other performance hurdles---effectively cap them at levels well below $100K.

What will the prospective teacher workforce look like under that scenario? Why would anyone want to work in this soon-to-be low-paid trade if your job security is akin to that of a football coach who might be terminated if this year's team falls below 500?

Question for Team Christie: where- oh-where is any hard evidence that a viable student-performance-driven teacher evaluation system actually exists?

kallikak said...

A test we'd like to see as part of the teacher-evaluation debate:

At the start of the school year, swap 6 teachers from Tenafly HS with 6 counterparts from the worst-performing HS in Newark. At the end of the year, rank the teachers using a VAM or other student-performance-driven metric.

Who will make the cut????

NJ Left Behind said...

That’s a highly unlikely proposition. Poor urban students are twice as likely as wealthier to get novice teachers, uncertified teachers, teachers without content knowledge. See this link ( for further discussion.

But it’s really interesting to contemplate what would happen if counties had the ability to assign teachers to schools to improve access and equity for poor students, as you suggest. A good argument for county-wide contracts, or even state-wide ones.

Duke said...

Isn't it possible - I'd even say likely - that a good teacher in Millburn may be a bad teacher in Newark?

Then how would county-wide assignments help?

Joe Schilp said...

Kallikak, regarding your statement, "What will the prospective teacher workforce look like under that scenario? Why would anyone want to work in this soon-to-be low-paid trade if your job security is akin to that of a football coach who might be terminated if this year's team falls below 500?"

Teachers are very well paid and won't be leaving the trade to go work in another state, where the pay is lower. They're also not going to work in private industry, where they know there's no huge pension, no lifetime bennies, no summers off and no 185-day work year. It just ain't happening, especially now that there are no jobs out there.

kallikak said...

Duke gets it: does our perception of 'success' more heavily reflect the context or the actor?

As regards my other comment: rational actors (in this case, prospective teachers) tend to view their career choices in terms of lifetime potential earnings. Christie's proposals would effectively cap later-years' pay and retirement benefits for most teachers while also severely limiting their compensation via a jump to the supervisory track (where comp levels are being capped and compressed as we speak).

Teaching as a profession competes with other lines of work in the minds of prospective job candidates. In NJ, it just became a lot less attractive on a relative basis. Eventually, the current recession/mini-depression will end and we will see if our dog can hunt.