Bridging Differences returns from summer hiatus today. Deborah Meier and Diane Ravitch’s blog resumes with a Ravitch post that takes issue with the $5 billion “Race to the Top” fund. “As I predicted on this blog, President Obama and Education Secretary Arne Duncan are now the spear carriers for the GOP’s education policies of choice and accountability,” Ravitch writes.
What is extraordinary about these regulations is that they have no credible basis in research. They just happen to be the programs and approaches favored by the people in power. Under normal circumstances, the Department of Education would need congressional hearings and authorization to launch a program so sweeping and so sharply defined. Instead, they are using the “stimulus” money to impose their preferences, with no hearings and no congressional authorization.
Ravitch is also troubled by the administration’s push to tie teacher evaluations to student test scores.
I commend to our readers the response to the RTTT regulations by Professor Helen Ladd, an economist who has studied teacher evaluation for many years, as well as the one by Paul Barton, who has studied education issues for many years. What both of these responses clearly demonstrate is that there is no research basis for the priorities favored by Secretary Duncan.
“This will be an interesting year,” Ravitch concludes. ”But also a very dangerous year for American public education.”
More Diane Ravitch: She weighs in on the NYC school report card controversy in today’s New York Post
Ravitch’s claim that there is no research here is inaccurate. The very study that she cites on the effectiveness of charter schools, for example, expressly found that charter caps are harmful.
Comment by Stuart Buck — September 11, 2009 @ 12:21 pm
Stuart,
As I understand it, the CREDO study does not say that charter caps are harmful. On p. 45, it says:
“Differences across states in their charter school policies help to explain part of the observed differences in student results. This study reveals that state laws governing charter school operation have an important impact on student academic growth. Specifically, the presence of a charter school cap correlates with -.03 standard deviations of academic growth. Similarly, states permitting multiple entities to serve as authorizers for charter schools also exhibit negative academic growth, approximately -.08 standard deviations. However, states that allow charters to appeal an adverse decision on either initial charter applications or renewals experience slight positive academic growth, about .02 standard deviations.”
First of all, correlation and causation are not the same. Second, the policy of having multiple entities authorize the charter schools seems to have an even stronger negative correlation with student performance than charter caps do. And states that have caps but allow charters to appeal a negative decision show “slight positive academic growth.”
Now, something puzzles me here. They call 0.2 standard deviations “slight,” but they apparently regard -0.3 standard deviations as “important.” Where is the cutoff between slightness and importance?
Diana Senechal
Comment by Diana Senechal — September 11, 2009 @ 7:17 pm
I think the generally accepted cut-off is 0.25 standard deviations as an “educationally significant” effect size.
Comment by kderosa — September 11, 2009 @ 8:30 pm
Thank you, Ken. For anyone interested, I found a concise explanation of effect size by Howard Bloom, senior methodologist at MDRC.
Comment by Diana Senechal — September 12, 2009 @ 9:13 am
.03 standard deviations isn’t large by many measures. Nonetheless, it’s at least a point in favor of the view that charter caps are a bad idea. The view that charter caps should be left alone is the view that has zero evidence in favor of it.
Ravitch seems even less familiar with the scholarly literature on merit pay; it would be one thing to disagree with the relevant studies, but to say they don’t exist?
Comment by Stuart Buck — September 12, 2009 @ 10:25 am
In my school, most teachers are already working as hard as they can.
How would merit pay make any difference?
Comment by Ben F — September 12, 2009 @ 11:00 am
Merit pay with group incentives might encourage people to cooperate and study each other’s lesson plans, so as to find the best way to teach particular concepts. (See the book “The Teaching Gap” for a defense of “lesson studies”). Merit pay could also, in the long run, encourage better teachers to enter or stay in the classroom (energetic 30-year-olds of high caliber sometimes tend to be dispirited if they they see that all the extra rewards are based on meaningless statistics like being 55 years old with a masters’ degree).
Comment by Stuart Buck — September 12, 2009 @ 12:10 pm
Charter schools are an attractive CHOICE to many poor/minority families in our urban districts because they allow parents to transfer their children out of disastrous urban public schools with regulations that protect trouble-making students. This, of course, prevents other students from learning. Charter parents voting with their feet to leave regular public schools have finally been able to voice their disapproval over the day to day operations of their neighborhood school. The regulations that allow behavioral problems to remain in regular public schools are the stimulus for parents to place their kids in charter schools. Charters call their own shots as to who they keep once the lottery has taken place. Public schools are not afforded this luxury. This freedom given to charter administrators is the eight hundred pound elephant in the classroom no one is talking about. This CHOICE has previously been afforded only to families of wealth. Now, poor/minority families have a choice beyond parochial school. I would contend Obama and Duncan are on the right track by demanding states lift their charter caps in order to be eligible for RTTT monies. I applaud their efforts.
As for test scores being used to evaluate the (in)effectiveness of teachers, I also support Obama and Duncan on this policy. The existing system US public schools employ to evaluate teachers is an embarrassment to the teaching profession, an absolute embarrassment. Anything, I repeat, ANYTHING would be an improvement over the subjective evaluation system used to embellish the performance of 98.5% of every public school teacher in America. IT MUST BE TERMINATED.
Are there problems and/or potholes with using test scores in evaluating teachers? Of course there are. But those indiscretions can and should be worked out between the local teacher union and district administrations.
Hold the high stakes decisions that could potentially be attached to these scores in abeyance for some predetermined grace period (i.e two to three years). Use the test scores only to help improve instruction. No one loses their job or is denied tenure during this grace period until many/most of the potential snags can be remedied. Tweak this new system. Folks like William Sanders have worked to improve value-added-assessments and they’re close – very close to being what we need to determine the effectiveness of our teachers.
With all due respect to Diane Ravitch, and I have nothing but the utmost regard for her body of work, I believe she’s being a bit hasty on RTTT and her judgments on this program I consider to be somewhat premature.
Comment by Paul Hoss — September 12, 2009 @ 3:37 pm
Paul Hoss asks “Are there problems an/or potholes with using test scores in evaluating teachers?” And answers yes, but it can be “worked out.” This is equivalent to asking “Are there problems and/or potholes with sending men to mars?” The technology, Bill Sanders notwithstanding, is not there.
I discussed the problems with Sanders’ model my May and December Phi Delta Kappan Research columns. The best short, readable work on the topic is Henry Braun’s “Using Student Progress to Evaluate Teachers: A Primer to Value Added Models.” http://www.ets.org/Media/Research/pdf/PICVAM.pdf. I summarized Henry’s monograph in my February, 2006 Kappan column.
There was also a special issue of the Journal of Educational and Behavioral Statistics devoted to VAM’s, but if you’re not an expert in statistics, you’ll find only a couple of the pieces accessible. The closer you get to the topic, the more impossible it looks.
The issue of value-added also reveals one of the many internal contradictions in the rapidly growing collection known as The Inane Ramblings of Arne Duncan. Arne says the tests we have are bad. I agree. He demands, though, that states develop data bases so that student performance can be linked to teacher performance. And what will be in these data bases? Scores from the lousy tests.
Jerry Bracey
Comment by Gerald W. Bracey — September 12, 2009 @ 4:33 pm
Stuart,
If our principal said, “Study each others lessons”, we would do it. We don’t need merit pay to get there. What we need is a principal who possesses sound ideas about education instead of the half-baked stuff that comes out of administrator conferences and education schools. The merit pay movement seems premised on the notion that, if only teachers really tried hard, they’d teach better. In my experience, teachers DO try very hard –executing the flawed ideas about education that administrators and professors foist upon them.
And I don’t see many successful young teachers bailing out of the profession because veterans make more than they do. I reckon that many bright young teachers leave because of harsh and unsupportive working conditions.
Comment by Ben F — September 13, 2009 @ 2:35 am
Jerry,
Braun’s article articulates the main drawback to a fair VAM as one of randomization. As someone involved in the placement process for three and a half decades I would have to consider this concern to be a mole hill, not a mountain, especially in lieu of what it could harvest.
Obama could easily imply/mandate to states interested in receiving RTTT monies that randomized placements of students and/or teachers be mandatory to qualify for the funds which according to Braun’s findings would greater validate the process.
If random placement is the major roadblock in this process for school administrators then get them out of the way and bring in a trained cohort of new administrators who are willing and able to rectify this minutia. Beyond that, have the local teacher union oversee the process to EQUALLY protect all its dues paying members. Come on! This is not as challenging as some are attempting to make it.
Massachusetts MCAS tests are not lousy. They’ve been refined over time and are quite possibly the best in the country. Other states could easily use MCAS as a model, validating their results in the process. They might not want to until they discovered their reluctance would effectively foreclose any reasonable opportunity to qualify their state for federal funds.
Comment by Paul Hoss — September 13, 2009 @ 8:48 pm
Paul,
Are you kidding?
You write:
“Obama could easily imply/mandate to states interested in receiving RTTT monies that randomized placements of students and/or teachers be mandatory to qualify for the funds …”
Have we gotten to the point where people are so sure of their pet theories that they would have federal micromanaging to that point? You would seriously consider that level of top down mandates, over-riding people on the ground who might want to consider the people side of education? We would subordinate human preferences, not allowing for subjective, local decisions in order to make some command and control models more effective? What about different teachers and kids having different talents, different chemistry, different feelings, different desires? After all we are talking about peoples’ children!?! You expect parents to cede that authority to D.C.? Different children and different adults click differently with each other. You want a federal Big Brother to ban adjustnebts based on personalities?
Perhaps you are just being argumentative. Otherwise you are crossing the line into social engineering.
Comment by john thompson — September 13, 2009 @ 10:18 pm
John,
You’re probably correct about Obama mandating random placements. He need not micromanage to that degree. Thanks for the slap back to reality on that one. What I am saying is local districts should not be able to use student/teacher placement as the excuse not to go toward VAM’s. They’re simply a tool in the evaluation process whose time has come and something as insignificant as placements should never be allowed to be the deciding factor against their use. And I say the placement process to be insignificant because I believe far too many school personnel think placements are the be all and end all in the educational process. THEY ARE NOT. They might be a cog in the process but they are certainly not the whole enchilada. Kids need to be exposed to a variety of teachers during their formal school experience and teachers certainly the same.
Comment by Paul Hoss — September 14, 2009 @ 12:55 pm
Here’s a response to Ravitch’s post, including cites on merit pay. http://mid-riffs.com/?p=302
Comment by Stuart Buck — September 21, 2009 @ 10:52 am