Network News

X My Profile
View More Activity

Posted at 9:09 AM ET, 03/15/2010

Willingham on school choice

By Valerie Strauss

My guest is cognitive scientist Daniel Willingham, professor at the University of Virginia and author of “Why Don’t Students Like School?”

By Daniel Willingham
Education historian Diane Ravitch’s book, "The Death and Life of the American School System", has been well publicized and I won’t review it here, other than to encourage you to read it, as I encourage you to read anything that Ravitch has written. She is a brilliant scholar.

I have been fascinated by the response to the book, and in particular to Ravitch’s conclusion on charter schools, and more broadly on the idea of school choice. The hope was that school choice might lead to an array of benefits: leaner bureaucracy, satisfied parents, motivated administrators, and, ultimately, better student learning.

That was the theory, the model of what was supposed to happen. Ravitch concludes that the data don’t fit the prediction; they are uneven. I believe a fair summary of Ravitch’s position is: “If school choice were going to have a profound, positive influence, we’d know that by now.”

What do you do when a theory makes a prediction and the observed data don’t fit? This is a common problem in science, but there is not a set of guidelines for an appropriate response. Ravitch’s response is one: abandon the theory. Other possible responses have appeared in critiques of Ravitch’s book:

You can claim that the data actually do fit the model, as Mark Schneider does.

You can claim that there aren’t enough data to make a judgment yet as Paul Peterson does.

You can claim that the wrong prediction was tested, as Rick Hess does.

You can claim that the experimenter doesn’t really know how to interpret the data, as Andrew Coulson does, or that, at least in this case, the data have been analyzed improperly, as Nelson Smith does.

Another option, not taken up by any commentator that I’m aware of but that seems the most reasonable to me, would be to modify the theory and thus the predicted outcome. School choice will not help under all circumstances, but it might under some circumstances. The next task would be to generate hypotheses about the conditions under which school choice ultimately results in improved schooling.

So which response is the right one? Don’t scientists have rules of thumb as to when we abandon theories, when we know that we have enough data that they have been suitably tested, and so on?

No, we don’t. It is one of the chief topics over which philosophers of science have argued, from the logical positivists, to Karl Popper, to Thomas Kuhn to Imre Lakatos: On what basis do scientists discard ideas, and on what basis should they do so?

Paul Feyerabend, philosophic gadfly, delighted in pointing out that for any rule of thumb about how science ought to progress, one could look at the history of science and find an instance of a great discovery that would have been missed, had the rule of thumb been respected.

In one of Feyerabend’s compelling illustrations, Galileo ignored plain data, right before his eyes. Galileo championed Copernicus’ model of a heliocentric solar system—but if the earth is spinning, isn’t it obvious that a rock dropped from a tower should not fall near the tower, but away from it?

It doesn’t, but Galileo overcame that obvious bit of data in his thinking. Thus, Feyerabend argues, even the obvious dictum “a theory ought to fit the observed data” should not be respected one hundred percent of the time.

But that brings us to another vital point from the philosophy of science, emphasized by many, if not most philosophers: Theories should not be evaluated solely on how well they fit observations in the world. They should be compared to one another.

Don’t ask, “Does school choice lead to better student learning?” Rather, ask “How does school choice stack up against our other options?”

More than a free market of school choice that is implemented universally, I would like to see a free market of competing ideas for school reform, systematically compared.


Follow my blog all day, every day by bookmarking And for admissions advice, college news and links to campus papers, please check out our new Higher Education page at Bookmark it!

By Valerie Strauss  | March 15, 2010; 9:09 AM ET
Categories:  Daniel Willingham, Guest Bloggers  | Tags:  Daniel Willingham, charter schools, guest bloggers  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: On Obama (and Jay Mathews) and teacher accountability
Next: Former Bush official backs student loan reform


The thing is, School choice, is not a theory. It is an extrapolation of "free market" economic theory. In the wake of the financial crash, even enthusiasts like Alan Greenspan have abandoned that theory.

The effects of standardized testing and school choice are not theoretical. Ravitch documents how they are undermining public education. Checker Finn speaks for the "reformers" in openly admitting that the aim is to "blow up the system."

No critic of the book has refuted the effects Ravitch describes or has presented any alternative course of action.

An alternative, which Ravitch describes, is to return to the original intention of charter school--to ensure that each is a laboratory for testing instructional options, not for testing and comparing teachers and students on measures that are instructionally insensitive.

Standardized tests leave instruction a black box. They shed no light on the instruction that has been delivered or how the instruction, whatever it was, can be improved. "Better tests" constructed and interpreted using the same theory will yield no better information.

There are alternatives described in papers at the Social Sciences Research Network:

Posted by: DickSchutz | March 15, 2010 12:59 PM | Report abuse

I have this weird idea.

Let us test all of the children that enter the public school in a poor area and all the children that enter a public school system in an affluent area.

It is not impossible to have these tests since NYC does this type of testing on limited number of children already where parents want to apply for better and special schools.

The results of these tests would probably show a large difference in the children and would save us from the continuous mistake of believing that the problem must be the difference between schools and/or teachers.

But this would not be politically correct and it is far better to continue to believe that our world is the center of the universe and does not move.

Posted by: bsallamack | March 15, 2010 2:29 PM | Report abuse

I would say that the debate over school choice suffers from the same problems that much of education reform suffers; that of the expectation of a simple answer. I am not sure that Ravitch would advocate abandoning school choice per se, but rather, abandoning school choice as a panacea. Her criticism of "fads" is really a criticism of simplistic thinking when it comes to education reform, whether it is thinking that montessori will work for everyone, or KIPP will work for everyone. My impression of her work (and I haven't read the book yet) is that she would urge us not to abandon school choice and look for something that works, but acknowledge the opportunity costs of looking for anything that "works" in the simple way that policymakers and the p
public want it to work.
One unfortunate piece of collateral damage of the great hope of charters and accountability (and every other simplistic solution to education reform) is that it drastically undermines the value in the profession of teaching. If education is merely an engineering problem, waiting for the right technology, or the right checklist, or the right universal recipe, then we don't need to invest in training our teachers, or supporting them in their efforts to practice their craft.
Not that I really disagree with anything Prof. Willingham said above. But I think the idea of some sort of Darwinian (more like Spencerian) competitive marketplace of ideas, where the "best" win out and are pitted against each other head-to-head, ignores the likely reality that there are no overall winning solutions, but rather a complex ecology of strategies and principles that work in some circumstances more than others. This is not to say that standards are useless, or that there are no universals to learning (Prof. Willingham's book offers several fine examples) just that these standards and universals should be implemented with care by the people who know the students in front of them: the teachers.

Posted by: formerDCPSstudent | March 15, 2010 7:56 PM | Report abuse

"I would like to see a free market of competing ideas for school reform, systematically compared."

Hear, hear. Unfortunately, for at least the past 20 or more years K-12 education has been dominated blowhard policians, ambitious superintendants, and good-old snakeoil salesman -- each with a remedy that will cure all ills. How delightful it would be to have a bit of humility on the part of all involved, genuine open-mindedness, and a real curiosity about finding out what works. Exhibit A that we don't yet have that is Arne Duncan, with his clairvoyant ability to identify the worst 5% of schools and his package of untested tools to fix them. Thank goodness our medical system doesn't work like this or we'd still be applying leeches to cure disease.

Posted by: dz159 | March 15, 2010 8:32 PM | Report abuse

//What do you do when a theory makes a prediction and the observed data don’t fit?

Why do you ignore the obvious answer to this question? The "theory" is wrong! Time for a new "theory", huh? The theory was bogus from the beginning, and many children have been harmed.

The obvious answer is: END THIS FAILED EXPERIMENT before more children are hurt. That answer doesn't seem to come to you, for some reason.

Posted by: editor2 | March 17, 2010 1:11 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company