Network News

X My Profile
View More Activity

Posted at 3:00 PM ET, 09/13/2010

Obama ed policy lacks scientific support -- Willingham

By Valerie Strauss

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

By Daniel Willingham

“We will restore science to its rightful place . . .”

So said President Obama in his inaugural address. Although he was not referring to education policy in particular, he made it clear in his first major speech on that subject that he planned to take scientific evidence seriously when formulating education policy, saying “Secretary Duncan will use only one test when deciding what ideas to support with your precious tax dollars: It’s not whether an idea is liberal or conservative, but whether it works.”

His actions contradict this claim.

The relationship between scientific knowledge and the application of that knowledge is complex and has many dimensions. One dimension in particular is salient here: how confident must we be in a scientific finding for it to merit application (or better, attempted application) in policy?

In standard philosophical views of science, all scientific knowledge is considered provisional. We always acknowledge that our understanding is incomplete, and that a better, more complete theory might be proposed tomorrow.

But a theory can still be reliable enough to be useful. To use a time-worn example, Newton’s description of universal gravitation and three laws of motion proved ultimately to be incomplete.

But these laws provide an amazingly good fit to observed data most of the time, and they are certainly useful.

So a theory doesn’t have to be perfect for it to be applied.

But how imperfect can it be? What should our criterion be for scientific support before we declare “we know this is true, and education policy should reflect it?”

That’s always going to be a judgment call, but given the state of the data, I have to call the president’s judgment into question.

The administration’s education policy has four pillars. These were outlined in that first education policy speech, and they were represented in the requirements for Race to the Top applications.

But of the four, only one could be said to have clear scientific support.

The well-supported policy initiative is an emphasis on early childhood education. Aside from ample research by developmental psychologists showing that the early childhood years are a critical time for learning, economists have conducted persuasive studies showing that early childhood intervention programs can have lasting and profound effects on at-risk kids.

Better-educated kids are more likely to be tax-paying contributors to the economy and less likely to be incarcerated or on public assistance. Thus, in the long run intervention programs for at-risk kids more than pay for themselves.

For the other three policies that the administration emphasizes—teacher evaluation using student data, a greater number of charter schools, and improved standards—the data are more mixed.

Much has been written in the last month about the reliability and validity of value added measures of teacher performance based on student test scores, subsequent to the Los Angeles Times funding the calculation of these scores for Los Angeles-area teachers and then publishing the individual results.

But long before that, measurement experts had published reviews of the scientific literature with a general consensus that such models are a promising way to evaluate some aspects of teacher performance, but the resulting data are not yet reliable enough to be applied.

That’s why the Board on Testing and Measurement of the National Research Council wrote a letter to Education Secretary Arne Duncan expressing, among other things, concern over the emphasis on value added measures in the Race to the Top competition.

How about charter schools? The key question is whether kids who enter charter schools learn more than comparable kids entering other public schools. The data on this question are mixed.

Some studies indicate that the answer is “no,” and others suggest that the answer is “yes.”

It would be nice if the reduced bureaucratic burden that typifies charter status automatically led to better student outcomes. That seems not to be the case. Rather, it seems probable that some charter schools are well run and effective and others are not.

The best hope for this approach may be that status as a charter may allow for more rapid change in ineffective schools or for the closure of the school. If there is a reliable method to employ that strategy, we’re not there yet.

What about improving standards? On logical grounds, tough standards seem like a good idea; surely it’s better for teachers to aim for excellence than to aim for mediocrity.

But tough standards are not sufficient. We know that because there has been variation in state standards in the past, and there is no relationship between the quality of the standards and student achievement on the National Assessment of Educational Progress, according to an analysis by Russ Whitehurst & Michelle Croft.

I’m not concluding that any of these three ideas--teacher evaluation based on student test scores, a greater number of charter schools, or tougher standards--is bad or hopeless.
But it is inaccurate to claim that in emphasizing policies based on these ideas, Mr. Obama and Mr. Duncan are “restoring science to its rightful place.”

Scientific support is not the only arguments one can advance to support a policy. Other arguments ought to be tried, because the scientific support is inadequate.


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

By Valerie Strauss  | September 13, 2010; 3:00 PM ET
Categories:  Charter schools, Daniel Willingham, Education Secretary Duncan, Guest Bloggers, National Standards, Race to the Top, Research, School turnarounds/reform, Standardized Tests, Teacher assessment  | Tags:  common core standards, daniel willingham, duncan and policy, duncan and reforms, naep, national assessment of educational progress, obama and education policy, obama and inaugural address, obama and school reform, president obama and education, race to the top, race to the top requirements, science and obama administration, teacher assessment  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: Student loan default rates increase -- especially at for-profit schools
Next: Obama picked wrong school to give speech


Well, Obama should read this article regarding scientific research in education (based on Feynman's views on the matter) and consider it in relation to formulating education policy:

Classroom Research and Cargo Cults
by E. Donald Hirsch Jr.
All the trappings of scholarship, missing only real results

And yes, I do believe that there is indeed some scientifically sound research in the realm of education, but I am inclined to believe that the Obama/Duncan duo clings to the pseudo science that has more holes than Swiss cheese.

Posted by: shadwell1 | September 13, 2010 11:41 AM | Report abuse

Why is it that so many buy into the promotion of methods with no support from the data?
Schools appear to be the scope of action but "schools" are not the problem it seems. How do you deal with the environmental issues when the environment is outside the purview of the school?
It sounds like Duncan is copping out by hammering on a popular scapegoat (teachers, unions, public schools) and hoping he gets out before it's time to pay the piper.

Posted by: zebra22 | September 13, 2010 1:05 PM | Report abuse

The writer is correct that science is always in flux. But Al Gore says the science is settled on a major matter (anthropogenic global warming). That is something to be scared of, when someone like Al Gore has so little understanding of what science is. Apparently Obama suffers from the same disease. Time to give the man a pink slip.

By the way, who is proofreading these articles? It is about education, so I would like to see some corrections in the copy.

Posted by: spellerman | September 13, 2010 2:04 PM | Report abuse

It is quite easy to say, "This teacher has students with high scores, she is excellent and this teacher has students with low scores, he is ineffective." It would be nice if it were that easy.

It is too bad that there is no real reform in education today.

Posted by: celestun100 | September 13, 2010 4:09 PM | Report abuse

"Secretary Duncan will use only one test when deciding what ideas to support with your precious tax dollars: It’s not whether an idea is liberal or conservative, but whether it works."

This is not a reference to science but more an argument for objective data. There is a difference. Whether a policy works depends more on objective data than science. I realize this is semantics but it's important to differentiate.

Of the three policies you refer to in RttT, I believe you overlooked one for this discussion.

But first charters; why are Obama and Duncan pushing so hard for charters when the research on their effectiveness appears conflicting? As our first African American president, it's understood Obama is obliged to do as much as he can for this cohort of students. While he didn't go to the extreme of vouchers, with charters he has finally made available to poor/minority families a choice as to where to send their children to school, a choice previously afforded only to families of wealth. So while the objective data on charters is conflicting, his heart is in the right place.

Linking student test scores to teacher evaluations goes beyond controversial as it apparently is rebuffed by the pyscho-metrician community - TO THIS POINT IN TIME. What's interesting about this policy however is it is gaining traction in pockets around the country. Even more interesting is the Al Shanker-like acceptance of this direction by one of the two major teacher unions in the country (AFT) and especially its president, Randi Weingarten. If the objective data is so overwhelmingly opposed to this policy, how is it that teacher unions, and in effect teachers, have resigned themselves to such a notion? It's because they realize there is some degree of validity/reliability to such a concept. The performance of their students should in some way be linked to their level of effectiveness, at least over time (4-5 years of data). This takes such variables as family background into perspective and still allows for the value added by a teacher over a period of time.

The last policy you cite in RttT, standards, is not as significant as Obama/Duncan's idea of shuttering or dramatically altering chronically performing schools. Not sure if the review of the literature is complete on this one, but again, the administration is reaching out to the poor/minority community in an attempt to make opportunities better for them.

In all of this debate about education reform we have to remember NCLB and RttT were not written with the kids from Beverly Hills or Scarsdale in mind. These federal policies were developed to address the needs of youngsters in communities like the Bronx, Watts, and Roxbury. As such, the policies emanate more out of a degree of pragmatic caring than they do strictly from a scientific viewpoint.

It’s not whether these policies are liberal or conservative, but whether they work for the targeted population. As a public school teacher, I believe that they will.

Posted by: phoss1 | September 13, 2010 5:21 PM | Report abuse

Lacks scientific support? I've already seen on so called "progressive" websites that we should obviously roll back President Obama's initiatives. Health Care reform doesn't necessarily have scientific support, but everyone (progressives) love the idea of that.

The article does state that evidence is "mixed" on some of the areas of President Obama's ed policy. One area for instance, charter schools, is shown to give uneven data. Much of this has to do with uneven charter schools, which has to do with weak charter laws and poor authorizing boards (which is basically said in the article).

Also that "tough standards are not sufficient," again, fair, but duh.

Posted by: delray | September 13, 2010 6:02 PM | Report abuse

I wonder if one of the reasons Mayor Daley is retiring is because of the failure of Blarney Duncan's education program in Chicago.

Posted by: educationlover54 | September 13, 2010 7:10 PM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company