That sample is 10 times larger than any previous study connecting genetics to behavior-related areas like education, a landmark change, considering most prior studies of this sort have used groups of a few hundred to a few thousand people. It’s not even like looking for a needle in a haystack, but more like looking for a needle you think might be there in one wheelbarrow full of an entire field of hay.

Tiny Differences

After combing through and comparing more than 2 million individual genetic locations, the researchers found the equivalent of three metal filings, hinting at the presence of those needles. One genetic variant was found to be associated with the number of years of overall education a person attained, and two others were associated with college completion. Individually, these mutations could only explain about .02 percent of the difference in people’s educational attainment. Altogether, the three SNPs explain about 2 percent of the difference, or a month’s worth of schooling in the course of a person’s life. 

Potential Uses

Cesarini and his colleagues are adamant that they have not found an “education gene,” and in fact there are no implications from the study for education policy, practice, or law. We’re nowhere near genetic screening for preschoolers to target early-education interventions, and the researchers note that genes can change in response to education conditions, too.

“We’d like to understand what channels through which these SNPs are influencing educational attainment … but this is very, very preliminary and it will be very challenging to do that,” Mr. Cesarini said. “Obviously, these snips don’t have a proximal or direct effect; they affect attainment through a very complex chain of influences and phenotypes.”

However, the researchers note the findings do point to areas for further study: The SNPs associated with education levels are on genes that have been previously associated with specific health, nervous system, and cognitive processes, and other analyses suggest they may be linked to learning and memory processes as well as neuron development.

“Ultimately we would like to understand, if you discover genetic risk factors for problems of low educational attainment, you’d like to understand what policies could be in place that would influence those risk factors,” Mr. Cesarini said. “There’s a common misconception that if something is a genetic effect you can’t change it. That’s not true … Sometimes you are in a very good place to change that relationship.”

Education Mosaic

To be clear, this would predict an individual student’s likely educational trajectory, but it could help control for differences in populations for social scientists looking at large policy or program changes. “Can 10 or 15 percent be valuable in a social setting? It would be, but only in a situation where the intervention would be very, very expensive and you need to keep the sample as small as you can,” Mr. Cesarini said.

The study adds to a complex mosaic of factors influencing education, from genetic variations large and small, to environmental aspects like lead exposure and toxic stress. And all of that is before taking into account the more direct and obvious effects of parents, teachers, curricula, education policies, and so on and on.

A recent meta-analysis of genetic impacts on educational attainment by researchers at Northwestern University in Evanston, Ill., found genetics and environment contribute in different ways to education depending on the population, and suggest education is less influenced by genetics than other social areas. 

For example, the researchers found, “Children from the same household are likely to differ more in extroversion, alcohol consumption, smoking habits, and even on their IQ, than they do in their educational attainment.”

Decades of cognitive, social and economic research show people with high IQs are often just as likely as those with lower intelligence scores—sometimes even more likely—to fall victim to an array of cognitive biases and other bad thinking strategies.

For example, a student may be very knowledgeable, but not intellectually curious and closed to information that conflicts with his or her previous understanding. A high IQ may reflect a student’s ability to understand how to calculate compound interest, but it wouldn't necessarily ensure that he or she would choose $100 in a month over $20 tomorrow.

This spring, the John Templeton Foundation awarded a three-year, $1 million grant to develop a “rationality test” to Stanovich and colleague Richard F. West, a psychology professor emeritus at James Madison University in Harrisonburg, Va. The assessment is still 18 to 24 months away from field testing, Stanovich said, but the basic framework is in place. 

“Rationality is a more encompassing construct than intelligence,” Stanovich said, but because our irrational behaviors have been such a popular subject to study, he believes there are enough existing types of cognitive tasks to develop a good assessment. “We should not shirk from measuring something just because it is logistically difficult.”

There's no video of Stanovich's talk at APS, but he goes into detail about the cognitive problems with existing intelligence tests in a 2010 talk at Portland State University in Oregon, below.

At the same time, in a letter to Acting NSF Director Cora B. Marrett, Rep. Smith said he had “concerns” about the “intellectual merit” of five grants and asked for peer review and program officer notes on their approval.The chairman also began circulating a proposed bill that would change the grant review process at the NSF and potentially at other federal research agencies.

The bill, called the “High Quality Research Act” in a draft obtained by the American Association for the Advancement of Science’s newswire ScienceInsider, would require the NSF to certify, on an online website, that any grant-funded research is advancing national health, prosperity or welfare; not duplicating other federal research; and “is the finest quality, is groundbreaking, and answers questions or solves problems that are of utmost importance to society at large.”

The proposal would expand on similar language introduced in the fiscal 2013 budget agreement back in March by Republican Sen. Tom Coburn of Oklahoma. That language, ultimately included in the final budget, barred funding of any political science research except studies the NSF director certified as relevant to the U.S. economy or national security.

The squabble carries high stakes for education. In the fiscal 2013 budget, the National Science Foundation’s budget for science, technology, engineering, and mathematics education alone is more than $1.15 billion. By contrast, the federal Education Department’s research agency, the Institute of Education Sciences, had $597.3 million to spend on education research this fiscal year.

Kenneth Prewitt, a professor of international and public affairs and director of the Scholarly Knowledge Project at Columbia University, New York, takes a similar view.

“If you weaken science anywhere, in effect you weaken it everywhere,” Mr. Prewitt said. “If there are other areas of contention in science—as there certainly are in the ed reform movement—insofar as there are legitimate political interests that don’t want these particular findings made available, they can make sure that research never gets done.”

Mr. Prewitt—who argued in an earlier Science editorial against Sen. Coburn’s budget language restricting political science—told Education Week that trying to judge the practical value of foundation research is difficult, and adding political litmus tests would make it more so.

“Ahead of time, you simply don’t know what good social science and natural science is going to put into play. Trying to stop it seriously undermines the national interest,” Mr. Prewitt said. “Science is not designed to be a quick answer to an immediate problem, and if you try to make it do that, you miss out on a lot of fundamental truths that may later prove useful.”

For example, "Fifty years ago, what did people think about early childhood learning? They didn’t even think it took place. But research went on, got better and better, and now, in 2013, we think of learning even in the womb, and we are busily doing all kinds of early childhood programs," he said.

The proposed change to peer review has drawn widespread criticism from science advocates and Congressional Democrats. President Obama and Mr. Holdren criticized the move last week in separate addresses to the American Association for the Advancement of Science.

“You can have a general testing anxiety that might go across content areas, but children often have anxieties that can affect one particular domain, and it often plays out very similarly whether it is anxiety about math or about reading,” Ms. Beilock said. “There’s been, I would argue, a paucity of work examining these anxieties in young kids.”

In a forthcoming study, "Math Anxiety, Working Memory, and Math Achievement in Early Elementary School," in the Journal of Cognition and Development, Ms. Beilock and University of Chicago doctoral candidate Gerardo Ramirez tested 154 1st and 2nd graders in both math and working memory, a brain function associated with decisionmaking and academic performance. They then separately gauged the students' math anxiety.

The researchers found that math anxiety hamstrung the math performance of the students that might otherwise be expected to perform best--the ones with higher working-memory capacity. Students with better working memory tended to use more complex strategies to solve problems, while those with lower working memory resorted to simpler strategies, such as counting on their fingers.

“Working memory is important in lots of academic subjects,” Ms. Beilock said, but, “In math, juggling numbers in your head and keeping some things in and other things out, it’s a place where having this cognitive horsepower at your disposal can be particularly important.”

Most research on the problem has focused on students in secondary school and college, when fears around numbers have been found to hurt students' performance in higher math classes like calculus and turn them away from pursuing careers in science, technology, engineering and math fields. 

But even preservice and working teachers, like Miss Anne Thorpe, recall dread when it came to poring through arithmetic.

On the plus side, the self-confidence and encouragement of parents and teachers can help math-anxious students learn to relax, too. Laura Bilodeau Overdeck, who as a child had "memorized perfect squares for fun," started giving her own children math story-puzzles at bedtime along with their fairy tales. That tradition grew into a blog for friends and then an online community, Bedtime Math, which provides daily math games and puzzles for parents to do with their children.