The study, published this month in the National Bureau of Economic Research's online working paper series, found that on test days many districts fed students high-energy foods with low nutritional value because the empty calories gives students a short-term mental lift much like carbohydrate loading energizes athletes, said David Figlio, a UF economist whose research focuses on the design and evaluation of education and social policy.
"We find significant evidence that school districts, particularly those threatened with at least one failing or sanctioned school, respond by giving students more empty calories on testing days," said Figlio, who also is a faculty research fellow for the National Bureau of Economic Research. "These are calories found by nutritionists to have substantial very short-term cognitive effects, but no long-term benefits."
Studies have shown that students perform better at academic tasks in the several hours following a large dose of glucose or empty calories, which improves knowledge or cognitive ability in addition to boosting energy.
Figlio and Joshua Winicki, an economist with the American Institutes for Research, did a detailed study of elementary school lunch menus from 23 randomly selected school districts in Virginia during the 1999-2000 school year, comparing nutritional content on days of state-mandated tests with that of non-test dates. The study was funded by The National Science Foundation.
As part of the national 2001 "No Child Left Behind" Act, states must evaluate schools based on the percentage of students meeting proficiency standards on state curriculum-based examinations. Schools with large numbers of students failing the test face sanctions that may include loss of state funds.
In districts with at least one school in danger of facing sanctions, school lunches averaged 863 calories during the testing period, compared with 761 calories before and 745 calories after, Figlio said.
Despite the calorie boost, however, the nutritional content of the food was lower on test days, he said. Vitamin C averaged 34 milligrams during the testing period as opposed to 37 milligrams before and 44 milligrams after the test. Vitamin A averaged 531 retinol equivalents, or RE, during testing, but 615 RE before and 672 RE afterward. The dietary changes were subtle but may include corn dogs instead of hot dogs or different types of pizza, he said.
The empty calorie approach seems to work in improving test scores – at least in the short run. Among schools threatened with sanctions, those that increased calories the most during test days saw the largest gains in student achievement, he said.
Although the sample size was small, the results suggest that raising calories by an average of 110 resulted in higher pass rates of 11 percent for mathematics, 6 percent for English and 6 percent for social studies, he said.
"This study provides another piece of evidence that evaluating schools on the basis of aggregate student examination scores is not a silver bullet," Figlio said. "This study, as well as others looking at school responses to high-stakes testing, suggests that these test scores are more manipulatable than one might naively assume."
Other research, for example, finds that schools alter their discipline policies during the testing period or move more low-performing children into special-education classes in which students often aren't required to take these tests, he said.
Diet is a plausible way to improve test scores because students who don't reach the minimum acceptable threshold in school accountability systems are overwhelmingly from low-income income families eligible for subsidized lunches, he said.
The researchers focused on Virginia, a state with a school accountability system in place since 1995, because they had access to detailed elementary lunch menus collected by the U.S. Department of Agriculture. But similar results likely could be found anywhere in the country, he said.
"Performance incentives are performance incentives, so I don't see anything special about the Virginia exam that would make Virginia more or less prone to this phenomenon than anywhere else," he said.
Figlio said he doubts that schools increase calorie content because they want to reward students with their favorite foods for the stress they endure on test days.
According to one school district in the sample, elementary students' favorite meals, as measured by sales, are pizza, cheeseburgers and tacos. Yet schools in the sample served these items on only 35 percent of testing days, compared with 34 percent of non-testing days, he said.
Writer: Cathy Keen
Source: David Figlio