MIT neuroscientists find even high-performing schools don’t influence their students’ abstract reasoning
States commonly rely on standardized exams to measure school quality and make important decisions about student advancement. Passing those exams often factors into whether students receive a high school diploma, and research has shown that strong performance on such tests predicts future educational attainment as well as adult employment and earnings. These assessments primarily measure what psychologists call crystallized intelligence — the knowledge and skills students acquire through schooling and study.
In a new study led by MIT neuroscientists in collaboration with education researchers at Harvard University and Brown University, researchers examined whether schools that raise students’ standardized test scores also produce gains in fluid intelligence: the capacity for abstract reasoning, working memory, and rapid information processing. The answer, based on the study’s findings, is largely no.
The study followed nearly 1,400 eighth-graders across Boston public schools, including traditional district schools, exam schools, and charter schools. The researchers compared gains on the Massachusetts Comprehensive Assessment System (MCAS), a statewide standardized exam, with performance on independent measures of fluid cognitive skills such as working memory capacity, processing speed, and novel problem-solving.
John Gabrieli, the Grover M. Hermann Professor of Health Sciences and Technology, professor of brain and cognitive sciences at MIT, and senior author on the forthcoming Psychological Science paper, explains the central question: if a school succeeds in raising MCAS scores and thereby improves students’ opportunities for college and beyond, does that success also come with improvements in other cognitive abilities? The research found that, although some schools produced meaningful improvements in MCAS scores, they had almost no measurable effect on students’ fluid reasoning or related cognitive capacities.
Quantitatively, schools accounted for a substantial portion of variation in MCAS performance — about 24 percent for English and 34 percent for math — but they explained less than 3 percent of the variation in combined measures of fluid cognition. In one fluid-reasoning task used in the study, students selected which of six images completed a visual puzzle, a test that requires integrating shape, pattern, and orientation information and applying reasoning to a novel problem. These tasks are intentionally different from curriculum-based assessments because they require flexible thinking rather than recall of learned material.
A particularly informative part of the analysis compared roughly 200 students who had entered lotteries for admission to oversubscribed Boston charter schools. Many of these charter schools have a track record of boosting MCAS scores. Students who won the lottery and attended high-performing charter schools showed significantly higher math MCAS scores than those who did not win admission, but their performance on fluid cognitive measures did not improve correspondingly. That pattern suggests that school-driven gains in crystallized knowledge do not automatically translate into gains in fluid abilities.
Amy Finn, an MIT postdoctoral researcher and lead author of the paper, emphasizes that fluid tasks differ from curriculum-based work because they require attention to changing task dimensions and the application of reasoning in unfamiliar contexts. “It’s the application of reasoning skills in novel situations,” she notes, which helps explain why success on traditional school tests does not guarantee comparable improvements in these kinds of skills.
The researchers stress that the study is not a critique of schools that raise MCAS scores. On the contrary, improving students’ knowledge and academic skills is valuable: better reading comprehension, stronger mathematical skills, and higher standardized test scores all open doors to higher education and career opportunities. The key implication is that gains in crystallized knowledge should not be assumed to produce parallel gains in fluid cognition.
Because fluid cognitive abilities — such as working memory, attention control, and inductive reasoning — reliably predict academic performance, the authors argue that educational policymakers and practitioners should consider whether targeted practices can be added to curricula to strengthen these skills. Some intervention studies have reported improvements in aspects of executive function and reasoning through focused training, but evidence remains mixed about which approaches produce consistent, transferable gains across diverse student populations.
The research team plans to continue tracking the cohort, now in 10th grade, to monitor how academic outcomes and life trajectories unfold over time. They are also participating in a related study of high school seniors to examine how standardized test results and cognitive abilities relate to college attendance and graduation rates.
Notes about this cognition and psychology research
The study was conducted in collaboration with the Center for Education Policy Research at Harvard University, Transforming Education, and Brown University. Funding came from the Bill and Melinda Gates Foundation and the National Institutes of Health.
Contact: Anne Trafton – MIT
Source: MIT press release
Image Source: Image credited to Wikimedia Commons user Onderwijsgek (Attribution-Share Alike 2.5).
Original Research: The research will be published in Psychological Science.
Keywords: fluid intelligence, crystallized intelligence, standardized tests, MCAS, education policy, cognitive skills, working memory, executive function