It's Time to End Timed Tests
Why do we believe that speed reflects intelligence?
The outrage this year over the attempts of the rich and infamous to rig the college admissions process in favor of their children has focused new attention on an old issue: purchasing a diagnosis to qualify for extended time on standardized tests. During my 18 years as an assistant head of school, from the late '80s to the early 2000s, the increasing number of students with psychologists' recommendations for extended time made me suspicious. While students with real learning difficulties are legally entitled to a number of "reasonable accommodations," others seemed solely interested in being allowed more time on tests—tests given by classroom teachers and, especially, the SAT and ACT.
A few conversations with a psychologist who offered families this diagnostic service confirmed my suspicions that not all of these diagnoses were legitimate. He explained the complexity and inexact science of arriving at a diagnosis and spoke of the pressure from parents whose goal was to obtain the recommendation for extended time. He recounted instances when he resisted the pressure, so the parents procured the recommendation from someone else.
I began to wonder about the system itself. Why do we bother with timed tests? Why do we believe that speed reflects intelligence? As teachers, we see all sorts of students who work at different speeds, which produce both intelligent and not-so-intelligent results.
I taught English, so my tests tended to require essays or short paragraphs. Annie always aced them. Ideas seemed to just flow from her pen as she hunched over her paper. She usually finished early. Bill also wrote quickly, but he was trying to list all his thoughts before he actually started organizing and writing his responses to the essay questions. At the bell, he was often rushing to finish. Sally was slow. She tapped her pen on her palm and stared into space, thinking about what she wanted to say, searching for that first sentence that was always so important to get her launched. She produced wonderful essays written outside the classroom, but her results on timed tests were poor.
Circumstances play a significant role in performance. Harvard University researchers Kurt Fischer and Thomas R. Bidell noted in a 2006 paper, "Children (and adults) show distinct levels of competence under different conditions." Their research suggests that learning involves a process of building, regressing, and rebuilding neural skills. Rather than memorizing facts and procedures, learning math requires that we build webs of interrelated skills for math. Understanding the Civil War means building a conceptual understanding of the Civil War. At first, these skills for thinking are fragile, requiring nurturing conditions of high support.
Typically, in less supportive circumstances (anxiety over a high-stakes test, being asked to tackle an unfamiliar problem), the skill collapses (regresses) and must be rebuilt. Regression actually plays an essential role in learning. Rebuilding these networks results in increasing stability. Our obsession with speed ignores the reality of learning.
Speed is less a sign of intelligence than a sign of the automaticity that many experts evince, and it seems nuts to expect young learners to be experts, let alone experts in the many different fields they are required to study in school: expert mathematicians, historians, scientists, writers, readers, linguists, artists. Experts have developed deep knowledge, understanding, and experience that can allow at least some of them to function quickly and intuitively, usually in a particular field and under certain circumstances. Why would we expect students to have developed this sort of intuitive ability?
Neuroscientists Mary Helen Immordino-Yang and Matthias Faeth explore the development of "skilled intuition" in young learners—the integration of "emotional with cognitive knowledge." They write, that while "students may be slower to build the full representation of the material, without the development of sound intuitions undergirding their representations, it is likely that the students will not remember the material in the long-term, and that even if they remember it in an abstract sense, they will have difficulty applying it to novel situations."
Many tests ask students to do exactly this: quickly apply to new situations skills and understanding that they have just begun to develop. They simply aren't ready. Our emphasis on speed in school is antithetical to stimulating meaningful learning, the sort of learning that we claim is the goal of education.
As we ask students to approach a problem, we need to understand the many factors that can affect the speed with which they might solve it: How complex will this student find the problem? How much experience has that student had with problems like this one? What's his skill level? How methodical does she tend to be when working through problems? How likely is this problem to initiate the collapse of his fragile understanding of the concept being tested? What is her stress level? The answers, even if we could know them, will vary for each student.
Researchers have helped us understand that all brains are different. A normal brain or an average brain is a myth. Yet school practices continue to reflect the myth. Some vague notion of an average brain sets the expectations for the speed with which students "ought" to complete a task. The results—the failures, the mediocre performances, the frustration felt by both students and teachers—suggest there is a problem with this approach. As the stakes go up, as grades and scores determine who will gain entry into the "best" colleges, desperate people opt to cheat.
In my own classroom, I found a simple solution: extended time for everyone. The results were gratifying. Less stress eliminated one major cause of skill collapse. Students could work at their own speed—read questions carefully, organize their thoughts, start again, think. All students could move through tests at whatever pace was comfortable, and more students performed well. Most still finished within the time for which I had designed the test, but a few stayed an hour or more later (though, interestingly, not most of the students who qualified for extended time).
Extending the time for all seems a more productive approach to establishing equity. Perhaps it's time for the SAT and ACT to take this small step—at least until we finally align our approach to assessment with our growing understanding of learning and the brain.
Vol. 39, Issue 07, Page 16Published in Print: October 2, 2019, as Don't Stop, Don't Put Down Your Pencils