Sometimes a study is published that has the promise of being a cultural game-changer. It carries the potential to reset the way a significant portion of the population thinks or behaves regarding an issue. Last week a study that raises questions about the value of standardized test scores as a predictor of college success was released. My hope is that the findings will launch a long-overdue cultural conversation about the "gatekeeper" role these tests play in the college admissions process. The study tracked 123,000 students at 33 test-optional public and private colleges and universities over 8 years and concluded that high school GPA is actually a better predictor of college success than SAT or ACT scores.
Not that shocking, right?
Even correcting for potential grade inflation, it makes sense that students' grades over the course of four years in high school would be a better predictor of how they will fare over four years in college. The habits of mind and approaches to work that lead to success in college are difficult to demonstrate during a single Saturday spent filling in bubbles in a high school gym.
There are many reasons these contrived tests are limited predictors, but one of the most glaring is that they do not reflect the way students make important intellectual decisions or fulfill academic work requirements. For example, the tests do not reward students for using an academic strength in one area to compensate for a weakness in another. If a student is a math wiz, but struggles with being a slow reader, it is not possible for her to take extra time from math section and apply it towards reading the verbal section at a slower pace. Outside of the testing process, students regularly leverage their strengths to offset their weaknesses. In fact, students who seamlessly compensate their weaknesses with strengths perform better in school overall.
Additionally, the tests do not measure curiosity. They do not measure the ability to formulate an original idea. They do not measure creativity. They do not measure work ethic.
There is the complex issue of Time. No one will argue with the tests' ability to measure of a student's test-taking speed. But when did our nations' colleges actually decide that answering questions quickly was a valuable metric for measuring a student's intellectual abilities? I'm unaware of research that supports the idea that educators learn anything additional about students' depth or breadth of knowledge by measuring how quickly they can recall information or express what they have learned. There are many things that contribute to a student's test-taking speed, and it is not always influenced by lack of mastery.
Even in college, it rarely matters how long a student spends on a project -- students are evaluated on the quality of their work. Did their work meet or exceed expectations? Did it meet the deadline? They will likely not be asked, "Did it take you two hours or four hours to prepare for that presentation, or to write that paper or research that problem?" As a teacher, I recognize that certain thinkers take more time to process information, even if I am not always sure of the underlying causes of the delay. What matters most is that the student finds a work style that allows them to produce timely quality projects.
This new study is obviously big news for dyslexics, who rarely do well on standardized tests. But dyslexics are just one of many groups that are unfairly handicapped by the overemphasis on tests scores in the college admissions process. The accommodations offered to students with learning disabilities are made available in an inconsistent, and often random, manner. Both ACT and SAT tests use different measures for qualifying students for accommodations and both agencies' rulings are idiosyncratic and impossible to predict. Students with learning disabilities are commonly granted a time accommodation by one test agency, yet denied it by the other.
Testing agencies determine who "deserves" extra time -- and who doesn't because apparently they don't think that every student should be given the opportunity to express what they know. Even given that unfortunate reasoning, the agencies are deluding themselves if they think that everyone who is "deserving" of extra time is able to access the complex process of securing a time accommodation. There are a variety of reasons students who would benefit from extra time don't get tested, including well-meaning adults' resistance to labels and misunderstandings about how and why that kind of testing is needed. But most obstacles point to limited resources -- a school's or a family's. In private schools, the learning evaluations needed to access accommodations are rarely offered internally. Families must bear the burden of the steep financial costs of independent evaluations. Meanwhile, public schools offer free internal screening for learning disabilities, but students have to be failing in school to qualify.
According to William Hiss, the lead author of the study, standardized tests were initially utilized to discover untapped talent, otherwise obscured by remote geography, poor access to good schools, or other obstacles related to socio-economic inequities. Ironically today, research done by Stanford University and College Board has shown that standardized test scores rise with parental wealth. There are many potential explanations for how or why the tests have strayed from serving the original objective of discovering talent among the less privileged. While factors like childhood circumstances, race and parental education clearly play complex roles, many also point to the expensive and sophisticated test preparation and multiple retakes that elude lower income students. In any case, all biases seem reason enough to make optional reporting of test scores a standard choice in every college admissions process. If we genuinely want our colleges and universities to be rich and diverse meritocracies, we would do well to question and reform our overreliance on flawed standardized testing metrics that often obscure, rather than reveal, academic potential.