October 26, 2015 | By Kelly Zunkiewicz
Isabel is a senior at Lennard High School in Hillsborough County Public Schools, and a member of my AP Calculus AB course. In the process of writing her a recommendation letter, I reviewed her student summary: a list of her scores on all the district, state, and national exams she’s taken in her first three years of high school.
I counted 63 exams. That’s 40 district exams, six community college exams, nine state level exams, six national exams, two AP exams… and a partridge in a pear tree.
She’s not alone. A study released over the weekend by the Council of the Great City Schools showed that students in big-city school districts will take roughly 112 standardized tests by the time they graduate from high school.
I’m a math teacher. I love data—and I need it to teach well. Tests can and should be a key way for me to find out whether my students are actually learning what I’m trying to teach them. But I think we can all agree that the number of tests we’re talking about here is too high, and I’m glad the Obama administration has pledged to tackle the problem. I disagree, however, with the administration’s proposal that we cap testing at two percent of instructional time. If we’re serious about fixing testing, we can’t start by focusing on a particular quota for how much time we’re allowed to spend on it. Instead, I’m much more concerned with figuring out which tests are actually useful to us, and which ones aren’t—and going from there.
Let me explain what I mean. Formative and summative assessments occur every day in my classroom. They provide concrete data on my students’ performance, and my own. I complete an item analysis on quizzes and tests to identify trends in student performance by standard, and adjust my instruction based on what my students are learning.
And when my students receive a graded test, they immediately join their classmates in making corrections—learning from their mistakes and filling in the gaps in their knowledge. The learning experience of an assessment is what makes it useful to me and my students, and it’s what I want them to focus on even more than the grade they’ve earned.
The problem is, far too few standardized tests provide this critical learning experience for teachers and students. Of the 63 assessments Isabel has taken, less than ten have provided her or her teachers with anything other than a final grade. And results from many of them weren’t available for weeks or even months—too late to have any real impact on teaching and learning. In other words, most of these tests were not time well spent.
How can we ensure that the tests we give our students support their learning process, rather than take time away from it? Sure, some of Isabel’s 63 exams might be duplicative, or unnecessary. We can and should get rid of those. But it’s also possible that some of those tests could be useful if we had access to better and more timely data from them.
That’s particularly true for tests required by individual school districts, which represent the biggest chunk of tests Isabel and other students take (even though state- and federally-mandated tests get most of the attention). Right now, many of these district-level exams only tell us how many students missed each question—not which individual students, which means I can’t use that data to provide truly differentiated instruction. And while some of these tests do collect data that would help me target instruction to meet students’ needs, the process for accessing that data is incredibly convoluted and time-consuming—far beyond what even the most dedicated (and tech-savvy) teachers can manage in their busy schedules.
Teachers need to be encouraged to use data to inform their instruction, which means useful data must be readily accessible to us. The technology exists to give us near-immediate results and opportunities to personalize each student’s learning plan. In particular, teachers need score reports with breakdowns of students’ performance by individual standards, and analytical support to be able to interpret the data they’re seeing. If Isabel did poorly on, say, MA.912.C.1.12—“understand and use the Intermediate Value Theorem on a function over a closed interval”—I need to know right away, so I can discuss the concept with her and provide resources to improve her knowledge. This data can also help students embrace assessments as an integral part of the learning process.
As the Obama administration and others look to improve standardized testing, I hope they’ll do more than just search for a magic number of tests that’s “just right.” Instead, let’s figure out which tests are actually useful—or could be, given the right changes—and get rid of those that aren’t doing the job.
Kelly Zunkiewicz is a 2014 Fishman Prize winner and a member of the TNTP Educator Editorial Board.
To read this blog post on the TNTP Blog website, click here.