Testing Should Help

A TED talk from April 2013 by “creativity expert” Ken Robinson—who has been criticized of late for being idealistic and clueless—makes a really simple point: tests should be diagnostic. He states, “Standardized tests have a place. But they should not be the dominant culture of education. They should be diagnostic. They should help.”

At this writing, most students in Wake county have recently undergone the annual ritual of high-stakes testing, during which time virtually no other learning could take place at each school.

There’s plenty to despise about the state-mandated standardized testing that occurs every May. Here are just a few of the common criticisms that angry parents have piled on in recent years:

  • Testing wastes an inordinate amount of instructional time, as teachers are compelled to provide practice tests and to teach students test-taking skills throughout the school year. In addition, children seem to learn practically nothing after they’ve completed their testing regimen, even though they typically have an additional week or two of school remaining.
  • Preparing for tests also wastes an enormous amount of teachers’ time, as they attend required training courses, fill out paperwork, and clear their schedules to administer these tests.
  • The tests create a great deal of stress among students starting in the third grade.
  • They create a great deal of stress among teachers and cause many to hate their jobs, or even to leave the profession entirely.
  • Because the main subjects that are tested are math, reading, and science, other subjects are often starved of resources. And these other subjects might be the ones that bring joy to individual students, the ones that motivate them to come to school each day.
  • The tests create a financial windfall for a few testing providers, whose behavior sometimes resembles an organized crime network.
  • The results of testing do not seem to accurately reflect the reality of the experience at a given school.

In another TED talk, famed education reformer Geoffrey Canada points out that by the time students take their high-stakes, state-mandated end-of-grade tests, they’re already done with those subjects and might not ever study that material again. Canada explains the problem from the perspective of an educator:

I’m a tester guy. I believe you need data, you need information, because you work at something, you think it’s working, and you find out it’s not working. . . . But here’s the problem with testing. The testing that we do … is in April. You know when we’re going to get the results back? Maybe July, maybe June. And the results have great data. They’ll tell you Raheem really struggled, couldn’t do two-digit multiplication – so, great data, but you’re getting it back after school is over. And so, what do you do? You go on vacation. You come back from vacation. Now you’ve got all of this test data from last year. You don’t look at it. Why would you look at it? You’re going to go and teach this year. So how much money did we just spend on all of that? Billions and billions of dollars for data that it’s too late to use. I need that data in September. I need that data in November. I need to know you’re struggling, and I need to know whether or not what I did corrected that. I need to know that this week. I don’t need to know that at the end of the year when it’s too late.

Unlike the high-stakes end-of-grade and end-of-course tests that our students actually take, a targeted series of shorter tests would identify students’ areas of weakness. Useful, diagnostic tests would enable administrators to aim additional resources at building students up in those very areas. A whole raft of research suggests that testing can be a powerful teaching tool if deployed effectively—and not just your standard “diagnostic” testing, either. An Education Week article by Wendi Pillars encourages teachers to give tests that ask students to think about a topic well before they are ready to pass a comprehensive exam on that topic: “The mere act of guessing engages your mind in a different and more demanding way than straight memorization or being fed answers. . . . Somewhat counterintuitively, doing this regularly also sets up your classroom community for understanding that the value of knowledge lies in the processes of learning, rather than just the outcome.”

Using tests to help students become more efficient learners is an idea that makes plenty of sense.

Pillars urges the use of additional testing whose aim is to help consolidate learning. However, many, if not most, K-12 teachers already feel overburdened with responsibilities for administering tests and submitting grades in a timely manner. The pressure is on to keep gradebooks up to date for those eager parents and students who live and die by up-to-the minute grades and GPAs they can check in PowerSchool. And additional testing leads inevitably to more opportunities for failure.

Pillars advocates for an approach that helps students grow accustomed to periodic failure, an unpopular notion among today’s children and their parents. She writes, “Providing your students with a unit pretest they’re destined to fail is actually beneficial for learning.” And perhaps she’s correct, but today’s students—who feel ever more pressure to produce stellar transcripts to accompany their college and scholarship applications—are not energized, but rather demoralized by frequent failures in the classroom. Even if it promoted the much-lauded character trait of “grit” or “perseverance,” a shift to “failure-based” education would surely be challenging for our underpaid NC teachers.

And yet, well-timed and thoughtfully crafted tests certainly aid learning for most students. A thoughtfully constructed exam encourages students to consolidate and synthesize recently acquired knowledge and ideas. Pillars advocates for an approach that requires students’ “effortful retrieval” of knowledge and that “promote[s] active repetition of information.” She points out, “Better recall can occur through the use of testing!”

Studies of learning and retention back her up. Writing about psychological studies that showed a retention gain from “pre-testing,” Benedict Carey claims, “wrong answers on a pre-test aren’t merely useless guesses. Rather, the attempts themselves change how we think about and store the information contained in the questions.” He explains the phenomenon of improved retention through pretesting, or testing students’ knowledge of a subject before they’ve even studied it, as a “basic insight” that “[t]esting might be the key to studying, rather than the other way around.” He continues, “a test is not only a measurement tool. It’s a way of enriching and altering memory.”

Another concept at which Canada expresses bafflement is the three-month summer break, which is anathema to recall. The long vacation period has been termed nonsensical by multiple education researchers, and when you place it in the context of high-stakes testing, as he does, the schedule makes even less sense. In Canada’s view, with end-of-grade tests administered in May, you potentially identify all of the areas in which a student failed to master the required material. And then the school year ends, and you allow that student to forget even the aspects of the subject that s/he actually did grasp during the school year.

With such a long break, and no ability to view the test results beyond a score that is mailed to your house a couple of months after your child completes the tests, there’s no opportunity for even the self-motivated student to perform additional work in the identified areas of weakness. No test can be effective if students are never allowed to see the questions that they missed! It’s just like reading a murder mystery and finding that the last chapter is missing, so that you never find out who committed the crime. It surely feels like a huge waste of time to both teachers and students … and it is.

In an ideal world, we would replace end-of-grade tests in grades 3-12 with shorter tests that came earlier in the year, tests whose results could be discussed in class. Teachers would kick off the school year with short diagnostic tests in each subject—tests that include some of the material in the curriculum that has not yet been covered, and some material that the student should already know from the previous year. (Many teachers probably do this already.) They would go over the results and provide the correct answers with the class during a friendly, non-punitive lesson, and find ways to engage the students in finding out more about the correct answers as homework.

While simultaneously using the results of these tests to target additional resources at students whose test results indicate that they need them, the teacher could then frame the lessons that follow as attempts to explore the questions that the class was unable to answer on the test.

A more comprehensive exam taken a few weeks later could include the questions from the previous diagnostic test and build on them with more probing questions that encourage synthesis of the knowledge that has been gained. As Pillars suggests, teachers could then space out quizzes throughout each semester in order to consolidate the gains already made. The point to keep in mind throughout the year is that testing needs to be both diagnostic and helpful. A quiz that simply “proves you did the homework” is not a helpful tool; a test that asks you to consider a different aspect of material that you are covering this month can be a very helpful tool.

But we can’t just eliminate standardized testing. We still have to monitor student academic progress, performance gaps among different groups of students, and teacher effectiveness. And we have to compare academic progress among schools so that we can ensure that every student is getting a sound education. Standardized tests could provide useful data for such evaluations if their many flaws could be addressed.

Without eliminating state-mandated end-of-grade tests, perhaps teachers could be given tools to analyze and use test results to help students in the areas where they have weaknesses. The “end-of-grade” state-mandated tests could be administered earlier in the school year, reserving time to diagnose student weaknesses and then provide extra instructional time in the weakest areas of the curriculum. (Obviously, the tests would need to be graded swiftly—and we have the necessary technology to do that!)

The benefits of such an approach would include the following:

  • Parents would have useful information on which to base interventions, either
    immediately, or during their child’s summer break. If I see that my child is below grade level in reading, of course I want to enroll her in a reading course, or help
    her to practice at home.
  • Less instructional time at the end of the school year would be wasted.
  • The final month or so of the school year could be spent on differentiated learning: assisting students who are below grade level to catch up in specific areas and pushing students who are at or above grade level to achieve at a higher level.
    This portion of the school year would be a good time to bring in those parent volunteers if the state legislature won’t provide enough money for qualified teacher assistants to help with the differentiated instructional time.
  • With less build-up to the tests, less of a sense that the tests were utterly comprehensive, and less emphasis on their effects on promotion to the next grade, students would probably feel less worried about them.

Further, even if standardized tests were administered two months before the end of the school year, their results would still be very useful for deriving student growth measures (important components of each school’s annual School Performance Grade), comparing the performance of different groups of students, and gauging teacher effectiveness.

Something to think about.