How to make standardized tests more useful for teachers

In the past few weeks, students across Colorado have pushed back on the rollout of the state’s new set of science and social studies exams called Colorado Measures of Academic Success (CMAS). Some seniors opted out of taking the exams, and they took to social media and the streets to argue their cause. They aren’t alone — parents, educators, and other community members have also complained.

But so much of the discourse around CMAS is focused on issues that, quite frankly, won’t move the conversation forward. The situation is so polarized that finding a realistic solution is almost impossible. In my opinion, we should focus on two questions: when should we assess students and how do we make the tests useful for teachers?

We must first recognize and acknowledge that we live in a society that demands accountability. Taxpayers and society in general, want to know if our schools are effective. So let’s start the conversation with the question, “How much testing is necessary to check the effectiveness of our schools?”

I agree with Mark Tucker, a national expert on public education policy. His proposal is to test all students in kindergarten, fourth, eighth and tenth grade, with random testing of second and sixth graders.

Why? Kindergarten is to make sure students are ready to move on (and yes, we can and do test kindergartners). Fourth grade is when students should be able to decode text, eighth grade is when students should be ready for high school, and tenth grade is when students should be ready for a community college program, with two years left to catch up.

Right now, we test all students from third grade until graduation. That’s too much. Think of this in sports terms: athletes have plenty of practice between games to make mistakes and correct them without worrying about their overall win/loss record. Districts and schools should utilize testing throughout a student’s academic career, especially with struggling students, to monitor their progress. But these tests should not be the high stakes exams used by the state to evaluate schools.  More practice, with fewer games.

Even with pared back exam schedules, the testing process should be more transparent and responsive.

The current state exams work primarily as autopsies, rather than checkups. State exam data is released months after the exam has been administered, much too late to be useful as a tool for the educator. Every attempt needs to be made to get teachers information on how their students are doing and address any deficiencies before students leave our classrooms.

The switch to the fall testing window from spring does little to address this if results are not quickly disseminated. In high school, many courses are semester long. By the time results are back, the student has moved on.

Even when the data is released, it’s too vague to help. Useful test results impact how and when I teach to a standard. They also give me specific information on individual students that I can use to check in on their progress in my class.

Right now, I don’t know how my students do on individual questions or how the questions addressed the standard. To help the data inform my practice as a teacher, it is not enough for me to know what standard the question addressed. I need to know what the question asked of the student and how the question was asked. This would allow me to pinpoint what changes I need to make in how I teach the material.

To use the standardized tests, I have to trust them. The onus to that build trust rests on the testing companies. Teachers should be involved in writing the questions and they need to release the actual test questions. I realize this is a difficult demand. Releasing test items is expensive, since every question made public would need to be replaced. In addition, many testing companies also claim intellectual rights to the questions. But the Colorado State Department of Education can write contracts with testing companies that require these companies to release exam items and to require them to involve practitioners in writing these exams.

If we were to bring more transparency and timeliness to the exams, teachers would be more comfortable with using the results of the exams as a part of their overall evaluation. I am willing to be held accountable to student achievement results as long as I have trust in the exams. I know we still have work to do to ensure student data can be used to evaluate teachers, but this would be a move in the right direction.

About our First Person series:

First Person is where Chalkbeat features personal essays by educators, students, parents, and others trying to improve public education. Read our submission guidelines here.