Opinion: Teacher tests can right past wrongs

Emery Petchauer
The Michigan State Board of Education meets in Lansing in  May of 2016.

In 2013, the State Board of Education attempted to increase the readiness of new teachers by revising the exams they have to pass in order to earn a license. The results were a disaster. In the first round of testing, pass rates for the history exam fell from 71 percent to 25 percent. Social studies fell from 65 percent to 22 percent. Elementary education fell from 83 percent to 49 percent.

The steepest drop was on the new Professional Readiness Exam that admitted students into teacher education program. Pass rates fell from 82 percent to 22 percent. These pass rates led to enrollment drops in teacher education programs across the state. Some programs have yet to recover from this sudden shock to the system.

Soon, the state will have an opportunity to right these wrongs. As part of the Department of Education’s “students first” initiative, teacher education programs in Michigan have begun a deep process of redesign that will change the grade levels that educators are certified to teach, clinical hours required, and more. After this process, the state will be in a position to rewrite these exams once again.

Avoiding the mistakes of 2013 will mean resisting some of the conventional logic about tests and teachers.

First, we must resist the logic that says pass rates drop because teachers today are less prepared compared to years past. This is wrong. A sudden drop in scores like we saw in 2013 happens when exams are hastily adopted or not properly piloted. This sudden drop also happens when the exams are made by a company with little connection to the deep work of teaching.

Second, conventional logic would say that tougher tests make better teachers. Doesn’t a high tide raise all boats? A series of research studies from Dan Goldhaber at the University of Washington shows a more complex picture. Yes, there are instances where teachers with higher exam scores produce greater gains in student learning. But the evidence of this “student learning” is not related to the real-world problems we want and need students solving. The evidence is doing well on standardized tests, a very narrow gauge of learning.

Third, we must resist logic that says these exams are a fair and uniform measure. Again, research from Goldhaber is helpful here. The accuracy of these teacher exams changes based upon the race and gender of the teacher taking it. This is like a scale that gives a different reading depending on a person’s eye color. What’s more, other factors beyond a teacher’s test score can be more influential. In some instances, a racial match between African American students and their teacher has a greater impact on student learning than how well the teacher scored on their exam.

When the state of Michigan rewrites its teacher exams, it will be crucial that the exams go through an extended pilot period before they have consequences on aspiring teachers and programs. This pilot period must include multiple test administrations with diverse pools of prospective teachers. If these exams do not predict the deep learning we desire from students -- not just their standardized test score -- the exams should be thrown out. When the state does reach a point where it requires these new exams, the results should never trump more direct indicators of teacher quality. Instead, the exams should be one of many indicators that shapes how the state supports and licenses teachers.

Children and families deserve teachers who will bring out the very best of their educational potential. They also deserve a process that will bring these teachers into their classroom.

Emery Petchauer is associate professor in the College of Education and College of Arts and Letters at Michigan State University and author of "Navigating Teacher Licensure Exams: Success and Self-Discovery on the High-Stakes Path to the Classroom."