Why Can’t Our Certification Tests Measure What’s Really Important for Teachers to Know?
I recently sat for the four-hour Massachusetts Test for Educator Licensure (MTEL) in History. The test — required certification for history teachers in the state’s public schools — consisted of a hundred multiple-choice questions and two short essays on subjects ranging from Bronze Age inventions, to the nuances of Newton, to the ramifications of the Stamp Act, to the development of post-Cold War European nations. As I walked out of the testing cubicle, I was bleary eyed — and disappointed.
The state’s website proclaims that the test ensures that “educators can communicate adequately with students, parents/guardians, and other educators and that they are knowledgeable in the subject matter.” I agree that aspiring history teachers should have a broad, integrated, and thorough understanding of the world’s history. Yet a multiple choice test of discrete facts plucked from across a millennium strikes me as a poor ruler for measuring how well teachers can animate the past.
Most states have a variation on this test. Florida has the FTCE, Virginia the RVE, Ohio the OAE. Nearly two dozen states in total have created their own tests. Many other states choose to use or adapt the Praxis tests administered by ETS. From talking with teachers and comparing sample tests (I specifically compared history subject tests), many of these tests appear roughly equivalent, despite some different packaging.
But what does distinguish one test and one state from another is what is considered necessary to acquire a teaching license. There are no standard requirements across the 50 states. In some states, one need only take and pass a series of these tests to secure a five-year license to teach. In some states a passing grade on the same Praxis history test differs by up to 24 points in two different states. In other states, candidates need to amass courses in pedagogy and content, student teaching, and university recommendations.
Preparing for the grab-bag MTEL exam has made me think back to last year’s discussions at the Harvard Graduate School of Education about what should be required of new teachers and how different countries assess their quality.
In Finland, where teaching is considered by many to be as prestigious a career as one in medicine or law, there are no shortcuts into the profession. All wanna-be teachers must apply to the same rigorous degree system, and then spend five years attaining a master’s in education and in their core content area. (Roughly one out of 10 primary school teacher hopefuls are accepted into master's programs.) Rather than tests, each Finnish content teacher has completed a master’s in their field. Singapore has a similarly rigorous track.
Such systems might be unworkable in whole for the United States, given its size and its decentralization. But perhaps countries like Finland and Singapore have models and programs that can be adapted to help us better assess high-quality hopeful teachers.
The MTEL tests were written nearly two decades ago by educators trained in the 1970s. We teach now in classrooms where obscure facts can be readily accessed by a few taps on an iPhone. And while it is still important that our teachers have a strong grasp of historical facts, it is perhaps even more important that they are able to explain larger concepts, make connections across time, and connect with and motivate an increasingly diverse student body.
What if our entry tests challenged prospective teachers to do more than regurgitate facts? What if, in addition to correctly identifying historical dates and theories, these tests attempted to identify effective, inspirational instructors?
There have been some efforts to rethink these tests. EdTPA, a test developed by a team from a Stanford University and implemented with the American Association of Colleges for Teacher Education and Pearson Education, is one such attempt. Piloted in 2014, the test is now being used at different levels in 35 states, but there has been increasing backlash — about everything from the breadth and expanse of materials required, to the reports of unqualified test correctors, to the monetary incentives of Pearson.
What might a better test look like? The test I just took measured a fraction of my daily work. Here are a few ideas about how we could measure the rest:
- Instead of simply testing my own understanding of the 1929 stock market crash, what if the test challenged me to describe how I would explain the crash to sixth-graders or 12th-graders in a lesson plan? (Having recently taught the Great Depression, I’m all too aware how challenging it is to teach.)
- What if we had prospective teachers watch and analyze videos of real classrooms, identify the use of different techniques, and suggest areas where the teacher could improve?
- What if, to measure a teacher’s troubleshooting skills, we provided them with a series of scenarios where students had misunderstood key effects of World War I? We’d ask teachers to provide ideas for how to address each instance of student confusion.
- What if the test asked me to outline a month-long unit on major religions of the world? Not the particulars, but the arc of unit, with ideas for major projects and a list of essential questions to guide students’ study.
Such a test would obviously be much harder to grade, possibly more subjective. But with careful thought about the goals we want to reach, we might just create a test that could measure what really matters.
Get Usable Knowledge — Delivered
Our free monthly newsletter sends you tips, tools, and ideas from research and practice leaders at the Harvard Graduate School of Education. Sign up now.