Will A “Bar” Exam for Teachers Improve Student Performance?

Continued from previous page

Tougher tests will almost certainly result in many more failures among test takers. When Massachusetts instituted new licensing rules requiring prospective teachers to pass a new elementary math test, 75 percent of applicants failed in 2009, the first year the test was administered, according to Stotsky.

Many educators worry that the unintended consequences of a teacher’s “bar exam” could be to weed out many of the people they want in the classroom. “The only thing I'm certain a bar exam will do is keep out minorities,”says Chrystina Russell, the principal of a highly rated public middle school in Harlem who strives to make sure that her teaching staff resembles her black and Latino student body. Adds Nadelstern: “I don’t trust the testing industry to put together a test that is inclusive on the basis of gender, sexual persuasion or race.”

One reason that they see such an exam as disadvantaging people of color is that native speakers of Spanish and vernacular African-American dialects are at a disadvantage on written tests based on standard English because they often come from communities that are much more homogonous both ethnically and linguistically than those in which other immigrant groups live, according to Nadelstern, who adds: Speaking Yiddish didn’t get you very far if you grew up in an Irish, Italian and Jewish neighborhood.

The AFT counters that it has modeled its proposal on the so-called edTPA, formerly the Teacher Performance Assessment, which, instead of the prevailing multiple-choice  tests and written essays,  emphasizes a detailed “review of a teacher candidate's authentic teaching materials,” including video clips of instruction, lesson plans, student work samples and analyses of student learning. The aim of edTPA tests, which are subject-specific, is to evaluate how a prospective teacher actually teaches a given subject to real students, rather than her theoretical knowledge of teaching.

So far seven states, including Illinois, Washington and Minnesota, have adopted the edTPA, which was developed by Stanford University and the American Association of Colleges for Teacher Education; test administration and scoring are being coordinated by Pearson, the education company. New York State is adopting edTPA beginning in 2014.

The edTPA assessment is not without controversy. Last year, students at the University of Massachusetts at Amherst, one of 200 colleges in close to two dozen states where the new assessments were being piloted, refused to participate, saying the requirement that students submit 20-minute videos of their teaching practice was too onerous. They also questioned whether the test evaluators, who are being paid $75 to review each assessment, will put in the two to two-and-a-half hours necessary to do a thorough job. While Pearson helps to recruit scorers, Stanford notes that the scorers are experienced teachers who have been specially trained to evaluate the assessments.

Taking the edTPA will cost prospective teachers $300, about double what licensure tests cost today, but still far less than other licensing tests, such as those for nursing or the law.

Such controversy may be why the AFT, rather than adopting the edTPA outright, has asked The National Board for Professional Teaching Standards, which established the standards for so-called master teachers, to convene a gathering of educators and other “stakeholders” to design a multi-dimensional entry bar that could well include edTPA, minimum GPAs for teacher-prep candidates, as well as new student-teaching requirements.

With many states already signed up for edTPA, there is a danger that the AFT’s  “bar exam” could result in yet another addition to the national licensure smorgasbord — rather than a national consensus.

“There is a tension here with the AFT’s desire to get to a single national system,” concedes Linda Darling-Hammond, the Charles E. Ducommun Professor of Education at Stanford Univ., and a contributor to edTPA’s development. “How to get there is not immediately clear. We already have a patchwork. It’s conceivable that you could end up with more than one addition to the current schema.”