69传媒

Education programs across 19 states are piloting a performance-based assessment for teacher-candidates that potentially could serve as a common prelicensing measure for new teachers.

Based on a test in use in about 30 education schools in California, the includes a 鈥渢eaching event鈥� requiring teachers to extensively document and submit for review artifacts of their planning, instruction, and ability to assess and respond to student needs.

Five of the states taking part in the work鈥擬assachusetts, Minnesota, Ohio, Tennessee, and Washington鈥攈ave committed in legislation to use a performance-based licensing test, and officials have signed memoranda of understanding agreeing to adopt the assessment if it proves to be technically valid and reliable.

Supporters of the initiative see in the work an opportunity to focus on classroom-based effectiveness at the precertification benchmark鈥攁n area that has not received much attention as policymakers tackle the tenure-granting and annual evaluation processes.

鈥淚magine a time two years from now when we can show people, predictably and validly through standardized instrumentation, a teacher who鈥檚 ready for a placement,鈥� said Peter J. McWalters, a consultant for the Council of Chief State School Officers, a project partner.

Classroom Performance

States鈥� current licensing systems require a dizzying array of standardized tests, most of which gauge content knowledge and pedagogical skills through multiple-choice questions. Only Arkansas, California, and Ohio now use instruments that judge prospective teachers鈥� actual classroom performance.

That landscape is poised to change under the Teacher Performance Assessment Consortium, a group of 19 states working with the CCSSO; the American Association of Colleges for Teacher Education, or AACTE, a Washington-based membership organization; and the Stanford Center for Assessment Learning and Equity, based at Stanford University, to develop an exam embedded in preservice teachers鈥� student-teaching experiences.

State members of the consortium include California, Colorado, Iowa, Illinois, Maryland, Michigan, Missouri, New York, New Jersey, North Carolina, South Carolina, Virginia, West Virginia, and Wisconsin, in addition to the five 鈥渁ccelerated鈥� states that have signed memos of understanding.

The centerpiece of the new exam consists of a common 鈥渃apstone teaching event鈥� that all preservice teacher-candidates must complete. The event recalls aspects of the certification process for the National Board for Professional Teaching Standards.

Over a three- to five-day period, a teacher must plan and implement an instructional unit, use assessment data to tailor instruction to students who have difficulty mastering the material, and reflect on his or her experience. A 15-minute videotape of the teacher鈥檚 instruction and other artifacts are scored according to a common guide by university professors and, potentially, other stakeholders, such as deans, school district officials, or master teachers.

To reflect the diversity of offerings in the teacher education world, each institution may customize three other assessments in specialty areas such as technology or the instruction of special populations. The results will be used to provide teacher-candidates with feedback on their performance before the capstone teaching event.

Common Definition

In addition to serving as an initial licensing requirement and providing information that districts can use when making hiring decisions, the exam could help preparation programs improve.

鈥淥nce you鈥檙e finished doing it with a group of students, you get information about where students are encountering weaknesses in the program, and it actually becomes a part of what the faculty grapples with at the end of the year,鈥� said Pat A. Wasley, the dean of the college of education at the University of Washington, in Seattle. 鈥淲e just haven鈥檛 had that kind of information before.鈥�

When multiple stakeholders take part in the scoring, she added, 鈥測ou can learn collaboratively about how you commonly define high-quality teaching, and at the same time, you can argue about what should be of greatest value.鈥�

Plans include linking the exam to professional certification and setting up a national repository of the videotaped teaching events that could be integrated into the preservice curriculum, said Sharon P. Robinson, AACTE鈥檚 president.

鈥淲e will have a national database of productivity measures in programs that are developing the education workforce,鈥� she said.

If many states ultimately adopt the exam, they could compare information on their graduates, potentially easing issues with license reciprocity, Ms. Wasley said.

Questions of Use

The exam is being piloted in at least three institutions in each of the member states. A pilot year based on a full state sample will begin this fall, and a large-scale field trial will take place in the 2011-12 school year. Among other actions, researchers will examine the pilot data for the relationship of scores to other teacher tests, and to determine whether scores are predictive of student achievement during candidates鈥� first two years of teaching, said Raymond L. Pecheone, the director of the Stanford center.

Experts said that the consortium鈥檚 success likely will depend as much on the policies that surround the assessment鈥檚 implementation as it does on its technical quality.

Project partners are still mulling over several aspects of the exam that, if scaled up, might pose challenges.

Scoring may have to be changed. Mr. Pecheone envisions a more controlled, possibly regionally based scoring and auditing system to ensure fair and comparable results.

Other experts harbor philosophical concerns. Daniel Goldhaber, a researcher at the Center for Reinventing Public Education at the University of Washington in Bothell, drew a distinction between policies that seek to use a licensing exam to screen out teachers who appear to be less effective and those that use such exams as a mechanism for improving effectiveness.

In the past, most teacher exams have been shown to be somewhat predictive of classroom performance, but there is less evidence to show that they actually help cultivate better teachers, he said.

鈥淧olicymakers don鈥檛 generally get this distinction,鈥� Mr. Goldhaber said. 鈥淚f there isn鈥檛 a human-capital value associated with [a performance-based test], it鈥檚 a very expensive assessment to use to screen out such a small percentage of the workforce.鈥�

For Edward Crowe, a senior adviser to the Princeton, N.J.-based Woodrow Wilson National Fellowship Foundation who wrote on teacher-training program accountability, the test鈥檚 mettle will also depend on whether states and preparation programs are willing to use its results to make tough decisions about teacher and program quality.

鈥淚f [the assessment] doesn鈥檛 support accountability for individual teachers, if they鈥檙e going to report 96 percent pass rates, and if weak programs in these states are not forced out of business,鈥� Mr. Crowe said, 鈥渢hen I would say we haven鈥檛 gotten anything better from this.鈥�