Groups Honing Real-Time Teacher-Performance Exam
Tests would simulate classroom situations
Prospective doctors take a licensing exam in which they diagnose the ailments of mock patients. Airline pilots dodge wind shear in flight simulators. Even 16-year-olds follow a driving examiner's directions—turn right, parallel park—to earn their coveted driver's licenses.
Should a similar licensing standard apply to prospective K-12 teachers?
That's the thinking behind a new suite of assessments under development by TeachingWorks, a nonprofit launched by University of Michigan education faculty members, and the Educational Testing Service, the purveyor of the venerable Praxis suite of certification tests.
Within a few years, the groups hope to release a new exam capable of measuring teacher-candidates' ability to execute key aspects of instruction, such as leading a class discussion and asking probing questions to gauge students' understanding.
As currently envisioned, the exam would rely partly on simulations in which a prospective teacher would interact with actors portraying bewildered parents, 6th graders struggling with fractions, or a rowdy just-back-from-recess class.
The goal is for states to use the exam to supplement or even replace current licensing exams and to ensure candidates enter classrooms with a basic threshold of acceptable skill.
"We should debate less about entry standards to teacher education, and we should agree more on what the exit standards are," said Deborah Ball, the dean of the University of Michigan's school of education and the director of TeachingWorks.
The TeachingWorks and ETS project comes as interest in measuring and improving teachers' hands-on performance is growing.
Already, a number of performance-based licensing exams exist, notably Stanford University's edTPA and the ETS' own Praxis Performance Assessment for Teachers.
Those two exams require a teacher-candidate to assemble a portfolio over several months demonstrating his or her ability to teach a lesson and tailor it for specific students, and include a videotaped segment of each candidate's instruction.
By contrast, the new exam, tentatively called the National Observational Teaching Exam, or NOTE, would try to gauge a candidate's teaching skills in real time—and with limited time to prepare.
"The idea is that you're asked to do something specified at a given moment of time, not give an example of your very best practice," Ms. Ball said.
The exam will be based on Ms. Ball's work to codify a list of practices all novice teachers should master and be able to demonstrate before taking charge of their own classrooms. These so-called "high leverage" practices have become the backbone of the University of Michigan's teacher-preparation programs.
TeachingWorks, a separate body set up by Ms. Ball, consults with education programs to research, refine, and spread the practices.
From a list of 19 core practices, TeachingWorks and the ETS have tentatively settled on four that will serve as the basis for the exam. Specifically, the exam will gauge a teacher's ability to:
• Conduct a classroom discussion;
• Teach a specific concept, such as dividing fractions, through modeling and examples;
• Elicit and interpret a student's thinking and tailor instruction in response; and,
• Meet constructively with a parent or guardian to discuss a child's progress or behavior.
Controlling for Variables
With TeachingWorks providing the curricular basis for the exam, the ETS' role is to help translate the skills into a testing context.
It is conceptually a difficult task. Indeed, with up to 30 students in a classroom, teaching contains potentially even more variables than driving, flying a plane, or removing tonsils. But ETS officials say technology has made it increasingly possible to control for those variations.
For example, the teams are considering using avatars—"virtual" students represented on a computer screen and controlled by trained actors—to approximate a classroom environment. One such example already exists at the University of Central Florida, in Orlando; its simulated classroom, populated by five mimed students with distinct personalities, has been used by a number of teacher-preparation programs.
To ensure standardization of the exam, the organizations envision giving the prospective teachers a premade lesson plan and making sure that they can carry it out, rather than having them devise the content on their own, as is the case for existing performance exams.
Cost could pose a challenge to the endeavor, because performance-based exams tend to be more expensive and longer than other kinds. ETS officials declined to put a dollar estimate on the exam, though they acknowledged that it will have to be considered in the context of novice teachers' low salaries.
"The practical part is how you make the exam short enough that people will take it, and at a price that people will be able to afford it," said George Powell, the vice president of teacher licensure and certification programs for the Princeton, N.J.-based ETS. "We're talking about some pretty sophisticated ways to measure candidates' abilities."
It's also far too early to know whether states and education programs would embrace the novel format. Already, some have been wrestling with the policy implications of portfolio-based tests like the edTPA.
But the project appears to be a go: Prototypes are already being tried out in a handful of teacher-preparation programs across 10 states.
By 2015, the groups plan to begin larger pilots and to have parts of a test for elementary teacher candidates ready for states to adopt by late 2016.
Vol. 34, Issue 08, Page 7Published in Print: October 15, 2014, as Groups Honing Real-Time Teacher-Performance Exam