School districts wrestle with effective ways to judge instructors' performance
By Erin Richards of the Journal Sentinel
Nov. 14, 2010
Norma Mortimer moves about her high school classroom with confidence born of 41 years' experience.
Directions to students are clear; she knows when to push for an answer and when to let a question hang.
The English teacher formerly taught music, composed and arranged marching band music, and performed at the Bristol Renaissance Faire.
"It all adds into what I bring to the classroom," she said.
Once every three years as a tenured teacher, performance evaluations provide her with feedback, something she looks forward to even though she knows she's not slipping.
Still, evaluations never flag what she considers her weakest area - teaching effectively when the class is in small groups. Last year, she never received her post-evaluation conference with the principal.
In the growing national debate on how to raise the quality of public school teaching in America, performance evaluations have become both a lightning rod and a sticking point.
Most evaluation systems in public schools provide little information to properly assess teachers' strengths and weaknesses. And because teachers are rarely dismissed over their performance, formal evaluations seldom carry much weight.
In the push to create more great teachers and raise student achievement, President Barack Obama's administration has championed stronger performance evaluations, tied to student test scores, and bonus pay systems to better place, support and reward excellent educators. The administration also wants school districts to justify the dismissal of those who are ineffective.
There's wide agreement that current evaluations don't accomplish much; there's little agreement between reformers and unions on how best to improve them.
In Milwaukee Public Schools, veteran teachers complain about being asked to sign off on evaluations when their principals never observed them. Principals can be overwhelmed with so many teachers to evaluate, and nonacademic matters to worry about, that it's difficult to do fair and complete analyses.
Mortimer, who teaches at Samuel Morse-John Marshall School for the Gifted and Talented, was rated exemplary in her last performance evaluation. MPS rates teachers as either unsatisfactory, in need of improvement, satisfactory or exemplary. Last year, more than 95% of 2,424 teachers evaluated received favorable reviews.
The number rated unsatisfactory in the state's lowest performing district: five.
A national discussion
Across the country, leaders in both the education and political arenas are increasingly focused on strengthening teacher evaluations by basing them in part on student test scores. Critics argue that tests - even ones that measure growth from the beginning to the end of the year - don't reveal everything about teacher quality.
"We're seeing different models now (for teacher evaluations) partly because we know that evaluations are generally bad and that teaching really matters," said Elena Silva, research director at Education Sector, a nonpartisan Washington, D.C., education policy think tank.
The urgency for change is rooted in research. Studies show the most important school-based factor determining student success is the quality of the child's teacher, and that low-performing students benefit from highly effective teachers even more.
Eric Hanushek, a Stanford University education and economics researcher, has popularized studying teacher effectiveness by looking at how much academic growth teachers' students make over the course of a year. As early as 1992, based on results of his research in Gary Ind., Hanushek reported "the difference in student performance in a single academic year from having a good as opposed to a bad teacher can be more than one full year of standardized achievement."
Critics of using test scores to evaluate teachers, or value-added measurements that isolate the impact of a teacher's effect on student learning, often worry that too much emphasis on scores may overlook the intangibles that may contribute to good teaching, such as motivation and presentation skills. Also, test scores may not pick up other nuances of the school environment.
Still, student achievement data is the best information researchers have to measure gains in performance, the ultimate goal.
Sabrina Laine, director of the National Comprehensive Center for Teacher Quality, said no state has yet perfected a model evaluation system. But efforts are under way, including:
• Tennessee and Delaware, which were rewarded in the first round of the Obama administration's $4 billion Race to the Top competition for offering teacher merit pay and factoring student test scores into evaluations. The competition offered millions of dollars in federal schools money to states with the best plans to improve education.
• Florida, Georgia, Hawaii, Maryland, Massachusetts, New York, North Carolina, Ohio, Rhode Island and Washington, D.C. - the winners in the second round of Race to the Top - all have instituted rewards for effective teachers and principals.
• Louisiana, where a new law requires half of a teacher's evaluation by 2012-'13 to be based on growth in student test scores.
• Colorado, which passed legislation this year that ties teacher and principal evaluations to student growth and makes it tougher for teachers to get tenure.
• Thirteen states with districts implementing TAP: The System for Teacher and Student Advancement, a model that defines and measures effective teaching, links evaluations to professional development, and allows teachers to earn bonuses based on growth in student test scores.
Wisconsin, for the most part, has been on the sidelines of the reform movement - and it has cost the state.
During the first phase of Race to the Top, U.S. Education Secretary Arne Duncan singled out as "ridiculous" a provision in Wisconsin state law that banned the use of student assessment data in teacher performance evaluations. According to the rules, states with such a ban would be disqualified from the competition.
Wisconsin lawmakers subsequently altered the language and dropped the firewall. But the new language specified that any changes to teacher evaluation plans would need to be bargained with the teachers unions.
Rep. Brett Davis (R-Oregon) said the watered-down legislation eliminated one firewall and added 426 more - one for every district in Wisconsin that would have to negotiate changes to evaluations.
Still, Deborah Mahaffey, assistant state superintendent for academic excellence, says the prospects for change in Wisconsin are good.
"We all understand that the move is toward teacher effectiveness, and that there will be some measure of student achievement in determining teacher effectiveness," she said.
Obstacles to fair systems
Creating more fair, comprehensive and useful evaluation systems faces roadblocks from the start.
First, involved parties have to agree on a definition of quality teaching. Then they have to decide what factors to use to measure it.
If student achievement is to be a factor in the evaluation, what tests or student work should be considered? How much of that should count toward the total evaluation score?
"Test scores can only be used for certain teachers because not all teachers teach in tested subjects and grades," Silva said. "You get into this debate where people say, let's use test scores for 42% of the evaluation, or 62% of the evaluation. There's no real rationale for any particular percentage, whether it's half, 51% or 49%."
In Wisconsin, state law requires districts to conduct a personnel evaluation in the first year of employment and at least one every three years after that, Mahaffey said. Individual districts are left to determine the characteristics of their formal teacher evaluations.
Most are based primarily on principal observations, and practices vary.
Some go further than others. In the Cedarburg School District, principals submit three different assessments: a standards rating for each teacher based on how well he or she implements the state's 10 teaching standards, another score based on education consultant Charlotte Danielson's Framework for Teaching and a narrative on the teacher's performance.
Superintendent Daryl Herrick and his administrative staff read all the evaluations and require principals to defend the rankings. Principals are evaluated on how well they evaluate their teachers.
"The process serves as a clearinghouse for performance concerns," Herrick said. "Mediocrity has become a tough challenge in schools."
Herrick said he doesn't have scientific proof that changes to the evaluation system have caused better academic results, but achievement in the district based on ACT and SAT scores, as well as the annual state tests and district assessments, has been at its highest in recent years.
"That was about when we started to focus much more on teacher quality," Herrick said.
In the Hamilton School District, teachers may undertake a voluntary research project as part of their evaluation. The project allows them to experiment with different teaching techniques in class, under the guidance of their principal, and then reflect on their research as an addendum to their evaluation.
In MPS, principals use a formal evaluation tool that outlines 11 teacher expectations, such as knowledge, classroom management and attitude. MPS is working with the Milwaukee Teachers' Education Association to revise that tool and provide better information about teacher quality, Human Resources Director Karen Jackson said.
"The motivation up to this point in MPS has not been about how we have a professional dialogue about how to improve teaching and learning," said Dan Lotesto, a teacher at Rufus King High School who has been on both the teaching and administrative side of the evaluation process.
Sid Hatch, deputy executive director of the MTEA, said the union was obligated to work with the district and create a new evaluation system for schools receiving federal grant money for improvements. But the evaluation committee realized the system should be taken districtwide to be effective, he said.
Hatch said the evaluation committee is "kicking around" the idea of using growth in student test scores as a component of the evaluation.
"There are some things you can't measure in teaching," Hatch said. "Student performance cannot be the sole factor in teacher evaluations or I'm not interested in putting my name on it."
For Mortimer, who has taught in four districts in three states, the issues surrounding teacher quality and performance haven't changed much over the years. If she could wave a magic wand, the evaluation process would be conducted by principals who spend lots of time in classrooms, have experience as high-quality teachers themselves and offer ongoing, constructive feedback.
"I feel passionately that our kids need teachers who want to be in the classroom, who want to be with kids and all their issues, foibles and delights," she said.
But will desire still be a measure of teacher effectiveness in the new world of education reform? Some teachers are great with kids but can't move the needle on test scores. Others may be unimaginative but adept at raising proficiency levels.
Silva, from Education Sector, said she believes stronger evaluation systems will do a lot to better distinguish between educators, and to improve the quality of teaching.
"Principals will tell you, 'I know the good teachers,' " Silva said. "But that doesn't make an evaluation system. That's not enough."
Justin Snider of The Hechinger Report, a nonprofit, nonpartisan education news outlet affiliated with the Hechinger Institute on Education and the Media at Columbia University, contributed to this report