This is an archived article that was published on in 2011, and information in the article may be outdated. It is provided only for personal research purposes and may not be reprinted.

West Valley City • JoAnn Crawley sits in the back of a classroom, perched on a camouflage-printed bucket with a padded seat. The Monroe Elementary principal quietly observes her subject, a third-grade teacher, and marks off items on a checklist.

The fishing bucket lets the avid angler sit without interrupting the class. But she's watching for something much less tangible than a snag on her line: the signs of a good teacher.

How best to evaluate teachers has become the subject of nationwide debate. A robust evaluation process can help school administrators keep effective educators, strengthen those who are struggling and weed out consistently poor performers. But the debate has grown rancorous as a movement to use student test scores in measuring teacher performance gains momentum.

Most teachers sail through their district's evaluation process even when their students don't meet grade-level goals.

In a survey of 12 school districts in Colorado, Ohio, Arkansas and Illinois, The New Teacher Project found that 99 percent of teachers are deemed "satisfactory" on their final exams. And in Utah's largest districts, fewer than 1 percent of teachers are dismissed each year for poor performance.

Meanwhile, one-third of Utah students are not considered "proficient" on state math and science exams. Minority students are even further behind. Last year, 73 percent of white students scored "proficient" in math, compared to 44 percent of Latino students and 46 percent of black students.

Value-added modeling • Under federal requirements set by No Child Left Behind, students are tested every year to measure whether they're hitting certain goals. Growing concern about American students' poor showing on standardized tests compared to their international peers, coupled with persistent gaps between white and minority students, has led to a movement to grade teachers partly based on student scores.

One method being discussed nationally — and litigated — is value-added modeling, which attempts to quantify a teacher's influence on student achievement, while controlling for outside factors such as poverty. The analysis measures student growth over time by tracking each student's test scores year over year.

"Value-added scores provide the only clear measure of teacher effectiveness — a measure that is systematic and objective," says Eric Hanushek, a leading researcher on value-added analysis and a senior fellow at Stanford University's Hoover Institution.

Replacing the bottom 5 or 10 percent of teachers with just average teachers, Hanushek says, would vault the United States to No. 1 in international math and science scores. Currently the U.S. is ranked near No. 30.

Value-added analysis has been endorsed by the Obama administration and adopted by a number of school systems, including in Chicago and Washington, D.C. In New York, the teachers' union is suing to stop the release of the scores. States planning to adopt value-added methods won points in the competition for federal Race to the Top funds.

This summer, the Los Angeles Times sparked a teachers' union boycott of the newspaper when it did the analysis the school district had not, publishing effectiveness scores for 6,000 elementary school teachers in Los Angeles.

Utah collecting data • Utah's State Office of Education (USOE) is looking at value-added models as it revises professional teaching standards and creates Utah-specific guidelines for evaluating principals. The work is being done by a committee of education professionals, including the head of the Utah Education Association.

"Certainly, student achievement has to play a role" in evaluating teachers, says UEA President Sharon Gallagher-Fishbaugh. "My concern is there has yet to be any sort of definitive research to validate the efficacy of value-added assessments. It is a definite gray area."

In August, the Economic Policy Institute, a Washington, D.C., think tank focused on the interests of low- and middle-income Americans, released a report warning that value-added data can be unreliable. The authors deemed "unwise" proposals in some states to give student test scores up to 50 percent of the weight in teacher evaluation and compensation decisions.

"Such scores should be only a part of an overall comprehensive evaluation," the report concludes. "While there are good reasons for concern about the current system of teacher evaluation, there are also good reasons to be concerned about claims that measuring teachers' effectiveness largely by student test scores will lead to improved student achievement."

USOE already has a data site that can show value-added scores for teachers, but it's not being used to rate educators. And despite a law passed in the 2010 Legislature, the state office has not made the data publicly available because of the cost involved in purchasing an expanded software license. The state office also has stopped publishing annual reports of state test scores by school after lawmakers scrapped the U-PASS accountability system.

"The more that we can empower parents with information about how their school is doing, how their teacher is doing, how we are spending taxpayers dollars, the more we will get them engaged in education," says Judi Clark, executive director of Parents for Choice in Education and a supporter of transparency in schools. "We need to continue to make education a community commitment."

Guiding instruction • Many Utah schools are using student assessments to drive instruction, if not to evaluate teachers.

"Student achievement data should be used to determine gaps in teaching," says Sydnee Dickson, USOE's director of teaching and learning. "The data should be used for purposes of [teacher] improvement — not purposes of punishment."

Of the state's five largest districts, only Davis factors in student achievement data when it rates teachers. The district does not use value-added modeling but is researching the technique, says educator assessment coordinator Suzanne Cottrell.

In teacher evaluations, districts are required by state law to consider "multiple lines of evidence," such as classroom observations by principals, student and parent feedback, a teacher's self-evaluation, student achievement data and the teacher's participation in professional development activities.

Provisional teachers must be evaluated twice a year. But once they reach career status, typically after three years, they receive formal evaluations once every three to five years.

"The frequency has to increase for all those teachers," says Daniel Weisberg, vice president of policy and general counsel for The New Teacher Project, a New York-based nonprofit that consults with districts on teacher evaluation systems. "The idea that a veteran teacher doesn't need feedback is wrong-headed."

Every teacher should receive an evaluation at least once a year, Weisberg says. In surveys his group has done, both veteran and novice teachers say they want more frequent and more meaningful feedback. His group recommends that annual evaluations be tied to employment and compensation decisions.

Sen. Howard Stephenson, R-Draper, plans to sponsor a bill in the upcoming legislative session that could cost a teacher career status if he or she consistently fails over time to produce adequate student progress. Provisional teachers are subject to annual contract renewals, but career teachers can only be fired after a more extensive due process.

"The public certainly wants to fire bad teachers. I think there's a myth out there that that work isn't done," Dickson says. "It's complex and it's hard to do. Our goal is to ensure that all students have high quality instruction every day."

In Granite district — Utah's largest with 68,000 students and 3,500 teachers — career teachers receive formal or "summative" evaluations once every five years that are filed in their personnel records. They have informal, formative evaluations, with goal-setting and feedback, once a year.

The district generally dismisses 25 to 35 teachers a year through termination, counseling teachers out of the profession or not renewing provisional contracts, says Mike Fraser, Granite's assistant superintendent of school accountability.

"Our teachers' association works with us, hand-in-hand, in helping us get rid of poor performing teachers," he says.

Classroom management • Like many principals, Monroe Elementary's Crawley conducts near daily "drop-in" visits that last less than five minutes. With her fishing bucket in tow, she aims to be in each of the school's 30 classrooms twice a month.

She uses a checklist with 15 items to assess student engagement, the classroom environment and behavior management. What percentage of students are on task? Has the teacher posted the lesson's objective on the board? Does the teacher speak in a clear voice and move throughout the room to maintain proximity to students?

Crawley leaves each teacher a note with "polishers" — things that can be improved — and things that are working well, or "keepers."

"The foundation to me is having the classroom-management piece. If that is solid, you will see kids who perform better," Crawley says. "That's the key. We want our kids to learn."

Tribune reporter Lisa Schencker contributed to this story. —

Teacher evaluations in Utah's five largest districts

The New Teacher Project, a national nonprofit that has helped establish teacher quality programs in 31 states, has published guidelines for teacher evaluations. "Teacher Evaluation 2.0" recommends: conducting formal evaluations of every teacher at least once a year; using multiple measures, including impact on student achievement; employing four or five possible ratings to describe differences in teacher effectiveness; and using evaluation data in making employment and compensation decisions about teachers.


Enrollment • 68,131

Frequency of evaluations • Twice a year for provisional teachers; every five years for career teachers.

Student achievement data used to measure teacher performance? • No.

Ratings • Satisfactory or unsatisfactory.

Number of teachers dismissed in 2009-10 • 25 to 35.


Enrollment • 65,452

Frequency of evaluations • Twice a year for provisional teachers; annually for career teachers.

Student achievement data used to measure teacher performance? • Yes, annually for provisional educators and every four years for career teachers.

Ratings • Unsatisfactory, needs attention or well functioning.

Number of teachers dismissed last year • Unknown. District says it doesn't keep track.


Enrollment • 64,351

Frequency of evaluations • Twice a year for provisional teachers; every three years for career teachers.

Student achievement data used to measure teacher performance? • No.

Ratings • Scale of 1 to 5.

Number of teachers dismissed last year • 18.


Enrollment • 48,411

Frequency of evaluations • Twice a year for provisional teachers; every three years for career teachers.

Student achievement data used to measure teacher performance? • No.

Ratings • Met standard or did not meet standard.

Number of teachers dismissed last year • 30.


Enrollment • 33,184

Frequency of evaluations • Twice a year for provisional teachers; every three years for career teachers.

Student achievement data used to measure teacher performance? • No.

Ratings • Met standard or did not meet standard.

Number of teachers dismissed last year • 15. Utah has value-added data on teachers but doesn't share it

Utah's State Office of Education (USOE) recently started running a newly designed data reporting site for teachers and administrators that includes testing data by student, class, grade, school, district and state. It's possible to see classroom performance of teachers over time, though they are identified by ID number, not name, said John Jesse, USOE's director of assessment and accountability at the state office.

But only educators have access to the site — not parents. In fact, last spring's state test results aren't available by school at all on the state office's website. USOE had published those results each year on its website but has stopped since a new law passed earlier this year suspended the state's U-PASS accountability system.

Education officials have said the new site is not available to the public for lack of money to buy anexpanded software license — an expense of about $240,000 a year.

"It's a significant increased cost," said State Superintendent Larry Shumway at a State Board of Education meeting in September.

But Sen. Howard Stephenson, R-Draper, said the data should be available to everyone. Stephenson co-sponsored a bill that passed in 2009 to make state test data, by class, available to the public.

"The student privacy needs to be protected, but anyone else who works in the system we ought to be revealing their efforts," Stephenson said.

"I think parents have a right to know that information," Stephenson said. "It shouldn't be something that is kept internal because what we find is there is the dance of the lemons in which teachers are transferred from school to school every few years after people get so upset they demand a change."

He said lawmakers would likely support funding to make that data public.

Lisa Schencker