As the daughter of two public shcool teachers, Jeanne Slavin Kaplan had always considered herself a "complete teacher advocate." That is, until site had two children of her own in Denver public schools. I have seen a number of things over the years that have made me reevaluate my position," Kaplan says.

For example, when her daUghter was in the third grade, Kaplan was unhappy with her teacher. So were some other parents. Kaplan tried to get her daughter out of that teacher's class and into another one. When she was unsuccessful, she and other parents persuaded the principal to get the teacher out instead. And the teacher was passed off to another Denver public school.

With pressure building from parents and others, education's accountability movement may soon catch up with teachers. As student test scores come under greater scrutiny from state and school officials, policy-makers and parents are beginning to ask how teachers are measuring up, too. "A lot of districts are now talking about a parrallel theme: We've got accountability for students, and maybe we should have accountability for teachers," said Allan R. Odden, a professor of education administration at the University of Wisconsin (Madison). In other words: If children can't be socially promoted, why should teacher's?

This year's National Education in Summitt endorsed a 10-state pilot project to examine performance pay for teachers. At the sunit. National Urban League President Hugh Price elicited strong applause with his pitch to free teacher hiring and firing from union collective bargaining. Sen. John McCain of Arizona, a Republican presidential hopeful, has endorsed merit pay. teacher tests, and ways to send potentially bad teachers down another career path. Denver is launching a pilot program to test the link between teacher pay and student achievement.

Although there is a growing consensus among lawmakers and educators that some form of teacher evaluation is necessary to infuse public education with more credibility, the "how" of doing that is tricky. Designing an systme that requires accountability without alienating good teachers will be a challenge. Just about everyone agrees that teachers must be on board for accountability to work.

Those who have studied effective schools, especially ones in low-income areas, say that weeding out bad teachers is an important factor in improving student achievement. Robert L. Mendro, chief evaluation officer for the Dallas Independent School District, who studied successful schools in that city, describes the practices of the effective principals he has seen in the schools. "They set this tone--that we expect our kids to learn. The thing they did was to be far less tolerant of bad teachers."

The question of teacher accountability sets up a new debate on measuring teacher success. Subjectivity has always been a sticking point for such measurements, but many researchers say that the availability of data will let them objectively determine teachers' effectiveness. At the forefront of this research is William L. Sanders, a statistician at the University of Tennessee. Sanders has developed a complex statistical model that measures the "value" a teacher adds to his or her students based on their change in test scores over a given year and that factors out such variables as socioeconomic status. The problem, he says, is that "without measurement. [evaluating teachers] is total], a political process."

WILLIAM SANDERS: The statistician has developed a complex formula for measuring the value a teacher adds to his or her students.

Sanders has been racking up thousands of frequent flier miles to spread the word, including some for a recent trip to Denver, and other researchers are starting to examine his idea. Sanders has found, for example, that children who have weak teachers for two consecutive years never overcome that setback. In Dallas. Mendro found that StUdents who had excellent teachers for three years in a row scored 40 percentile points to 50 percentile points higher on standardized math tests than students who had sub-par teachers for three straight years. Researchers in Massachusetts arid Mama have come to similar conclusions.

Drawing on the "value-added" research, William A. Galston, a professor in the. School of Public Mfairs at the University of Maryland, says that schools should give teachers the option of a lower salary with the potential for large bonuses if they can accelerate student performance. This bonus-based approach to teaching, he says, will encourage good teachers to take classes of low-achieving students because they offer a high potential for improvement.

And policy variations on Sanders' theme of measuring teacher ability by, student improvement are being discussed and tested in states such its Tennessee and cities such as Denver. In Tennessee, the state's assessment system allows Sanders' "value-added" numbers to be incorporated into formal teacher evaluations. But parents never see the, teachers' scores, and the system does not link those evaluations to rewards or sanctions.

In Denver last month, the school board and teachers' union forged a compromise setting tip the first plan in the country to tie teachers' bonuses to their own students' achievement. The two-year pilot program offers teachers an annual bonus of up to $1,500 if, with their classes, they meet two specified achievement goals by the end of the year. That compromise exemplifies the pressure unions are feeling to recognize individual teacher performance. "We were bashed for wanting to keep really crappy teachers." said Andrea Giunta, the president of the Denver Classroom Teachers Association.

In North Carolina, schools have been rewarded and punished based on their test scores; individUal teachers have not. But detailed measurement is on the way, said John Dornan, the executive director of the Public School Forum of North Carolina, a nonpartisan policy center. "All of the objections about rising one test for evaluations-as we get more sophisticated with management of information, those arguments ao out the window," he said. "We're suddenly getting into far more sophisticated territory than someone sitting at the back of the room with a checklist."

Proponents of judging teachers based on their students' scores argue that such an approach will further professionalize teaching. "There are very few types of pure output productions in the world. There are subjective judgments made of almost every.job and even, occupation in this world," said Site Edwards, the president of the Denver school board. "I'm not afraid of moving in this direction, because I think it will strengthen public education and I think it will strengthen the support for public education."

But some education researchers, Such as David Grissmer, who is a senior managenient scientist at RAND, a Santa Monica, Calif.-based public policy research organization, caution that inany factors affect student scores. "In the typical American fashion, we usually go too far in this stuff," Grissmer said. "It's such an appealing concept that they're going to try to move in that direction. We've got to be a little careful about creating a set of bad measures. I think we're on a slippery slope here." And Linda Darling-Hammond, the executive director of the National Commission on Teaching & America's Future, raised concerns about the incentives created by a student achievement-based evaluation. "When you reward teachers for student achievement, nobody wants to teach certain kids in certain communities," she said.

NEW INCENTIVES: A pilot program in Denver offers teachers bonuses of up to $1,500 if they meet specified achievement goals.

Feeling the pressure, some local teachers' unions are promoting their own answer to the accountability question-teachers judging teachers.

They are slowly beginning to support "peer review," a system in which experienced teachers make extensive evaluations of new teachers and veteran teachers who are struggling ill the classroom. They then make recommendations to help or, in extreme cases, to dismiss these teachers.

Tom Mooney, the president of the Cincinnati Federation of Teachers, cites his district's record over the 14-year history of the peer-review program. Of the 90 teachers sent to an "intervention" for a big professional development boost, onethird improved and two-thirds were removed from the classroom. Until this Year, only a handful of local districts employed peer evaluation, but interest is growing. California this year launched the first statewide version. And New Jersey Gov. Christine Todd Whitman said that a chat with Mooney at the National Education Summit piqued her interest in the program.

"It seems to be reaching critical mass in the last few years," Mooney said of peer review. "I think it's partly the search for effective approaches to accountability that are not just crude and stupid, but are based in professional standards. . . . You need to have a standard, and von need to have a way of enforcing that standard that is credible." But Amy Wilkins, a principal partner at the Education Trust, an education research group, says that peer review fails to "take the obvious next step," to base evaluations on student achievement.

Kaplan, the Denver parent, said that the best way to restore parents' confidence in teachers is to have teachers evaluate one another. The evaluating teacher would factor in several assessment tools, including student achievement, the context within which the teacher is teaching, the background of the students, and the quality of the school administration. The catch, she says, is that the evaluators would have to be willing to expel the bad teachers from the system. I do think there are a number of teachers who really shouldn't be teaching anymore," she said. "A lot of them are biding their time until retirement."

There may be ground for political compromise in Kaplan's value-addedpeer-review hybrid. The National Education Association, which advocates peer review, is willing to listen. Sanders' research "contributes to the debate and the dialogue, and could contribute to the substance," said Chuck Williams, the NEA's director of teacher quality initiatives. "But it could not and should not be the silver bullet."

And some union leaders concede that measurement is inevitable. "When we get to the point, and we will, where We can disaggregate to isolate the impact of a teacher on students learning vis-A-vis other factors, then we can raise this question of rewards and consequences, but until then, it's irresponsible," said Adam Urbanski, a vice president of the American Federation of Teachers and an advocate of peer review.

The real pressure to find a compromise is likely to come from such parents as Kaplan, who will be better armed with information as states start issuing report cards and ending social promotion, the practice of passing children from one grade to the next even when they're not academically prepared. As states move to end social promotion, parental pressure to differentiate between good and bad teachers will grow, Dornan said. And these evaluations will be easier to do. "You're going to open up a whole other layer of scrutiny," Dornan said. "But this time, it will be focused on classrooms, not buildings, and classrooms will be a proxy for teachers."

And, as Kaplan and Sanders pointed out, public schools have plenty of dynamic teachers, but the problem is that teacher quality is so uneven. "All professions have bad eggs," Kaplan said. "This one just happens to affect our society more than most."