More D.C. teachers will be at risk of losing their jobs for poor performance in coming years, under a revised rating system, even though standardized test scores will carry less weight in their job evaluations.

The changes — to be announced Friday — amount to the most extensive overhaul of a three-year-old evaluation system that has led to the firing of almost 400 teachers.

D.C. schools were among the first in the country to link teacher pay and job security to student achievement on standardized tests — an experiment that has drawn criticism from many teachers but inspired similar efforts in cities across the nation.

Schools Chancellor Kaya Henderson said the revisions are, in part, meant to raise expectations for teachers. They are also a response to complaints that the evaluation system is too rigid and too reliant on test scores, which don’t render a complete picture of a teacher’s work.

“I’m not stuck with what we thought was right in ’08, or too stubborn to ignore what we’ve learned over the last three years,” Henderson said.

Washington Teachers’ Union President Nathan Saunders welcomed some of the changes to IMPACT, as the evaluation system is known, and praised Henderson for making an effort to listen to teachers.

But he balked at a revision that could lead to the reclassification of hundreds of teachers who are now rated effective. Those teachers might fall into a new category, labeled “developing,” which would put their jobs at risk if they failed to improve.

“You cannot keep changing the d--- bar. We’ve got to be shooting at something, and the target can’t be moving,” Saunders said. “I don’t like it. I don’t think that our members are going to like it.”

Under the evaluations launched in 2009, the 4,100 teachers in D.C. public schools have been put into one of four categories based on how they score on a 400-point scale.

Those who score up to 199 are rated ineffective and are subject to being fired, as are those rated minimally effective (200 to 249 points) for two years in a row. Highly effective teachers (350 to 400 points), meanwhile, are eligible for bonuses of up to $25,000.

Most teachers — about two-thirds — have been rated effective, with scores of 250 to 349. School officials said that category has been overly broad, with teachers at the low end lagging far behind those at the top.

The new “developing” category will encompass teachers at the lower end, who score from 250 to 299. They will be targeted for more help and professional development, and if they fail to earn an effective rating after three years, they will be out of a job.

Nearly half of the teachers who scored effective in 2010-11 would have been deemed “developing” under the new system, officials said. An analysis of 2011-12 data was not yet available.

Saunders said the change leaves teachers feeling vulnerable to ever-shifting expectations in which they have no say. The evaluation system is not subject to collective bargaining by the union, and the chancellor can unilaterally adjust it.

“I’ve got a problem with so much power being in the hands of one individual and the potential for damage that the masses could receive as a result of that,” he said.

Henderson said she has to expect more of teachers if the school system is going to reach ambitious achievement goals in the next five years. Although students have made gains in recent years, fewer than half are proficient in math and reading.

Under the current evaluations, teachers are observed five times a year. They’re graded on their ability to meet nine standards, including managing time, explaining information clearly and correcting students’ misunderstandings.

For some teachers — those who teach math or reading in grades 4 through 8 — half of the evaluation has depended on how students fare on yearly standardized tests.

In the coming year, there will be less weight placed on such test results. Progress on citywide test scores will count for 35 percent of a teacher’s rating. But other measures of student achievement will factor into the ratings. Measures determined by principals and teachers, such as performance on final exams or early literacy tests, will count for 15 percent of an evaluation.

Henderson said she doesn’t expect that change to make much difference in final ratings but hopes it will help calm the anxieties that teachers have when so much of their future rides on test scores.

Several education reform watchers said reducing the reliance on test scores will bring the District more in line with other states and school systems.

“With all the concerns about the reliability and the validity of the scores, with them bouncing around from year to year, it makes sense for them to be a third rather than a half,” said Michael J. Petrilli of the Thomas J. Fordham Institute, a Washington think tank.

“This is exactly how you would want public policy to work — that they make improvements every few years.”

There are a number of other tinkers. For example, only four of five classroom observations will count in an evaluation. The fifth will be informal and strictly for feedback. Also, teachers who are consistently rated effective or highly effective will have only three formal observations.

If one observation yields a much lower score than the others, it won’t count — a recognition that it’s possible for any teacher to have a bad day.

Saunders said teachers will appreciate those changes.

Finally, the school system has changed the way it doles out performance-based bonus pay and salary increases.

The maximum annual bonus for a highly effective teacher will remain $25,000. But now salary increases, previously available to the highest-performing teachers in all schools, will be limited to those working in high-poverty schools — about three-quarters of the total teaching force.