Mount Vernon Elementary School Principal Gayle Smith had just her hopes and the 75 white ballots she had given to her teachers.

Smith had found what seemed to her the most effective program ever devised to turn struggling students into good readers. She wanted badly to bring it to her Alexandria school, whose scores on standardized tests were among the lowest in Virginia. But, thinking it over in her tiny office, she was not certain her teachers would let her use it.

The reading program, called Success for All, had achieved impressive results at other schools. But many teachers did not like it because it told them exactly what to do throughout their lessons. It gave them all the books their students would read, the questions to ask and the exercises to assign. Following a script was offensive to many teachers. It seemed contrary to what they had long believed about the importance of paying attention to the learning style of each child and varying their instruction accordingly.

It has become one of the central debates in American education: In the drive for higher standards and better test scores, is the classroom teacher who follows his or her own instincts an asset or a liability?

Throughout the country, more and more schools are using brand-name study plans that spell out lesson content and teaching strategy. The reform models -- with names such as Direct Instruction, Core Knowledge and Junior Great Books -- are spreading especially quickly among low-performing schools, where embattled administrators are often eager to overhaul the curriculum and set stricter guidelines for teachers to ensure consistency from one classroom to the next.

This was what Smith hoped to do at her crowded, two-story brick school in a mixed-income neighborhood of northeastern Alexandria. In this case, though, she could not simply impose the new reading program on her teachers. The inventors of Success for All, two Johns Hopkins University psychologists, would not let a school have the program unless 80 percent of the teachers voted for it in a secret ballot. The plan likely would fail without that support, they said.

The requirement was unusual. But most school administrators know that teachers must fully believe in a new instruction method if it is to succeed, and in that sense, Smith faced the same challenge as countless other principals in trying to overcome teachers' doubts.

The difference at Mount Vernon Elementary was that the school referendum forced into the open a debate about teacher freedom, classroom efficiency and the risks and rewards of change, a conflict that normally stays hidden below the surface.

Among the packaged learning plans that have come into vogue, Success for All is one of the most popular. Created for a Baltimore school in 1987, it is now used in 1,130 elementary schools nationwide, including 13 in the Washington area. The program's inventors at Johns Hopkins -- Robert E. Slavin and his wife, Nancy A. Madden -- have built it into a $45 million-a-year business with 265 employees.

The daily reading classes last 90 minutes and group children by skill level so they can move together at the same rhythm.

There is little emphasis on exposing children to classic literature. Most of the stories are written by Slavin or his colleagues and printed on cheap paper with black-and-white illustrations so that schools can afford to let students keep the books. Nor is much importance placed on previous teaching experience. Music, art and gym teachers join the pool of reading instructors so there will be no more than 20 children per class.

The biggest difference is the pace, noticeably faster than in most elementary classrooms. Teachers must complete all the stories and exercises in each lesson plan by the end of the 90 minutes, which forces them to work quickly.

During a recent Success for All reading class for second-graders at the Edison-Friendship Public Charter School in the District, one teacher used a kitchen timer to keep herself on schedule.

"You need to get started," she said to one child dawdling on a short writing exercise. "Time is running out." A few minutes later she said, "You have four seconds."

Much of the time was spent reading aloud in unison to reinforce the sounds of words. "Point, ready, read!" the teacher said. All 14 students began to read a story out loud. Then students paired up for quick critiques of each other's reading.

The teacher was interacting with the whole class during the entire period -- in contrast to a traditional reading class, where a teacher might split her 25 or so students into three or four clusters, assign them different books or chapters and divide her time among the groups.

Several education organizations have judged Success for All to have the best record of any program in boosting reading scores. Despite those accolades, many educators find the program disturbing. They say it denies children the excitement of making their own reading choices, leaves bright children bored and doesn't allow time to address the problems of slower students.

Success for All's lock-step approach "really dishonors the professional craft of the teacher," said Linda M. McNeil, co-director of the Rice University Center for Education in Houston, where many schools are using the program. Educators, she said, "do not find in this program, or any other package, the depth and breadth and the variety of reading styles that they need to get all their kids to read and to find reading purposeful and fun."

Slavin, co-director of Johns Hopkins's Center for Research on the Education of Students Placed at Risk, has heard the criticisms many times. When teachers complain that Success for All will stifle their creativity, he said, there are two possible interpretations: "One, that they really do have terrific ideas and really could do it better than our program," he said, "or two, `I'm lazy, I'm afraid, I've always done things this way.' "

The summer of 1998 was not pleasant for Gayle Smith. She tried to catch some relaxation at a New Jersey beach. But she was upset by remarks two Alexandria School Board members had made about Mount Vernon Elementary's abysmal scores on the Stanford 9 achievement tests. There was "something wrong at Mount Vernon," one of them had said. The school had scored in the 33rd percentile on the tests, 22 points below the average for Alexandria elementary schools.

As requested, she wrote a memo for the board describing the school's problems. One of every four Mount Vernon students was classified as having "limited English proficiency," and many came from immigrant families that were as illiterate in Spanish as in English.

But Smith, who had been at Mount Vernon three years, wanted to move beyond excuses and look for solutions.

She burrowed into the research on Success for All and came away impressed. In a study commissioned by five education groups, including the country's two largest teachers unions, the American Institutes of Research had given the Slavin-Madden curriculum its top rating among 24 education reform models, saying that it had improved reading scores significantly in school after school. At Pineloch Elementary School in Orlando County, Fla., for example, the portion of third-graders reading at or above grade level had climbed from 4 percent to 44 percent in 10 months.

Although some educators were bothered by the rapid pace of the lessons, the program's supporters said this kept children alert and increased the material covered.

Its admirers also stressed the extra tutoring and home visits that Success for All provided. Smith appreciated the value of tutoring. She and other non-classroom staff members at Mount Vernon already spent 30 minutes a day, three days a week with children who needed extra help.

Smith called Success for All headquarters and received a starter kit -- 50 pages of checklists, time lines and advice on how to finance the program.

As school began in September, she called a meeting of her leadership team, which included the elected representatives of teachers at each grade level. She told them she wanted to bring Success for All to Mount Vernon and planned a vote by all the teachers later in the school year.

First she would send out teams of Mount Vernon teachers to visit schools that already were using the curriculum, part of an orientation program Slavin and Madden insisted upon before a single vote was cast. She also urged teachers to research Success for All and any other models they thought might work.

The group was enthusiastic. Teachers shared Smith's worries about the school's bad reputation. Here was a chance to do something about it.

Ken King, the representative of the Mount Vernon kindergarten teachers, dove into his research on reform models. Slender, cerebral and 29, he was an energetic teacher who seemed destined to be a principal himself one day. He had attended an inner-city elementary school in Camden, N.J., and was lured back there to teach by a principal who wanted a young male role model.

King spent several nights scouring the Internet, examining all he could find on Success for All. Like Smith, he was impressed with researchers' findings.

But when he drove to Richmond with three other Mount Vernon teachers to see how Success for All worked at Overby-Sheppard Elementary School, he did not like what he saw.

The reading books, he thought, were drab and unappealing. In many classes, a few of the children were not paying attention. The Overby-Sheppard principal described increases in the number of children reading at grade level, but King was not convinced.

The four Mount Vernon teachers went to a hamburger restaurant in Richmond to share their thoughts before going home. Some of them liked the way Success for All unified the school, so that every teacher and administrator knew what everyone else was doing. But they had seen other outside reforms come and go, and they were worried about putting in the effort to learn a new system only to drop it later.

Some schools in Baltimore and Miami, they knew, had not had success with Success for All. Slavin and Madden argued that in many of those cases, the program had not been implemented correctly, but the failures still raised doubts.

Did the rigid schedule provide enough time to help struggling students? Although the 90-minute lessons were supplemented by 20-minute daily tutoring sessions for the weakest students, the tutors might not always be as well trained as classroom teachers.

Toward the end of the lunch, King confessed what bothered him most. "I wouldn't be comfortable teaching against the clock," he said. "I wouldn't be comfortable not being able to teach one particular tangent, something that relates to a kid's personal experience and allows me to teach from that experience."

One hundred miles north, music teacher Kathleen Baker and four other Mount Vernon staff members visited Fairfax County's Hybla Valley Elementary, the only Success for All school in Northern Virginia.

Hybla Valley, two blocks off Route 1 in a neighborhood full of small garden apartments, had just begun to use the program, but the principal, Tommy Thompson, was very enthusiastic.

Always suspicious of guided tours, Baker, 47, pulled some teachers aside for a private word. They insisted their support for the program was genuine. Some teachers told Baker that once they became accustomed to the rapid march of exercises, they had found time to inject their own ideas.

The teams reported back to Mount Vernon the next day. Baker's group was upbeat. King's group was mostly negative. A third group that had gone to Westhaven Elementary in Portsmouth had a mixed reaction.

For several more weeks, with a break for the holidays, the Mount Vernon staff talked it over, in the school's bright beige hallways, on the phone and at meetings. Should teachers be denied the right to set their own pace and address their students' individual differences?

At one of the schools visited, a teacher had said she liked to stand near the door of her classroom so she could hear whether her class was on the same page as the class next door. Some Mount Vernon teachers considered this robo-teaching.

Broader issues eventually surfaced. Some teachers said privately that Smith was serving as a front for a School Board that cared only about test scores.

As the debate reached a climax, the views of Mount Vernon reading specialist Sigrid Ryberg, 48, became very influential. Ryberg had trained many teachers in a program called Balanced Literacy. At a teachers meeting in December, Ryberg said that her method depended on children having a choice in what they read, and there seemed little time for that in the Slavin-Madden program.

In late January, Smith distributed the ballots at a faculty meeting. Each voter circled one of two sentences: "I would like our school to adopt Success for All," or "I am not interested in adopting Success for All."

King and Ryberg voted no, decisions they had made long before. Baker voted yes, mostly because she liked what she saw at Hybla Valley and thought a principal like Smith deserved faculty support. Holly Hill, a 28-year-old kindergarten teacher, decided at the last minute to vote yes, even though she had been with the team that went to Richmond and came away disappointed. "I thought it was worth a try," she said.

Each ballot was folded, stapled and placed in an envelope addressed to Success for All headquarters. Three days later, Smith's telephone rang.

"I am sorry," said the Success for All staff member on the line. "I am afraid you didn't get the votes."

The final result was 55 percent against Success for All, 45 percent for it -- far short of the required 80 percent support.

Smith sighed. "That's too bad," she said.

The outcome was unusual: Only about 10 percent of the schools that vote on Success for All turn it down, according to Madden. But on another level, it was not very surprising. The number of schools that drop the idea before even voting on it is probably larger than the number that accept it, Madden said.

It is hard to introduce change at an American public school, and sometimes even harder to make it last.

The cycle is a familiar one at many struggling schools. Parents and politicians demand improvements, prompting the principal to announce reforms. The new program does not yield quick results, leading to a debate over whether the plan is flawed, the execution is faulty or the public is too impatient. The pressure to make changes returns, and the old plan is abandoned in favor of a new one.

At Mount Vernon, Smith thought that Success for All's record made it as close to a sure thing as she would ever find in education. Now she is looking for another strategy, knowing that she does not have much time. "The new academic expectations are so broad and the results expected so quickly that we begin to feel disoriented," she said.

This spring, she called in her leadership team, and they agreed on some modest changes for this fall. They decided to reorganize class schedules so that every student will get at least 90 uninterrupted minutes of reading instruction per day. Each teacher, however, will continue to handle students as he or she sees fit.

Mount Vernon teachers say they have worked hard on their children's test-taking skills this year and assume the scores will improve. Smith says she sometimes wonders whether she pushed hard enough and whether she lacked the political skill to win faculty support for such a sharp change in teaching methods. Looking ahead to the next school year, she says, she will miss the training and the focus that Success for All would have provided.

"I thought I knew what I was doing," she said, "and all of a sudden I know nothing and I am starting from scratch and trying to figure all this out."

Mount Vernon Elementary at a Glance

Enrollment: 710

Students poor enough to qualify for free and reduced-price lunches: 76 percent

Students rated "limited English proficient": 25 percent

Percentage of students passing Virginia Standards of Learning tests in 1998

XMount Vernon

YState average

Third grade

English

X24%

Y53%

Math

X27%

Y62%

History

X6%

Y46%

Science

X23%

Y62%

Fifth grade

English

X31%

Y64%

Math

X15%

Y45%

History

X11%

Y31%

Science

X18%

Y57%

The Success for All Foundation

Headquarters: Towson, Md.

Employees: 265

Annual revenue: $45 million

Tax status: Nonprofit

Number of schools using its reading program: 1,130 (up from 128 in 1994)

Washington area schools using the program in 1998-99: Hybla Valley Elementary in Fairfax County; Mudd Elementary in Charles County; and Hendley, LaSalle, McGogney, Thomas, Webb, Wilkinson, Birney, Malcolm X, Miner and Shadd elementary schools and the Edison-Friendship Charter School in the District.

CAPTION: Kindergarten teacher Ken King leads a discussion of "The Very Hungry Caterpillar" as Tracey Harris, left, Leigh Ann Oliver and Johnetta Sowell make caterpillars. King opposed Success for All, saying he would be "teaching against the clock."

CAPTION: Left, Principal Gayle Smith gives first-grader Ishmil Hardwicke a hug after he finishes reading aloud. The two worked together for 30 minutes a day for 10 weeks as she and other faculty members tried to raise scores at Mount Vernon Elementary School.