Statewide student tests.

* Put too much emphasis on single assessments.

* Are just what we need to raise achievement.

* Don't go far enough in holding teachers and schools accountable.

* All of the above.

It's the last week in April, and more than 1,000 sixth, seventh and eighth graders crowd into the gym at De Leon Middle School in McAllen, Texas. They whisper excitedly as the school band begins to play the fight song. Cheerleaders cheer. Kids perform skits and sing songs. It is a pep rally of Texas-size proportions. But this particular pep rally is not about gearing up for a football game or a fund-raiser. It is about academics. The students and teachers at De Leon are getting revved up for the annual Texas Assessment of Knowledge and Skills test (TAKS).

You know something's changed when schools in Texas are holding rallies for tests. Although standardized testing has been a reality in many states for decades, a flurry of activity in recent years has resulted in a much higher profile and plenty of controversy for state assessment programs. Now, not only are students being tested, their teachers, schools and districts are being held accountable, as well. And the stakes are high.

"Testing for accountability has been with us in various forms probably for the last 40 to 50 years," says Joan Herman, co-director of the Center for Research on Evaluation, Standards and Student Testing (CRESST) in Los Angeles. "In the last 10 to 12 years, it's the consequences attached to performance that have been growing pretty drastically."

THE CONSEQUENCES

Indeed, more states are beginning to withhold high school diplomas and retain third graders for failing to pass state tests, igniting strong emotions in parents, students and even some school districts.

In Massachusetts last year, several school districts threatened to defy the state by issuing diplomas to seniors who failed to earn a competency determination as required under the Massachusetts Comprehensive Assessment System (MCAS). Phone calls from Education Commissioner David R Driscoll eventually quelled the rebellion, and a student-filed injunction to stop the graduation requirement was denied by state and federal courts.

Statewide, 92 percent of Massachusetts high school seniors passed the MCAS test. Even with remediation classes and retesting opportunities, however, 4,800 students didn't get their diplomas.

The stakes will continue to grow as states implement the complex requirements of the federal No Child Left Behind Act (NCLB) over the next decade. The new law requires states to test all students annually in math and reading in grades three through eight and at least once in math and reading in grades 10, 11 or 12, beginning in 2005-06.

Schools must use state-defined performance standards to show continuous gains--or adequate yearly progress (AYP)--in achievement for all students, including those with disabilities and limited English proficiency. Schools failing to illustrate adequate progress must offer school choice or supplemental services and may eventually face corrective action, which could include replacing staff, receiving technical assistance or even being turned into a charter school.

Advocates of high-stakes testing maintain that tough consequences are necessary to achieve the goals of education reform, which are that students and schools be held to a high standard, that students be tested on and measured against those standards, and ultimately that public school students be better prepared for higher education and the workforce. The goal of No Child Left Behind is unambiguous: that every student in America be deemed proficient or better by 2014.

States with fairly rigorous assessment systems already in place point to increasingly better test results as proof that high-stakes testing works.

"MCAS holds the whole system accountable, starting with the students," says Sylvia M. Smith, chief of staff for Senator Robert A. Antonioni of Massachusetts. She cites progress made by the class of 2003 as an example. "Between the class of 2002 and the class of 2003, there was a significant increase in the number of students who passed the English language arts and the math test, because that class was being held to the graduation requirement," she says. Seniors who failed the test in 2003 were allowed to take it again, however.

Florida Governor Jeb Bush announced in May that more students met or surpassed the proficient level in 2003 than ever before on the Florida Comprehensive Assessment Test (FCAT). Progress was particularly dramatic for minority students, with 51 percent of Hispanic fourth graders scoring at or above grade level in 2003, compared with 38 percent in 1998.

Despite the dramatic progress announced by the governor, about 2,500 protesters gathered outside his office in May to demand he suspend FCAT. Requirements for 2003 were that high school seniors pass the 10th grade exam to graduate.

"We've seen tremendous results," says Ohio Senator Robert A. Gardner, the "father" of that state's assessment program. There has been a dramatic drop in the number of schools on academic emergency or academic watch, while more schools have jumped to the excellent category.

"We had 71 schools last year that were in the excellent category. This year, we have 109," he says.

THE UNINTENDED CONSEQUENCES 

But critics point to predictions that harsh consequences for failing to show annual progress will tempt states to lower their standards in order to meet the federal act's proficiency targets, and that the federal formula will end up labeling perfectly good schools as failing. Officials in Iowa and Nebraska are concerned because their students already perform well on tests, so schools in those states may have trouble showing improvement in the annual increments required by NCLB.

In fact, three school districts in Connecticut turned down Title 1 funding under NCLB because of fears of "too much federal intervention" and the sanctions on "failing" schools imposed by the federal law.

And some question whether high-stakes testing actually motivates students and schools to perform better. Although studies to try to prove it remain controversial, the National Education Association maintains that high-stakes testing does not improve student achievement and that it may in fact produce more dropouts.

Michael Pons, policy analyst for the NEA, says that even though teachers welcome the emphasis on greater clarity in standards and expectations for students, the federal act's across-the-board requirement for adequate yearly progress is unachievable.

"It's the equivalent of saying every kid should be three feet six when they're in fourth grade," he says.

Representative K. Mark Takai of Hawaii also believes that while schools and students must perform better and make progress every year, the requirements of NCLB are a recipe for failure.

"The surprising element of this law is that it makes no distinction between a school in a well-educated and very affluent community and a school in an impoverished community," he says.

ONE SIZE CAN'T FIT ALL

Representative Takai says there is a strong, predictable correlation between a school's performance on assessment tests and factors that are beyond the control of the school and its staff. Those factors include attendance rate, percentage of disadvantaged students, percentage of limited English proficiency students, percentage of special education students and other factors.

"The federal government is dreaming if they expect those student groups to perform at the level being required by the federal law," he says.

Lynn Stockley, a seventh and eighth grade counselor at Carver Middle School in Tulsa, Okla., agrees. Even though students at Carver, a magnet school, do well on tests, Stockley contends that comparing urban and suburban schools is like comparing apples and oranges.

"It's really not fair to judge a school whose kids come without having ever read a book, and compare it to a school where kids go to the library and have libraries in their home," she says. "It's rather discouraging when you're one of those schools. You know you're working very, very hard and yet your scores are consistently low."

Critics of high-stakes testing also point out that simply scoring well on tests doesn't necessarily mean that students are learning what they need to know to be better prepared for the future. Common complaints are that some students simply don't test well (especially on timed tests), and that such tests force teachers to spend too much time teaching to the test.

"I am completely opposed to the notion that the school's job is to prepare for the test," says Scott Hollinger, principal at McAuliffe Elementary School in McAllen, Texas. "My belief is that the school's job is to teach a well-rounded curriculum and the test is to tell how well we're doing."

Current assessment programs, however, do not encourage a broad curriculum, he says. Many of his colleagues focus too heavily on reading and math, for example, because their students will be tested only in those subjects. Furthermore, he says, many schools are waiting until fifth grade to begin teaching science, because students aren't tested in science until then.

Barbara Provenzano, a fourth grade teacher at Forest Hill Elementary in West Palm Beach, Fla., maintains that the pressure to perform well on the FCAT drives out opportunities for creativity.

"We don't like the fact that everything is so assessment driven," she says. "As teachers, we have less energy to put toward projects, fun activities, activities that are more geared toward getting the students interested and involved."

Provenzano used to organize an end-of-unit region feast, during which her students would bring in different dishes from various regions they had studied. But that popular activity has fallen by the wayside, a victim of the prevalence of testing.

Senator Gardner of Ohio, an educator for more than 28 years, argues that there is nothing wrong with teaching to the test. "During the week I gave the kids information, and at the end of the week I tested them on it," he says. "So if that's what I wanted them to know, then that's teaching to the test."

DRIVEN BY STANDARDS

As the consequences become more severe, the tests themselves are growing more rigorous. Most experts agree--and No Child Left Behind requires--that states develop challenging standards and align their tests with those standards. As such, many states have spent considerable time developing their standards with input from educators, parents, and business and political leaders. Then they've developed customized tests to accurately reflect and measure those standards.

"We took a few years to develop our standards, then took a couple of years to make sure our tests were aligned with them," reports Mike O'Farrell, staff director for the Florida Senate Education Committee.

"Florida has done an excellent job of aligning tests with standards."

President Bush's home state of Texas did much the same. The TAKS program is the most recent incarnation of an assessment system that has been evolving since 1978.

"Now we have a uniform curriculum and a way of testing that uniform curriculum," says Adrienne Sobolak, deputy director of communications at the Texas Education Agency.

With higher standards come more complex tests. Before the 1990s, large-scale assessments consisted primarily of multiple choice tests, according to CRESST's Herman.

Since then, tests with more open-ended questions were developed to paint a more accurate picture of a student's knowledge and skills. Today, many state tests include writing assignments. And some states have experimented with portfolios, hands-on activities, teamwork skills and the like.

"Part of the impetus for all this performance assessment was trying to get at these more complex skills," Joan Herman says, "and understanding the problem-solving skills that we think are important for kids to know and be able to do."

But Herman says there has been somewhat of a retrenchment from the more open-ended tests because of the expenses associated with them.

THE COST FACTOR

A May 2003 report by the United States General Accounting Office (GAO) found that, in general, tests that are hand scored-those with essay questions, for example--are more time and labor intensive and more expensive than tests that are machine scored, such as multiple-choice tests. The GAO estimates that between FY 2002 and FY 2008, total state expenditures would be about $1.9 billion if states used multiple-choice tests. That figure jumps to $5.3 billion if states use a mix of multiple-choice questions and open-ended questions requiring written responses.

In light of the huge budget shortfalls being faced by almost all states, policymakers are concerned about their ability to adequately finance assessment programs. No Child Left Behind does provide states with additional money ($387 million in FY 2002), but questions remain about whether federal funds will cover states' costs. Policymakers wishing to incorporate more thorough, open-ended tests must find a way to finance them, or face compromising their standards in order to meet all the federal requirements within the fiscal constraints of the federal law.

Budget shortfalls are also threatening other aspects of assessment programs. Besides paying for tests, state funds are typically directed to high-performing schools as part of rewards and incentives programs, low-rated schools that need help in raising their performance, programs

for improving teacher quality, and myriad other school improvement efforts.

In Texas, for example, a program that rewards schools for excellence is no longer being financed, and legislators were looking at cutting funds for math and reading academies for teachers. In Ohio, Senator Gardner is struggling to find a way to restore teachers' professional development money, a victim of budget shortfalls.

A PRESSURE COOKER

While policymakers and administrators grapple with assessment systems at the state level, teachers and students are facing their own challenges in the classroom.

Testing is "a necessary evil," says Christena Southwick, a fourth grade teacher at Kendrick Lakes Elementary School in Lakewood, Colo. "There needs to be standards, some sort of leveling so that the teachers are all on the same page, kids get all of the same skills covered, and there's no unevenness in their skill development. That's part of not leaving any child behind."

On the other hand, she says, standardized assessments force all students to be tested the same way even though they learn in different ways and at different rates.

School testing also tends to raise the anxiety level in the classroom.

"The kids are nervous," says Provenzano. "We do our best to try and keep it low key, but on the other hand we can't afford to let it be too low key. We have to communicate how important it is."

Southwick agrees. "There's so much pressure on the teachers who in turn put pressure on the kids to do their best."

Back at De Leon Middle School in McAllen, Texas, Principal Jodie Ellis says that after "blowing off steam" at the pep rally, the TAKS test itself was a serious, but positive, experience. The kids know what to expect, she says, because the state has done a good job of communicating the standards to schools and providing resources, such as workbooks, for students who need extra help.

And students seem to be rising to the occasion. Nearly 90 percent of the members of the Texas Elementary Principals and Supervisors Association say that the Texas system is helping more children succeed, according to Executive Director Sandi Borden.

"So many more children--and that's real children--are learning at higher levels than they ever were before," she says. "We've narrowed the gap in a state whose children have become increasingly poorer. It shows policymakers that, indeed, their investments are paying off."

WHERE'S THE MONEY? New Hampshire's Experience

Finding the money to pay for mandated tests is a challenge facing most states. In New Hampshire, it became a real issue. The legislature wanted the state Department of Education to use money from the No Child Left Behind law to cover all testing. But federal money, argued the agency, cannot be used to pay for tests already required by the state, only to expand them to additional grades.

Convinced that federal money should be available, budget writers in September cut the $2.7 million appropriated for state testing in each of the next two years to $1 to force the issue. State officials argued they were getting mixed messages from the federal government on exactly what would be covered by federal funds.

The tests must be given for the state to comply with NCLB--or it risks losing $60 million in federal aid.

The situation has prompted the state to look at trimming some of the current state tests in an effort to save money. State Education Commissioner Nicholas Donohue proposed a plan in November that sets aside $2.4 million for next year's tests. Two thirds--or $1.6 million--would come from federal money through No Child Left Behind. The remaining $800,000 would be from state money, primarily an unspent portion of a special education grant received from the federal government. The proposal would eliminate testing in science and social studies this year and the language arts tests would no longer include a writing sample. The changes are expected to save about $600,000.

--Scott Young, NCSL 

Jane Carroll Andrade is a former State Legislatures editor who now writes from her home in Evergreen, Colo.