"It's that time of year -- testing time! My fever blisters have returned, my principal has very dark circles under her eyes, our testing coordinator's face is drawn with anxiety, the teachers' faces show deep lines of worry, and the students have rarely smiled during the last two weeks. This is my first testing experience in a school where our test scores determine so much of our future," blogs Betsy Rogers, a 20-year teaching veteran from Alabama who was named National Teacher of the Year in 2003.

Fulfilling the testing requirements
NCLB requires that states develop and implement math and reading standards by the 2005-6 school year and then test children in grades 3 through 8 to make sure that students are on track to achieve 100 percent proficiency by 2014. Schools must also test at least once during grades 10 through 12.

Key to President Bush's claim that NCLB does not impinge on local control of how students are taught, states select or design their own tests. In fact, the law specifically prohibits "federally controlled curriculum."

But the lack of a federal standard has led to a variety of approaches to developing the tests. Some states, such as Maryland, used commercial "off-the-shelf" tests and then modified them to match their standards. Other states, such as Massachusetts, used a more homegrown approach. Many of the states came up with hybrid testing regimens: Georgia, for instance, uses Harcourt's Stanford Achievement Test for grades 3, 5 and 8, but uses its own state-developed tests in grades 4, 6, 11 and 12.

The federal government provided NCLB grants -- more than $380 million per year -- to develop and design tests. If the state already has 3-8 testing, they can use these funds to implement or upgrade the tests.

By the 2007-8 school year, states must also have in place science assessments to be administered at least once during grades 3-5; grades 6-9; and grades 10-12. Further, states had to ensure that districts administer tests of English proficiency -- to measure oral language, reading and writing skills in English -- to all limited English proficient students, as of the 2002-3 school year.

No Child Left Behind requires that schools test at least 95 percent of the various subgroups of children, including students with disabilities and those with limited English proficiency. This provision is meant to rectify the widespread exclusion of students with disabilities from large-scale state and national assessments.

Reporting test results
Schools must "disaggregate" the testing data so that all student groups -- including poor and minority students, students with limited English proficiency, and students with disabilities -- are measured. This disaggregretation is designed to close the so-called "achievement gap" between highly proficient students and underserved, less proficient students by eliminating distortions and variations masked by school-wide averages.

The reporting requirements are costing states and school districts millions of dollars, according to the magazine Education Week's annual report on educational technology.

In a survey of state officials, 15 states reported the three years of NCLB had influenced their decisions to put in place bigger and better data-collection systems. The Education Week study concluded that the focus on data-management technologies is overshadowing the instructional technologies -- such as computers and learning software -- of past years.

"This is a new phenomenon based on NCLB," said Irene Spero, the vice president of the Consortium for School Networking, a Washington-based advocacy group for the use of technology in education.

Comparison with national assessments
As a condition of receiving federal education funds, states are also required to participate in the National Assessment of Educational Progress math and reading assessments for fourth- and eighth-grade students every two years.

The NAEP tests, developed by a rigorous process that includes a rotating board of governors, state legislators, local and state school officials, educators, business representatives and members of the general public, allow educators to see the differences between different state tests.

According to the Department of Education Web site, if there is a large discrepancy between children's proficiency on a state's tests and their performance on NAEP, that would suggest that the state needs to take a closer look at its standards and assessments and consider making improvements. However, the department is adamant that all decisions are left up to the states, and that the federal government has no jurisdiction over a state's test as long as it had approved the process of developing the standards and the correlation between the standards and the tests.

Different states, different tests
What NCLB has created is a system where standards and tests differ in either the content or the level of rigor. A study of 2003 scores by Achieve Inc., a bipartisan, nonprofit organization, shows that states which score similarly on the NAEP assessment, can score radically different on their state tests, revealing a dramatic difference in testing standards.

For example, in Louisiana, 20 percent of a sample of fourth graders scored proficient on the NAEP reading test, but only about 15 percent scored proficient on the state test. In the neighboring state of Mississippi about 18 percent of a sample of fourth graders scored proficient on the same NAEP reading test, but almost 90 percent scored proficient on the state test.

According to Scott Marion, vice president of the Center for Assessment, the difference says less about the tests and more about what each state considers "good enough."

"For many legitimate reasons, states have different views of the meaning of good enough (passing, proficient, or whatever term you want to use). It is really a language problem (not often a technical problem) because most states have chosen to call their good enough performance level 'proficient' which is the same thing that NAEP calls their target performance level," Marion said. "So when states that perform similarly on NAEP or another external measure have widely differing numbers of students that they each call 'proficient,' then we get into public communication problems."

Marion also points out that many people have criticized NAEP for setting achievement levels that are much too high.

According to Scott Norton, director of standards and assessments for the Louisiana Department of Education, Louisiana's state legislature passed a bill that says the state's curriculum should be as rigorous as NAEP. He says one of the challenges was adapting the state's norm-referenced tests to standards-based tests. Norm-referenced tests assess a student's broad knowledge, measuring performance against a specific group, such as a national sample of students of the same grade. Standards, or criterion-referenced tests, measure specific skills in relation to pre-established standards of academic performance.

Norton said there was a lot of pressure from parents to stick with the norm-based test because they could find out how their child performed compared to the norm. However, the state has compromised by adding a writing section to norm-referencing tests.

"The off-the-shelf tests don't test writing. We have two standards that say students should be able to write. So we added that to the Iowa Basic Skills test," Norton said. Such standards tests are harder to design, however, and more expensive to administer.

According to experts like Matt Gandal of Achieve, NCLB has made the work of education assessment experts both easier and harder: "Easier because every state was working on these issues and it had become the law of the land, but harder because there was a risk that states would be so focused on compliance that really good ideas, forward thinking ideas, would be held back. We have been encouraging increased rigor to meet the challenges of the workplace and college, and we fear some of these efforts could be unintentionally hampered."

As states devise strategies for meeting the timeline set by the Department of Education, they can ask for amendments to their standards and their assessment. Such strategies can include lowering their cut-off scores, resulting in more students passing the test, and changing the target rates of how to get to the 2014 goals. The Department says 37 states asked to amend their standards in 2004.

-- By Leah Clapman, Online NewsHour