Testing Systems in Most States Not ESEA-Ready
Most states will have to retool their testing programs to meet new federal mandates, a survey conducted by Education Week shows.
Based on a strict interpretation of the federal education legislation passed by Congress in December and set for signing by President Bush this week, few states currently meet its requirements for an unprecedented expansion of state testing systems.
The "No Child Left Behind" Act of 2001, a reauthorization of the Elementary and Secondary Education Act, will require states to give annual reading and mathematics tests to all students in grades 3-8 no later than the 2005-06 school year. Under the legislation, states may select and design tests of their choosing, but they must be aligned with the states' respective academic standards.
"I think it's going to be a big change for a lot of states," said Margaret E. Goertz, a co-director of the Consortium for Policy Research in Education, a federally financed research center at the University of Pennsylvania, "because they haven't been doing annual testing."
The tests also must produce individual interpretive and diagnostic reports on students—and itemized score analyses—that allow parents, teachers, and principals to understand and address specific academic needs. The tests must enable the results to be broken down by gender, racial and ethnic group, migrant status, English proficiency, disability, and income for each school and district.
It's the annual testing requirement in grades 3-8 that poses the most immediate challenge for states. But the law also requires that states test students at least once in grades 10-12 in, at a minimum, reading and math. And by 2007-08, states will have to give students a science test at least once in grades 3-5, 6-9, and 10-12.
States that do not comply risk losing some federal education funding, although historically the Department of Education has been loath to withhold money.
For Quality Counts 2002, released this week, Education Week surveyed all 50 states and the District of Columbia about their assessment practices for the current school year. States reported what tests they give, the subjects the tests cover, and the grades in which the tests are administered.
They also noted whether the tests are "norm referenced," meaning designed to measure how students in the state perform compared with a national sample, or "criterion referenced," which means they're designed to measure whether students have mastered a body of knowledge—in this case, the state's academic standards.
In rating the states for its annual Quality Counts report, Education Week gave credit to a state for having a "standards-based test" if the test was criterion-referenced, or if the state added questions to customize a norm-referenced exam to reflect the state's standards.
Based on that analysis, 17 states give some type of English and math test in each of the grades from 3rd through 8th. Fourteen of those states administer "comparable" norm-referenced or standards-based tests in English and math in those grades, meaning the results should be able to be uniformly compared across grades and across schools and districts within the state.
But only nine states administer standards-based tests in both subjects in every one of those grades, as the new law will mandate.
The law authorizes up to $490 million a year in federal aid to help states develop and administer the mandated tests. If Congress does not appropriate enough money in any given year—$370 million in fiscal 2002, and escalating in $10 million increments in each of the succeeding years—states may postpone the 2005-06 deadline for giving the tests, but they must continue to develop them.
Waiting for Clarification
One of the biggest questions state officials have is whether they will be able to use a combination of state and local assessments in different grades to meet the new testing requirements. Such flexibility could greatly reduce the number of new tests that states must devise and has been advocated by such groups as the Council of Chief State School Officers.
In a Dec. 18 letter to Sen. Susan M. Collins, R-Maine, Secretary of Education Rod Paige wrote that the law "does not preclude a mixed state/local system like Maine operates, provided the system is able to meet the other requirements in the legislation."
But Mr. Paige made it clear that states would have a difficult hurdle to jump. To ensure that local assessments were aligned with state standards and were of acceptable technical quality, Mr. Paige wrote, a state would have to establish a process to review and approve local assessments to be included in the state testing system.
"This process must ensure rational, fair, and equitable determination throughout the state for identifying schools in need of improvement," Mr. Paige wrote. In addition, the secretary's letter noted, "Maine would need to be able to aggregate, with confidence, data from disparate local assessments."
The assessments also must include the participation of all students, the secretary emphasized, including those with disabilities and limited proficiency in English, and the tests must produce individual student reports and itemized score analyses.
Mr. Paige also wrote that President Bush supports annual tests because he wants teachers and parents to have data that show student progress against state standards.
"Data from 4th grade results need to be relevant and helpful, for example, to the 5th grade teachers," the secretary said in his letter to Sen. Collins. "So, a state that uses both local and state tests needs to assure the secretary that the tests will be coordinated at least to that extent."
In Maryland, state schools Superintendent Nancy S. Grasmick predicted that the state would have to do some "tweaking" of its existing standards-based test, the Maryland School Performance Assessment Program, to produce individual scores for students, as the new law will require.
The state gives the MSPAP in grades 3, 5, and 8, and a norm-referenced test, the Comprehensive Test of Basic Skills in grades 2, 4, and 6. It gives a separate, functional-literacy test in grade 7.
"We really want to see, what are the requirements here," said Ms. Grasmick, "and we don't know that for sure yet. I think it's through the regulations that we're going to see what the options are."
The legislation requires that the federal secretary of education write final regulations for the standards and assessment provisions within a year of the measure's enactment. To do so, the secretary must establish a "negotiated rule-making process" that allows for the participation of various stakeholders including parents, students, educators, and education officials across the country.
The secretary also must establish a peer-review process, made up of panels of individuals knowledgeable about testing and accountability, to review each state's plan in those areas.
Questions of Comparability
Wayne Martin, who is the director of the state education assessment center for the Washington-based Council of Chief State School Officers, said, "I think the assumption is that [the tests] are going to have to be comparable, although this is going to have to be spelled out in the regulations, because I could make an argument that there's nothing in [the law] that specifically says that."
Sandy Kress, Mr. Bush's chief education adviser, said, "We believe that states must show, in their plans, a testing system that is aligned with state standards, that fairly and rationally evaluates schools across the state, and gives parents and teachers useful information on the annual progress of their children."
"It would be our view that the states need to show that their testing system meets the specific criteria that I just mentioned," Mr. Kress said. "Technically, that may not need to be 'comparability.' "
But Amy Wilkins, a policy analyst for the Education Trust, a Washington-based group that works to raise the achievement of poor and minority students, said the failure to spell out in the law whether the tests must be comparable across grades and schools was a major weakness.
"You want the 4th grade teacher to be able to interpret the 3rd grade data, and you won't be able to with this mishmash of tests over time," she argued.
"Lack of geographic comparability is really damaging to bringing greater equity," Ms. Wilkins added, because there's "no way to equate the scores across the districts."
Vol. 21, Issue 16, Pages 1, 26-27Published in Print: January 9, 2002, as Testing Systems in Most States Not ESEA-Ready