Problems Persist for Clinton’s Proposed National Tests

When President Clinton proposed new national tests 18 months ago, he said 4th graders and 8th graders throughout the country would be sitting down to take the exams in the spring of 1999.

Since then, his plan has been through a political whirlwind, and its starting date already has been delayed by at least two years. Moreover, the board given responsibility to create the voluntary exams still hasn’t solved many of the complex questions raised by Mr. Clinton’s proposal for 4th grade reading and 8th grade mathematics assessments.

About six months from the initial starting date, the National Assessment Governing Board has yet to negotiate such vexing issues as:

• How to write new national tests that would report individual student scores in the same way the National Assessment of Educational Progress reports the performance of groups of children in four categories: “advanced,” “proficient,” “basic,” and “below basic";

• How to offer accommodations to limited-English-proficient students and disabled children; and

• How to manage the task of buying and distributing the hand-held calculators that 8th graders taking the math test would need.

Now, even though a new report evaluating the testing initiative’s progress says the governing board, known as NAGB, and its contractors “have made a great deal of progress toward the goal of developing a [test] item pool of adequate size and of known, high quality,” it also finds lingering questions over such issues as accommodations and reporting of scores. “Without timely attention to these concerns, future progress is unlikely to be satisfactory,” evaluators hired by the National Research Council write in a report released last week.

AGB still doesn’t know if it will have the chance to address those questions. Congress and Mr. Clinton must come to an agreement this month about allowing the nonpartisan board to continue test development after fiscal 1998 ends Sept. 30.

The House will soon debate a fiscal 1999 spending bill that would prohibit any further test development. A companion Senate bill that has passed the Appropriations Committee would allow pilot-testing to occur on schedule in March. It is silent about what would happen beyond that.

Off the Agenda?

And the president himself may have hurt the case for his testing plan last week. At an Aug. 31 education round table, he didn’t mention national testing as one of his priorities for the fall. A White House summary of Mr. Clinton’s education agenda also avoids the issue. Instead, the president and the document emphasized class-size reduction and school construction.

The silence on testing led its opponents to speculate that Mr. Clinton would concede defeat on the issue.

“Documents we saw like this earlier this year and last fall had testing all over it,” said Jay Diskey, a spokesman for the leading foe of the Clinton plan, Rep. Bill Goodling, R-Pa., the chairman of the House Education and the Workforce Committee. “It’s puzzling to us,” he said of the proposal’s omission.

The Aug. 31 statements should not be interpreted as a change in direction, said Julie Green, the press secretary for Secretary of Education Richard W. Riley. Ms. Green said the administration stands behind an earlier August statement that says it “strongly objects” to the House spending bill because it “would bring a halt” to test development.

If test development extends past Sept. 30, it is clear that it won’t be as easy as President Clinton assumed when he proposed national testing in his 1997 State of the Union Address. (“Clinton Gives Top Billing to Education Plan,” Feb. 12, 1997.)

By March 1999, he said then, the federal government would have new assessments in 4th grade reading and 8th grade math. The tests would yield specific individual scores that would match the achievement levels set by NAEP, the existing national assessment, which tests only a sampling of students and does not provide individual scores. The new tests would be available to all children.

Mr. Clinton and his team assumed that NAEP simply needed tinkering to create an exam that yields individual results. But the assessment board has found that it can’t be easily tailored.

Take, for example, the problems in defining individual students’ achievement based on their answers to 45-minute portions of the NAEP exams. Because NAEP is a sampling of student achievement across a state, test administrators divide the assessment into sections for several groups of students to take. No student answers more than a small percentage of the questions on the overall test.

The expansive range of questions and the large sample size allow NAEP contractors to specify where groups of students score on the achievement levels of advanced, proficient, basic, and below basic.

But the proposed tests are intended to be given in two, 45-minute sections, which may not generate enough questions to pinpoint a student’s achievement level.

“I suspect that we’re going to end up saying: ‘We’re 90 percent sure ... that [a child’s] score would be in the proficient level,’” said Mark D. Musick, NAGB’s chairman and the president of the Southern Regional Education Board in Atlanta. “Is that going to be good enough? My view is that it could be.”

Quicken the Pace

AGB will be discussing the issue throughout the fall, but the National Research Council’s evaluators say the board needs to speed up its decision. “A clear vision of how results will be reported should drive rather than follow other test-development activities,” the research council’s report says.

The governing board also has failed to make timely decisions on how to accommodate the needs of children with a limited command of English or with disabilities, the report contends.

“In our judgment, a major opportunity for improved large-scale assessment is being lost in NAGB’s conservative approach to inclusion and accommodation” in the proposed tests, the evaluators say.

For pilot-testing, NAGB will rely on teachers and principals to decide which students are incapable of taking the tests. It also will offer large-print booklets, extended time, bilingual dictionaries, and other accommodations.

Such policies don’t break new ground, the National Research Council experts say in the congressionally mandated report. “They essentially involve retrofitting established assessment instruments and procedures to special populations of students,” the report says. “Another approach would be to design and develop assessments from the beginning that are accessible to and valid for all students.”

The assessment governing board will hold hearings on the issue this fall, focusing mainly on the issue of languages. But that may be too late, the evaluators say.

The calculator issue is another example of the difficulties of individualizing NAEP. Because the assessment is a sampling, only two-thirds of the students need to be provided with a calculator for the NAEP test in math. What’s more, the assessment isn’t given on the same day in every state. Those circumstances allow test administrators to ship calculators from site to site, and each device is used more than once every year.

But for the proposed new math test, every student would need to be lent a calculator. Because the exams would be given during the same week, the calculators would be used by just one student annually. That means NAGB would need to provide a calculator to every student taking the test. NAGB has made no decision about them for the proposed full-scale tests.