House Blocks, While Panel Settles On, New Tests
House lawmakers last week tossed President Clinton's highest education priority into political limbo.
The House voted overwhelmingly Sept. 16 to cut off federal funding for the creation of national tests in 4th grade reading and 8th grade math--just hours after a panel of educators and policymakers approved the design and content of the voluntary exams.
The Senate had already voted to support the tests, which are to be given for the first time in March 1999. The two chambers will have to settle their differences in a conference committee, leaving the tests' fate uncertain for now. ("Compromise Is Next Step in Testing Odyssey," in This Week's News.)
Members of the national test panel, convened under a Department of Education contract, were well aware last week that their months of work could end up moot thanks to congressional politics. But if the proposal for new tests measuring individual students' academic achievement survives, the committee's report could carry great weight: Its specifications are to be used by the contractor to write the tests that could be taken by hundreds of thousands of children.
Apart from actions pending in Congress, the panel's decisions have a "99 percent chance of surviving and becoming the national test," Gary Phillips, who is directing the national test effort at the Education Department, told the group at its two-day meeting here.
Even as the national test issue stirred debate on Capitol Hill, the panel's own deliberations were also contentious at times. Some panelists expressed interest in submitting dissenting viewpoints for inclusion in the group's final report to the department. They want to make clear their opinion that the tests should not be used to rank schools or determine grade promotion or retention.
Similar to the philosophical differences that have split educators and policymakers across the country, the panelists also sparred over how much of the "basics" should be incorporated into the tests, particularly for math.
In working through specifications for the 4th grade reading and 8th grade math tests--including the types of test questions, length of test periods, and methods of reporting scores--the panel wrestled with many of the same testing issues that perennially vex educators. The use of calculators on the math test and accommodations for students with disabilities or limited proficiency in English on both tests were two topics that prompted lengthy debate.
In the end, the group voted narrowly to recommend, but not require, the use of calculators. The panel tied 5-5 on the issue, but Chairman Wilmer S. Cody, the commissioner of education in Kentucky, broke the tie by voting to recommend use of the electronic aids.
The idea the panel endorsed is that of allowing students to employ whatever tools they are accustomed to using in the classroom. The math test could also be taken without a calculator.
A separate math committee presented the national panel with data from the 1996 National Assessment of Educational Progress that showed 83 percent of students nationally receive instruction in how to use calculators, and 67 percent of students use calculators on tests. The panel also heard that in the 1995-96 school year, 31 states said they permitted students to use the devices on some or all parts of their statewide assessments.
The test panel was charged by the Education Department with using the existing NAEP as its testing blueprint. And, in large measure, the panel was faithful to both the letter and the spirit of the national assessment's framework. But the specifications for the proposed new tests must differ from NAEP for technical reasons because the new tests would provide the scores of individual students, which NAEP does not.
On some matters, though, it was up to the panel whether to part from NAEP practices. The calculator issue was one on which it did so. The national assessment, mandated by Congress and given since 1969 to a sampling of students in several core subjects, gives math tests that allow calculator use on some parts and forbid it on others. The students taking part in NAEP are also provided with calculators to use on the math test. Students using calculators on the new national test would use their own or those provided by their schools.
On the reading test, the panel also made a significant departure from practice on NAEP's 4th grade reading test. The new national test is to include "intertextual items" that ask students to answer questions by drawing comparisons about two separate passages they have read on the test.
Such items are used on the NAEP reading tests for 8th and 12th graders, but not for 4th graders, "due to time constraints," according to the reading committee advising the panel. With each student spending more time on the national test than on NAEP, the reading committee wanted to include that type of item.
The national panel also endorsed going beyond NAEP in making a bilingual English-Spanish math test available for students who need it, as well as providing some accommodations on the new tests for disabled students and those with limited English proficiency. Advisers to the panel pointed to the use of such assistance on NAEP, but to date, that help has only been provided in small, experimental administrations of the national assessment.
Marking Time and Rigor
For most of its decisions, the 18-member national test panel largely followed the recommendations set forth by the math and reading committees, made up of teachers, administrators, and academic experts in the subjects. The two committees, whose memberships were separate from that of the test panel, held public hearings and came up with sample test items. ("Hearings Draw Few, But Varied, Remarks," Sept. 3, 1997.)
The test panel agreed with the math committee's advice that the math test should be designed so that most students could complete it in two 45-minute sessions, one on each of two consecutive days. But students would be allowed 60 minutes each day to complete the 90-minute test. The committee had earlier considered an untimed test.
"We're not trying to measure who is the fastest student," John A. Dossey, the chairman of the math committee and a professor of mathematics at Illinois State University in Normal, told the test panel.
The panel accepted the reading committee's recommendation for two 45-minute sessions over two consecutive days for the 90-minute reading test. Both subject tests are to include a mix of multiple-choice and open-response items.
The panelists also agreed with the recommendation that the reading and math tests should evaluate a range of ability levels. In addition to having rigorous items, the tests are to include questions that are somewhat easier than those on NAEP. That decision is meant to address longstanding concerns among some educators that NAEP does an inadequate job of measuring abilities of the lowest-achieving students.
Panel members also discussed how to address critics who have called for the math test to emphasize basic, computational skills rather than some of the more complex or conceptual thinking endorsed by the National Council of Teachers of Mathematics.
Debate centered for a time on adding items that would measure only basic skills. Members of the math committee, including the president of the NCTM, Gail F. Burrill, argued successfully that the items they suggested for the test measure a variety of skills and that the test both assumes and requires basic computational ability.
In a letter separate from its 150-page report, the reading committee recommended--unsuccessfully--that the name of the test be changed from the "Voluntary National Test in 4th Grade Reading" to the "Voluntary National 4th Grade Test of Reading in English."
The committee wrote in its letter that it wanted to "send a clear message that competence in reading in English is crucial ... without regard to what a child's first language is."
Members also wanted to make the point, according to the letter, that "for multilingual children the ability to read in English represents one important dimension of their ability to read, but not necessarily all aspects of their reading ability" because they may have literacy skills in other languages. Renaming the test, they said, would "more accurately characterize the nature of the test."
The proposed name change for the reading exam drew negative reactions from several members of the national test panel, including one representing urban districts.
Sharon Lewis, the director of research for the Council of the Great City Schools, a Washington lobbying group for big-city districts, distributed a letter to panelists that stated her group's opposition to the creation of a reading test only in English.
Ms. Lewis said the reading committee's proposal could curtail ongoing negotiations with the Education Department on the creation of reading tests in languages other than English. "We don't consider it a closed issue," she told her panel colleagues, "so I couldn't support renaming the test."
In an interview last week, panel member Arnold Fege, the director of governmental relations for the National PTA, agreed with Ms. Lewis.
"The federal government cannot send the message that we have developed very important educational measures for all children except for 3 million that may not read English," Mr. Fege contended. The PTA could not accept a national test of reading only in English, he said, adding that the issue numbered among the "bottom-line deal-breakers for us."
He said that he planned to submit a report this week to be included as a minority viewpoint in the panel's final report.
Mr. Fege also tried unsuccessfully to persuade the panel to clarify the purpose of both tests and their use. Among other changes, he wanted to prohibit states and districts from being able to rank students or schools on the basis of national test results. But he only succeeded in getting the national panel's reading report to reflect that the results are "not intended" to rank students or schools.