To Maintain Trends, NAEP Scales Back Math-Test Changes
WASHINGTON--Citing the need to be able to track student performance over time, the National Assessment Governing Board has adopted a scaled-back version of a proposed revision of its mathematics assessment.
The changes, which are expected to guide the development of the 1994 National Assessment of Educational Progress, are aimed at bringing the math test more in line with standards for instruction in the field developed by the National Council of Teachers of Mathematics.
The original version of the proposal called, among other changes, for increasing the number of open-ended questions and placing a greater emphasis on topics such as algebra.
But members of the N.A.G.B., meeting here this month, said the proposal would make the 1994 assessment too different from the 1992 assessment to enable NAEP to show trends over time.
The board approved several of the proposed changes. But it rejected the increase in open-ended questions and scaled back some of the topic shifts.
The decision represents a "difficult tradeoff'' that the panel is
likely to face increasingly as educators in other subject areas move
toward developing national standards, observed Chester E. Finn Jr., a
professor of education and public policy at Vanderbilt University and a
former chairman of the NAEP board.
On the one hand, NAEP must "maintain a scale to give us trend data,'' Mr. Finn said. "But we don't want the scale to be a permanent impediment.''
'Networks of Connections'
A Congressionally mandated project, NAEP is the only assessment program that can provide reliable national data on student performance trends over the years.
In January, for example, the National Center for Education Statistics, which oversees NAEP, released a report showing trends in science, math, reading, and writing during the past two decades.
In the past few years, however, NAEP has revamped its test frameworks to reflect changes in curriculum and instruction.
To prepare for the 1990 assessment--the first to provide state-by-state comparisons of student-achievement data--the Council of Chief State School Officers led a project to overhaul the math framework to make it align with the standards that were being developed by the math-teachers' council.
But critics said the changes did not go far enough. So NAEP's governing board contracted with the College Board to bring the 1994 test closer into line with the N.C.T.M.'s standards.
In presenting the proposed revisions to the board, John A. Dossey, the former president of the council and a consultant on the project, said the changes would emphasize students' abilities to solve math problems, to reason and communicate mathematically, to have confidence in their math abilities, and to value math as a discipline.
The proposals, according to a draft presented to the N.A.G.B., "reflect the increasing realization that student proficiency in mathematics is not the result of the interaction of discrete cells of knowledge with a discrete list of special mathematical abilities.''
"Rather,'' it states, "student proficiency in mathematics results from broad experience in forming networks of connections among mathematical ideas and skills.''
A New Framework
Specifically, the original proposal called for:
- Scrapping the "matrix'' framework that divided test items by ability levels, replacing it with one that simply separates items according to mathematical content.
- Creating "families'' of items that would enable NAEP to probe students' ability to define a concept, to apply the concept in a familiar setting, to use the concept to solve an new problem, and to generalize their knowledge about the concept.
- Shifting away from an emphasis on numbers and operations, and increasing the proportion of test items on algebra and data analysis, statistics, and probability.
- Increasing to 50 percent the proportion of open-ended items on the assessment. Currently, such items represent 50 percent of students' time on the assessment, or about 30 percent of the test questions.
- Allowing unlimited use of calculators.
In addition, the College Board called for a special study to measure 12th graders' abilities to use hand-held graphing calculators.
Such a study, Mr. Dossey said, would provide "baseline data on the most significant curricular change to take place in the '90's.''
Not 'Leading the Charge'
Although the members of the NAEP board said they supported the thrust of the recommendations, they contended that they could not accept all of the proposals while still maintaining consistency with the 1992 assessment.
Roy E. Truby, the executive director of the board, said changing the assessment would harm the ability to track progress toward the national education goals. The National Education Goals Panel, he noted, has decided to use NAEP results to gauge progress toward the goal of ensuring that students demonstrate "competency in challenging subject matter.''
"We told the goals panel we would give them a benchmark in 1990,'' he said. "Now, if we tell them in 1994 there will be a new benchmark, they will be very unhappy.''
To preserve the trend lines, the board voted to scale back some of the shifts in content emphasis, and to reject the proposed increase in time devoted to open-ended assessments. The panel also turned down the idea of allowing unlimited use of calculators, voting instead to conduct a special study to test such use.
Michael Glode, a member of the Wyoming state board of education and a member of the N.A.G.B., said the panel will continue to be forced to make such compromises as national standards for curricula and instruction are developed.
"We're aware of the fact that we'd like to be leading the charge, but we're probably not going to,'' Mr. Glode said.