Bipartisan interest in the uses of standardized tests in education has led the Congress to launch its own study of their validity and applications.
The ambitious $200,000 study, launched in January by the Office of Technology Assessment (ota), one of four investigative arms of the Congress, is intended to serve as a guide for federal policymakers to determine what steps the government might take to sponsor, monitor, or enhance the use of tests in education.
The study comes at a time when state officials, responding to pressure to improve student performance, are expanding mandated testing programs for promotion and graduation as well as for evaluation. And it reflects a variety of Congressional concerns--from the question of how testing affects minority students to that of whether tests intrude inappropriately into the personal values of students.
“There is growing interest in the Congress in linking federal funding assistance to student performance and other numerical indicators,” wrote Senator Orrin G. Hatch, Republican of Utah and chairman of the Labor and Human Resources Committee, in a letter to ota’s director, John H. Gibbons.
At the same time, another research arm of the Congress, the Congressional Budget Office, has been reviewing indicators of academic achievement in the United States at the request of Senator Robert T. Stafford, Republican of Vermont and chairman of the Senate education subcommittee. The cbo’s report is to be released sometime this spring.
The year-long ota study received initial support from Senators Hatch and Edward M. Kennedy of Massachusetts, the ranking Democrat on the Labor and Human Resources Committee, after it was brought to their attention by the ota’s project director, Richard Thoreson. Mr. Thoreson told the lawmakers his agency wanted to do more work in assessing educational technology and thought Congressional interest warranted a study of testing. In his statement of support to Mr. Gibbons last summer, Senator Hatch--the author of a controversial law to limit the psychological testing of students--said he was concerned about the potential invasion of privacy related to standardized tests. He also said that a federally financed study of testing was necessary because of the increasing use of standardized-test scores to “quantify levels of knowledge and to make comparisons among schools.”
Test scores, he noted, are influential in determining who is included in federally assisted programs, and they are used to gauge school systems’ effectiveness through such federally sponsored programs as the National Assessment of Educational Progress, which will receive about $5.4 million this year.
Senator Hatch also said the need for a national study of testing was underscored by the efforts of former Secretary of Education Terrel H. Bell to compare the academic health of states on his “wall charts,” which highlighted state average scores on the Scholastic Aptitude Test.
Scores ‘Justify Policy’
According to Senator Hatch, “interstate and international comparisons of standardized scores are used increasingly to justify policy initiatives and even to attract private investment in communities.”
But the major rationale for the ota study appears to lie primarily in the growing demand in the Congress that federal funds be linked to measurable outcomes.
In the background paper outlining Congressional interest in the study, Mr. Thoreson wrote that “standardized test scores provide ‘hard’ data about educational out-puts which can be used to make difficult decisions about where federal expenditures can or cannot be cut.”
In an interview last week, he said programs that could be linked more closely to standardized testing include adult education, vocational education, Chapter 1, and programs for students who have limited English proficiency.
High-School Bills
Bills introduced by Democrats in the House and Senate in this session, for example, would establish a program similar to Chapter 1 for high schools in which at least 20 percent of the students come from disadvantaged backgrounds. But the bills--HR 901 and S 508--would make continued funding contingent on two quantifiable outcomes: improvements in students’ scores on a nationally normed test of basic skills and progress in lowering the schools’ dropout rates.
The House bill, sponsored by Representative Pat Williams, Democrat of Montana, would provide $900 million directly to schools each year, while the Senate bill, sponsored by Bill Bradley, Democrat of New Jersey, would develop pilot programs over a two-year period at a cost of $100 million. In the third year, grants of $800 million would be awarded to states.
A ‘Dubious Proposition’
According to William W. Turnbull, distinguished scholar in residence at the Educational Testing Service and a former president of ets, national policymakers for8some time have argued that federal dollars should flow to districts where standardized-test scores are low and where there is obvious need. But today, he said, state leaders are talking about shutting down districts where performance is low.
Mr. Turnbull said that increasing aid or taking more punitive steps on the basis of standardized testing is “a dubious proposition,” because tests are limited in what they can measure and can do.
“We’ve almost reached the state where respect for tests has become a cult,” said Howard Matthews, an aide to Senator Hatch. He said Senator Hatch’s interest in a federal study of testing grew out of the concerns raised in a 1982 report by the National Academy of Sciences titled “Ability Testing: Uses, Consequences, and Controversies.”
That study, widely considered the most influential in the field, concluded that “overreliance on test scores is a widespread problem” in this country. The demand for quantified information tends to “dominate decisionmaking” to too great an extent, the study said.
It concluded that the tests, which are often normed to a general population, may be biased against minority groups; are often used to track students and result in permanent labels; are poor predictors of long-term performance; and are oriented only toward cognitive skills.
Scope of Report
According to Mr. Thoreson, the ota study will build on the findings of the nas report and other studies to point out policy options that the Congress may consider as it assesses the federal interest and role in standardized testing.
The agency, Mr. Thoreson conceded, has little experience in studying technologies related to education; its only previous effort was a 1982 report, “Informational Technology and Its Impact on Education.”
The new study, he said, was originally designed to look at all aspects of testing but has been reduced in scope somewhat. It will review current standardized-testing practices; minimum-competency and promotional-gates tests used by states; the extension and development of the national assessment; and--at the request of Senator Kennedy--ways in which tests are used for placement decisions or lead to self-selection for minority groups, particularly for students with limited proficiency in English.
Over the past few months, the ota has summoned testing experts from around the country to Washington to gain their perspectives. This month, the project’s organizers will begin to consult organizations representing teachers, administrators, superintendents, and other groups.
Observers have expressed con4cern that the scope of the project is too broad in light of the resources available. Budgeted at $200,000, the project has two full-time researchers.
“The ota reports carry significant amount of weight on the Hill, but the kinds of problems and questions they will be addressing are those that testing experts have been continually grappling with for years,’' said Laurie Garduque, director of governmental and professional liaison with the American Educational Research Association.
She said it would probably be more useful for the agency to limit the scope of its study to an analysis of a single research topic rather than taking on so many issues at once.
According to Mr. Turnbull of ets, the Congressional agency has the potential to assess “the adequacy of technology for the kinds of decisions people are asking it to make.”
But Mr. Turnbull said he doubted that the study would lead to substantive changes at the federal level, other than improving measurements used in assessing the academic health of individual states and the nation.
“As long as the education process is a state and local function, I can’t see that there is a federal role in testing other than in major data-gathering functions represented by naep or by ‘High School and Beyond’ and the National Center for Education Statistics,” he said.
