Achievement in the elementary grades is by some measures at its highest level in three decades, according to a report on testing trends released this week by the Congressional Budget Office.
The report, Trends in Educational Achievement, states that the decline in test scores among students in the upper elementary grades actually ended as early as the mid-1970’s, despite a “widespread misconception” that test scores stopped declining only a few years ago.
This trend in student performance in the early grades gained “relatively little attention,” the report notes, "[p]erhaps because of the greater notice accorded to tests at the senior-high-school level.”
The report is not particularly sanguine, however, about the overall level of student achievement. The report includes data showing that American students still fare poorly on international comparisons, and that many American students are not performing well even on fairly easy questions.
And although the test scores of high-school students have also improved, the C.B.O. analysis points out, scores on such examinations as the Scholastic Aptitude Test remain relatively close to their low points of he late 1970’s.
On some tests, gains also appear to have been greater on lower-level as opposed to higher-level skills, such as problem-solving. But the study notes that there have been improvements in both areas. For example, the percentage of S.A.T. mathematics scores above 700 rose steeply between 1980 and 1984, although it was still small in absolute numbers.
The report also concludes that the gap in test scores between black and nonminority students has been “slowly but appreciably narrowing in recent years.”
Although the difference in the average test scores of black and nonminority students remains large, improvements have been consistent from year to year, it states, “and could prove substantial over the long run.”
The test scores of Hispanic students have also improved, although there is less data available on this group, according to the C.B.O.
There have also been substantial gains in the performance of students from inner-city schools, the report says, noting that by 1981, a fourth to a third of the gap in test scores between disadvantaged urban communities and the rest of the nation had been overcome.
The budget office prepared the study at the request of the Subcommittee on Education, Arts, and Humanities of the Senate Committee on Labor and Human Resources.
The first volume of the study assesses trends in the achievement-test scores of American elementary- and secondary-school students over the past 30 years, based on a variety of measures--including the National Assessment of Educational Progress, the “High School and Beyond” study, the S.A.T., the American College Testing Program, the International Association for the Evaluation of Educational Achievement, and the annual Iowa assessment of student achievement.
A companion volume, Educational Achievement: Explanations and Implications, will be published this summer or fall. It will examine the most common explanations for the changes in students’ test scores, and the implications of those changes for educational policy.
Daniel Koretz of C.B.O.'s human resources and community-development division, prepared the analyses for both volumes.
‘Sizable Drop’
According to the report, scores on achievement tests first began a “sizable drop” in the mid-1960’s.
“The decline was widespread,” the report says, “occurring among many different types of students, on many different tests, in all subject areas, in private as well as public schools, and in all parts of the nation.”
The largest drops tended to occur for students in the higher grades. At the end of the decline, for example, he typical student in grades 6 and above performed at a level comparable to the 38th percentile of students before the decline began.
The report cautions, however, that the size of the decline differed so greatly from test to test that a different mix of tests--or a larger and more representative sample of tests--might have yielded a very different figure.
In general, test scores for students in grades 3 and below dropped little, if at all, during the same period.
The study found that test scores first began to improve with those children who entered elementary school in the late 1960’s, with each subsequent group of children typically outscoring those who began school the year before them. Test-score improvements have also been moving into the higher grades as these cohorts of children have aged.
Trend data from the annual Iowa assessment indicate, for example, that scores for students in grades 3 through 6 are at their highest point in 30 years; those for students in grades 7 through 10 have rebounded strongly, but are not yet at the previous highs for those grades; and scores for students in grade 12 have begun rising, but remain near the low point.
Upswing Ended?
Recent data from NAEP suggest, however, that the upswing in test scores is, for the moment, over in the youngest age groups and may end fairly soon in the higher grades, the C.B.O. reports.
Scores on NAEP tests administered to 8th graders are expected to level off between 1983 and 1987, while scores of seniors would level off between 1987 and 1991.
Because no other national data are available, the report cautions that it is “not clear” whether this leveling off is a general phenomenon.
It also suggests that although declining test scores during the 1960’s and 1970’s have often been cited to justify recent educational initiatives--some of which have even been based on theories about what caused those declines--many of the reform efforts “are not fully consistent with either the trends or the limited information on their causes.”
For example, the report notes, a number of recent proposals are aimed at improving curricula in the basic skills, even though test scores show some of students’ greatest problems to be in higher-order thinking skills.
The report adds that “the results· of standardized tests must be interpreted cautiously.” In particular, it says, trends that appear only on one test, or only among a set of very similar tests, “should be considered questionable.”
Copies of the report can be obtained free of charge by calling the Congressional Budget Office’s publications division at 202-226-2809 or writing to C.B.O. Publications, C.B.O., U.S. Congress, Washington, D.C. 20515.
