High Stakes: Test Truths or Consequences
It started as a method of motivation. Attach important consequences to individual performance on state achievement tests, officials in one North Carolina district reasoned, and students will take their studies, and the end-of-grade exams, more seriously.
The result, the officials say, has been an intensified focus on teaching the state curriculum, around which the assessments were designed, and a greater proportion of students performing at grade level in core subjects.
But some parents and testing experts are crying foul. Children who were held back a grade for failing to prove their proficiency on the reading and math tests, they argue, should not have been judged so heavily against a single measure when their academic futures are at risk.
As debate continues to swirl around President Clinton's plan for voluntary new national tests, and as more states turn to so-called high-stakes assessments of their students, both critics and proponents of such assessment plans are predicting that, inevitably, the results will be misused.
A Prime Example?
They are pointing to the Johnston County, N.C., district, which has been hit with a class action as a result of its policy, as a case in point.
"This is a prime example of a test that was developed for one purpose ... and applied for a purpose that is totally inappropriate and unintended," said Walter M. Haney, a professor of education and the acting director of the Center for the Study of Testing at Boston College.
"The North Carolina end-of-grade tests were designed to hold schools and school districts accountable," Mr. Haney said, and not individuals. "There is considerable potential for people trying to use the national test for similar decisions without stopping to examine whether, in fact, the content parallels the local curriculum."
Looking for a way to make students--and parents--more responsible for learning, Johnston County school officials passed a student-accountability policy last year. The policy, which went into effect with the test results released this past summer, calls for intensive remediation, and then retention, if necessary, of students who do not score at a proficient level on the state exams.
"We had a concern that kids were being moved along without being prepared to do the work at the next grade level," said James F. Causby, the superintendent of the 18,000-student district south of Raleigh. "There were no consequences for kids who didn't do well on the tests, and we weren't giving them the assistance they needed to succeed."
Soon after the results were released, a multiracial and multiethnic group of 14 parents filed a lawsuit against the district on behalf of their children and other students who were set to be held back under the policy. The plaintiffs argue that the tests, which are intended to rate districts and schools, are not valid for measuring individual performance.
"This was not the type of test that was intended to be used for this decision," said Kimberly West-Faulcon, a staff lawyer for the NAACP Legal Defense and Educational Fund, which is representing the parents. "It is a bad tool to identify kids in need of help, because using the test for individual assessment hasn't been validated."
Only last week, Texas officials were sued over the fairness of their high school exit examination. ("Discrimination Claimed In Texas Exit-Exam Lawsuit," in This Week's News.)
In North Carolina, Johnston County officials defend their policy, saying the tests are the best vehicle for improving schools.
"From our perspective it has worked wonderfully," said V. Vann Langston, the district's assistant superintendent for curriculum and instruction. "I'm in my 30th year in education, and I've never seen an entire school system as focused on teaching the curriculum and ensuring that all the youngsters are mastering the curriculum."
Students in the district and throughout the state are given one of three different versions of the North Carolina End-of-Grade Test, which was first administered in 1992, according to Lou Fabrizio, the state testing director. Students in the same classroom may be given different versions of the test. ("N.C. Gets First School-by-School Performance Results," Sept. 3, 1997.)
Combined, the tests can generally provide an accurate account of how well the curriculum is being taught in a school or district. Although one version does not provide for an accurate individual assessment, when a student takes all three versions of the exam, a clearer picture emerges of how well he or she has mastered the material, Mr. Fabrizio said.
Retesting and Remediation
Johnston County students who don't make the grade the first time around are offered several opportunities for remedial help and are given a different version of the test, one approved by the state for local use. Those students who continue to perform below grade level can be retained in their current grade.
According to state guidelines, a student testing at Level III Proficiency--the target for all Johnston County students--consistently demonstrates mastery of grade-level materials. The state school board calls for intervention--which may include remediation or retention--for students who do not meet that standard, Mr. Fabrizio said.
At first, it looked as though more than 2,600 Johnston County students would be affected by the policy after they failed to meet the standard on the first try. But after students repeated the tests or successfully appealed their retentions, the number held back dropped substantially, to about 500, or 6 percent of students in grades 3 through 8. The appeals were based on students' good grades throughout the year.
Johnston County officials claim the accountability program has already boosted student performance. Four years ago, more than one-third of students performed below grade level on the tests, yet just 1 percent of students were held back. More than three-fourths of the students scored well on the tests this year, while 8.8 percent of students were retained under the policy--and for other reasons, such as absenteeism. The other 16 percent were promoted based on the grades they earned and other academic factors.
But the plaintiffs say that the policy is unfair to students who do not perform well on tests in general, and is especially detrimental to minority and special education students. While 245, or 4.1 percent, of white students in the district's elementary grades were retained as a result of the policy, 207 black students, or 11.4 percent, were held back, as were 41 Hispanic students, or 10.2 percent.
The policy is modeled after one instituted several years ago in Transylvania County, south of Asheville, N.C., that has gone unchallenged. In that district, which is smaller and less diverse than Johnston County, student performance also has seen a significant boost, officials there say.
For that very reason, the National Association of State Boards of Education recommends, in a report released last week, attaching high stakes, such as high school graduation and grade promotion, to students' performance on state assessments. The report, however, emphasizes that the assessments should be just one part of an accountability plan. A successful high-stakes test should also be legally defensible and not adversely affect minority students, the report says.
Testing experts have urged officials to forbid high-stakes uses for the tests that Mr. Clinton has proposed for 4th grade reading and 8th grade math. So far, the U.S. Department of Education has not supported such restrictions.
"It would be desirable to have a prohibition, or at least guidelines, that say to states they are not to use the tests for reasons that were not intended," said Robert L. Linn, a professor of education at the University of Colorado and a co-director of the National Center for Research on Evaluation, Standards, and Student Testing, or CRESST, at the University of California, Los Angeles.
But U.S. Secretary of Education Richard W. Riley believes that such decisions must be made locally. The test results, however, "should not be used for high-stakes purposes ... unless they have in fact been validated for those purposes," Mr. Riley wrote in a letter last month to the Leadership Conference on Civil Rights.
Wade J. Henderson, the executive director of the Washington-based advocacy group, had expressed concern that minority students would suffer most if districts were not prohibited from linking promotion or graduation to the new tests.
Guidelines acknowledging the need for validating the tests for other purposes would be written by the test contractor and approved by the National Assessment Governing Board, the independent panel that oversees the existing National Assessment of Educational Progress, Mr. Riley's Sept. 11 letter said.
Mr. Riley's top deputy acknowledges that data from the new national tests would be compiled by local officials to make comparisons that may be unintended, such as the scores of schools and even classrooms within them rather than individual performance. If districts didn't do the analysis, newspaper reporters would, according to Marshall S. Smith, the acting deputy secretary of education.
"All of that I expect to happen because that's the way people do their business," Mr. Smith said in a recent interview.
For Johnston County, the decision may lie with the courts. In August, a U.S. District Court judge rejected the plaintiffs' request for a temporary injunction to prevent the retention of students this school year. District officials point to that action as a hopeful sign in the pending case.
"The judge's refusal to grant an injunction and the language he used," Mr. Langston said, "was very supportive of our position."
Staff Writer David J. Hoff contributed to this report.