Test Use Intensifies in Colorado, Texas

K-12 Testing

Like an out-of-control freight train rolling down a hill, test-driven schooling is gathering speed as high-stakes tests spread across the nation. The lack of evidence that such pro-grams actually produce improved learning does not appear to deter policymakers from implementing these programs, nor does the likelihood that they narrow and dumb down the curriculum to the long-term detriment of student learning.


In Denver, officials in school after school told a reporter for the Rocky Mountain News how they were revamping curriculum and instruction to fit the city's tests. In one school, "Teachers acclimated students by teaching lessons in formats similar to what they see on tests." The district's tests are entirely multiple-choice.

Last year, Colorado began introducing state tests that are not entirely multiple choice. Scores on these tests, however, are very low. On a four-point scale, few students scored in the top two levels. A 1996 law prohibits schools from promoting students from grade 3 to grade 4 if they do not pass the state reading test, although other data, such as students' classroom performance, can override test scores. Now, widely-backed legislation is pending that would make passing Colorado's tenth-grade test mandatory for high school graduation and would use test results to remove accreditation from low-performing schools which do not raise their scores.


Supporters of high-stakes tests point to Texas as a success story. Scores have risen on the state test, the Texas Assessment of Academic Skills (TAAS). The number of students reaching the "proficient" level on the grade 4 National Assessment of Educational Progress (NAEP) math exam also rose significantly from 1992 to 1994, though not on the grade 8 math exam.

However, a FairTest study showed that of the 14 cases of significant increases in the number of proficient students, only four were in states with a high stakes test, such as Texas (Examiner, Winter 1997-98). In fact, states with high stakes tests were slightly less likely to show significant gains. Additionally, Texas was below the U.S. average for percentage of students reaching proficiency on the 1996 grade 8 NAEP math and 1994 grade 4 NAEP reading tests, though TAAS has been in place since the 1990-91 school year.

The recent release of test results on the Stanford Achievement Test-9 (SAT-9) for Houston has cast further doubt on the meaning of rising TAAS scores. Houston scores were well below national averages on this norm-referenced, commercial achievement test, surprising Houston educators who had expected their students to perform better, based on TAAS scores.

Some observers suggested the reason for the discrepancy is that TAAS is too easy. However, Houston teachers probably are teaching to the state exam, as promoters of high-stakes testing support, but have not taught to the SAT-9. A serious mismatch between the city's curriculum and the SAT-9 could contribute to low SAT-9 scores, but it is most likely that the TAAS scores are inflated from teaching to that test. Neither the SAT-9 nor the TAAS multiple-choice tests indicate much about students' ability to think and use knowledge in tested subject areas. Neither should be used to control curriculum and instruction.

Despite the lack of evidence that high-stakes tests produce real learning gains and strong evidence that high-stakes tests spur teaching to the test that narrows curriculum and produces inflated scores, scores on TAAS now will be used to evaluate the state's 250,000 teachers.

Texas Gov. Bush also has been pushing a proposal to deny grade promotion to students in grades 3, 5 and 8 who do not pass the TAAS. The proposal must be approved by the legislature and, if passed, would first affect students who are kindergartners in 1999. Bush's opponent in this fall's governor's race, Garry Mauro, opposes additional high-stakes uses of TAAS.

Meanwhile, the Mexican American Legal Defense and Educational Fund (MALDEF) has filed suit against the TAAS, arguing it discriminates against minority students, many of whom have not received an adequate or equitable opportunity to pass the tests (see Examiner, Fall 1997). Minority students fail the test at a far higher rate than do white, Anglo students. The case is scheduled for trial this November.