Retention and High-Stakes Testing, Accountability and
Some Alternatives to Existing State Policies
February 20, 2003
Retention and High-Stakes Testing
In recent years Texas and other states have promoted the concept that school accountability is best determined by testing students. Testing of students is not a new idea. Students have been tested for decades using both locally-developed and standardized tests. But a new dimension has emerged in using a single test to make decisions concerning whether a student gets promoted to the next grade or whether a graduating student will receive a diploma.
Tests are increasingly used to make high-stakes decisions about students based solely or primarily on test scores. The push for using state test scores as the primary basis for promotion, retention and graduation decisions is based on the incorrect assumption that a single test score tells you all you need to know about student achievement.
Recent research on the Texas testing program, however, reveals that improvement in state test scores did not simultaneously result in higher test scores on national tests, and that despite rising state test scores, Texas students were not graduating in higher numbers or increasing their enrollment in college.
On the other hand, research has shown that students who are retained in grade do no better the next year. In many cases, retention leads students to drop out before they graduate.
Given evidence on the ineffectiveness of high-stakes testing, why do people insist that we do it? Part of the answer is that some think we must test all pupils in order to hold schools accountable for producing good results.
Distinguishing Between Student Testing and School Accountability
Given public interest and investment in public schools, it is appropriate and important that information be compiled that gives insights into how well schools are performing. Doing school performance reviews however, does not mean that all students have to be tested. For example, when testing for water quality, state officials do not test whole rivers, streams or lakes. They draw samples from those bodies of water that give them a relatively accurate measure of the quality of those systems.
Texas could continue to measure its schools’ performance, and do it much more efficiently and at much less expense, by moving to an assessment system that tests a sample of students from each school to get a picture of how each school is performing.
What Is Needed
Texas must change its testing policy in two important ways. First it must reject the use of any single test or other measure as the primary basis for making student promotion or graduation decisions. Even test makers point out that no single measure should be used in making important educational decisions. Grades, teacher opinions, and parent judgment should all be factors taken into consideration when making important life-altering decisions for pupils.
Second, Texas should abandon the inefficient practice of testing all students, and move instead to a process that uses a stratified, random sample approach that can yield the same information on school quality and performance at a fraction of the current multi-million dollar annual cost.
IDRA is an independent, non-profit organization that advocates the right of every child to a quality education. For 29 years, IDRA has worked for excellence and equity in education in Texas and across the United States. IDRA conducts research and development activities, creates, implements and administers innovative education programs and provides teacher, administrator, and parent training and technical assistance.