Standardized tests don't measure education quality

Standardized tests are the cornerstone of Ohio’s education “accountability system.” Test results are the dominant measure used to create report cards that judge the quality of education offered in Ohio’s schools and school districts and to shame and blame low performers.

Last year the state switched to the Partnership for Assessment of Readiness for College and Careers (PARCC) tests to measure school quality. The tests are aligned with the controversial Common Core standards. In July, the state rejected Common Core and jettisoned the PARCC tests. The 2015–16 measuring stick will be standardized tests created by the American Institute of Research (AIR).

There is still no report card for 2014–15. In January, after a six-month delay, the department of education released the non-test-based information that will go into the report cards. The reporting delay is blamed on calibrating what scores on the PARCC tests will receive what grade. This data will not be available until the end of February. Gratefully, full-blown report cards and their attendant sanctions will not be used until 2018. I hope school and district report cards never again see the light of day.

No matter how many times the state changes the standardized test it uses to measure student learning, the accountability system will still be unfair and an inaccurate evaluation of education quality.

Our accountability system uses standardized tests to measure student learning. The quality of education is measured in large part by the proportion of children who pass the standardized tests. Unfortunately, testing experts say standardized tests do not measure quality.

Basing a high-stakes system on an inaccurate tool has been called “accountability malpractice.” It has caused harm without making schools better places for children.

In his March 1999 Educational Leadership article, “Why Standardized Tests Don’t Measure Education Quality,” James Popham observed, “Asserting that low test scores are caused by the quality of education is illogical.” Popham identifies three things that influence test scores: what is taught in school, a student’s native intellectual abilities, and a student’s out-of-school learning. Tests really don’t measure what schools do.

The problem is twofold. Many variables outside the school affect test performance, and standardized tests are designed to compare students, not to evaluate their education.

A study cited in 2013 by Edward Haertel suggests that teachers are responsible for about 10 percent of student test performances, while 60 percent of what is measured is caused by out-of-school factors. The other factors include family background, neighborhood environment, peer influences and educational aptitude. I would add that unreliable computers and the test-takers' ability to endure a long test or apply themselves to endless tests are also factors.

Schools are only partly responsible for test scores, but the accountability system ignores this and uses scores to evaluate school quality. It doesn’t make sense.

Standardized tests are commercially developed assessments designed to rank and compare test-takers from across a large number of schools, districts and even states. Within a manageable amount of test-taking time, test items are selected that help differentiate high and low performers. Low performance is built into the test design! The tests exclude questions that the majority of test-takers might know—information teachers agree is important to teach. The tests sample the curriculum so there is no guarantee that what is being tested has been taught to everyone who is taking the test.

“To evaluate teachers’ instructional effectiveness by using assessment tools that deliberately avoid important content is fundamentally foolish,” wrote Popham.

Norm-referenced tests are a weak measure of quality, because their goal is to rank and compare kids, not measure mastery of specific content. Test scores do not tell us how much our children are learning and what kind of job our schools are doing, but we still use them to judge quality.

The fact that test designers use questions that rely on prior knowledge should also invalidate this system. Popham further observed, “One of the chief reasons that children’s socioeconomic status is highly correlated with standardized test scores is that many items on standardized tests really focus on assessing knowledge and/or skills learned outside of school—knowledge more likely to be learned in some socioeconomic settings than others.”

Testing experts know that standardized tests are the wrong tool for measuring the effect of schools on children. Unfortunately, the governor and legislature ignore this technical detail and persist. They deserve an F for their stewardship of our schools.

We can do better: Invest in teaching, not tests.

Susie Kaeser

Susie Kaeser is a longtime resident of Cleveland Heights and former director of Reaching Heights. She serves on the national board of Parents for Public Schools.

Read More on The Common Good
Volume 9, Issue 2, Posted 11:30 AM, 01.29.2016