We must walk fine line when it comes to testing

By Camilla Benbow

(Originally printed in The Tennessean on September 12, 2013)

Public education is not immune to the growing emphasis on data we see in other sectors. Many of the education reforms popularized in recent years promote using data to drive decision making at system, school and classroom levels. In turn, much of the data used by educators comes from testing.

In 2001, No Child Left Behind installed a new regime of annual testing in schools to demonstrate student gains. More recently, standardized test data has begun to be used to evaluate teacher performance.

But wariness about the number and frequency of achievement tests — and the stress they cause in children — has begun to crop up. Groups have formed around the country to encourage parents to opt-out of subjecting their children to yet more tests. One local Tennessee group, Stop the TN Testing Madness, has garnered nearly 300 Facebook “Likes.” Metro Nashville Public Schools decided recently that it will not administer the Stanford Achievement Test to K-2 students this year, and it is studying the number of other standardized tests it regularly administers.

Opponents say excessive testing narrows the curriculum. Endless test preparation, critics say, means less actual instruction time and fewer subjects covered. They point to countries such as Finland that test far less than we do but whose students routinely outshine ours. Proponents of standardized tests rightly argue that unless we measure student learning, we have no idea of whether efforts to improve education are succeeding or how students here in Tennessee compare with peers elsewhere.

My own colleagues generally agree that testing has a place in education but that much depends on the emphasis you put on it. They distinguish between formative assessments and summative assessments. Formative assessments are the many tools that teachers use to determine whether or not a student is learning while instruction is taking place. Summative assessments gauge how much a student has learned at the conclusion of instruction. Formative data can help teachers tailor instruction to meet student needs, and summative data tells you whether or not the instruction succeeded. Most parents will not object to tests they know teachers are going to use to identify gaps in student learning and to create plans to help students fill in those gaps.

Peabody Professor Kathy Ganske notes that both formative and summative data can be misused. High-stakes tests are just one indicator at one point in time. Formative data also can be misused by teachers who may not have the knowledge to interpret the results for instruction or may not be given the time to do so. “Testing data needs to be interpreted with caution,” Kathy says, “but it can help teachers make informed decisions about instruction specific to children.”

As big data permeates so many other areas of contemporary life, it is safe to say that achievement tests won’t be going away anytime soon. The challenge lies in making sure that these tests actually cover what we all agree it is important for children to learn — that test design is derived from clear instructional goals and not the other way around.

Camilla P. Benbow is Patricia and Rodes Hart Dean of Education and Human Development at Vanderbilt University’s Peabody College. Her column on education appears every other Thursday in The Tennessean Local section.