Monday, January 14, 2013

Abuse, Misuse, and Overuse of Standardized Tests

Fifty years ago American students and teachers were subjected to the administration of standardized assessments on a semi-regular basis. A portion of a school day was repurposed for the administration of the test. Most understood the need for standardized assessments; these tests had a particular purpose and provided meaningful information to policy makers and chief educational leaders. Furthermore, the results of these tests provided the means to make comparisons across buildings, districts, and even across states. The test results could be also be filtered by categories such as gender, ethnicity, and special populations. The information was general, but powerful at the bird's eye level only. After all, if a teacher wanted to know how she could improve her practice to better meet the needs of her individual students, she would be looking at evidence of learning at the classroom level.

Something alarming happened, though, over the last 15-plus years that caused a shift in how we use the results of these tests. Over time, the standardized test has been stretched so far and the results given so much weight that the standardized assessment has been abused, misused, and overused. We have lived through a strange transformation of a tool initially intended for policy makers be stretched into allegedly providing valuable instructional and curricular information to teachers. This simply is not the case.

In order to understand why large-scale standardized assessments cannot provide valuable information to teachers, we must remember the original purpose of the standardized assessment. Standardized assessment is administered and scored in the same way. Directions to students are uniform for any student taking the test. Testing conditions are also as standard as possible. This assessment experience allows a considerably large population of students to take the test, and the results of the test are then replicable and generalizable. This means that we can make overall statements about how a large number of students do on this test, and we can repeat this experience next year and be confident that the test is measuring the same things. Despite common practice and understanding, these tests do not have to be timed or in multiple-choice form. The important characteristic is that they are administered and scored in a consistent way.

The reality of our educational system and of our society is that we consistently look for the most "efficient" way. The efficient way to assess our schools usually involves multiple-choice items on multiple forms of a test that can be electronically scored. The results are a jackpot for psychometricians, a North Star for policy makers, and a learning thermometer for Superintendents. The results, however, do not provide any useful information for teachers of individual students in their classrooms. The standardized test was never intended for this purpose.

The claims of this assessment  to be valuable to the individual teacher emerged when we lost sight of the purpose of the assessment. To explain, when developing these and all other assessments, the purpose and appropriate use of the assessment must be the first consideration. Determining a clear and appropriate use of the test is the most important step of assessment development. This determination includes not just the plan for how the test is used, but also the identification of the users of the test. In essence, the purpose of the assessment answers the question of why the assessment is needed.

All who are affected by the assessment, including students and teachers, must understand why an assessment is given.

There are districts across the country that are attempting to evaluate teachers and administrators using tests that were never intended to be, nor can they ever be, useful at the classroom level. Even more distressing is that we are subjecting our students to yearly (at minimum) assessments that were never intended to provide anything but a bird's eye view.

If individual teachers are to be evaluated by how much their students learn, then standardized tests will never fill this role. Instead, we must use assessments that allow students to demonstrate what they know and are able to do. We must use assessments that include self-reflection so that students show ownership in their learning and can communicate the "then" to "now" journey. Most importantly, we must use assessments that provide teachers with rich information used to make instructional decisions that result in increased student learning.

Learning is complex and messy. The most efficient way to measure it must honor its complexity.

Next up? Our love of data.