Testing: A Small Window Measuring the Size and Depth of the Room

By Dr. David Weiss, Co-founder and Chairman of Assessment Systems Corp

For many of us in and around the academic orbit, test scores have been the ne plus ultra, the sine qua non of academic proof points. If you’re not up on your Latin, for many, test scores have been the unimpeachable evidence of successes and failures in teaching and learning.

In truth, using test scores as the sole, best determinant of education successes and student achievement was probably never the best idea. For one thing, most test scores have been, at best, a drive-by peek through a small widow to measure the size and depth of a room. For another, test scores are usually only as good as the test. Further, test scores have been used primarily as trailing indicators of learning, not leading indicators that can help shape instruction.

Fortunately, that reliance on test scores as the best metric for knowledge acquisition and skill attainment is changing. And that’s going to change the way we think about the tools and techniques we offer to help teachers fulfill their missions.

From Standardized Testing to Adaptive Assessments

First, testing is changing. Even the word is. The term is “assessment” now. Soon, the very idea of a standardized test – a test in which every student gets the same questions and is measured against some fixed point – will go the way of the encyclopedia. Adaptive assessments, measurement tools in which the next question is determined by how someone answered previous questions, are more accurate and more efficient. And, best of all, adaptive, computer-driven assessments allow us to harness AI and analytics in ways that make them not just accurate, but hyper-accurate.

Assessment designers and psychometricians are also working to spread the assessment process out over an entire academic tenure instead of loading it into pressurized testing snapshots. That approach will not only be far more accurate and insightful, it will allow us to know, for example, which students are struggling and what they’re struggling with and even why. Better assessments can inform future learning, exploring learning gaps and challenging exceptional achievers.

New Methods of Analyzing Assessments

Those changes in assessments – adaptive approaches, ongoing assessment across exponentially more data points and the new, powerful tools of AI – will change how we think about and frame education products and ideas because they will allow for the unprecedented testing of proposed solutions, not just students.

Soon, we will be able to do more than just speculate that a particular app or specific reading assignment enhances reading comprehension. We will know concretely that it increases vocabulary reasoning by 18% among eighth graders who have been reading above grade level for at least two years, as an example. And we’ll be able to cue teachers and other education leaders about the solutions and tactics that work best for each specific student or issue.

Perhaps even more importantly, testing education solutions will allow us to start culling those with mediocre or missing performance gains. To me, probably to most of us, that’s unquestionably good news. Better assessments can give us necessary clarity about what’s actually working in classrooms so we can focus on fixing what isn’t and scale what does.

Using Data to Enhance Teaching and Learning

And while all that may seem like a heavy burden to place on the future of assessment, it’s not any more than we’ve been putting on the back of “test scores.” Moreover, that burden isn’t on the future of assessment. Assessment technology available and in use today can already do all those things and more. The only open questions are when we swap data-driven and adaptive assessment for blunt test scores and how quickly we can shift our focus from where schools and students place on general metrics to using metrics to enhance teaching and learning.

Dr. Weiss is the featured keynote speaker at the International Association for Computerized Adaptive Testing Conference held on June 10-13 at the University of Minnesota. The conference will foster lively discussions and presentations from researchers, leading academics, test developers and school leaders about the best ways to assess learning. See the program here or visit IACAT.org.

Dr. David Weiss is widely considered the grandfather of adaptive testing. He is the cofounder and chairman of Assessment Systems Corp. and has been doing measurement research at the University of Minnesota for more than 50 years. His goal is to eradicate the low-fidelity anachronism that is paper and pencil testing. Dr. Weiss is a founding member and President Emeritus of the International Association for Computerized Adaptive testing, and has published hundreds of articles, book chapters, technical reports, and conference presentations.