Letter from Stuart Kahl

The following letter was sent to the editor of Education Week. Education Week published an edited version of the letter in the June 22, 2005 issue. 

To the Editor: 

“States Eyeing Expense of Hand-Scored Tests in Light of NCLB Rules” (May 25, 2005) highlights a significant risk that No Child Left Behind could pose to the value and efficacy of states’ large-scale assessments. For the U.S. Department of Education to suggest that states drop open-ended test items as a cost-cutting measure is tantamount to saying that assessment data is only meant to classify schools and teachers, rather than to actually improve teaching and learning. 

Open-response and multiple-choice items don’t measure all the same skills. In reading, do you want students to draw conclusions or pick from a list of someone else’s conclusions? In math, do you want students to arrive at an answer using approaches other than the ones you intend to test? One research-supported lesson of the “authentic assessment” movement of the 1990s is that sole reliance on multiple-choice items in high-stakes testing can have a negative impact on curriculum and instruction. 

Some of the richest assessments states have developed contained only open-response items. These tests were equally as reliable as multiple-choice tests; the nature of measurement error is just different for the two formats. Given the current environment, most states would find that approach impractical. For that reason we support balanced assessments that include both multiple-choice and open-response items. History shows that such assessments provide the quality of information educators need to improve curriculum, instruction, and most important, student learning. While all-multiple-choice assessments might satisfy “the letter of the law,” I certainly hope that the “spirit of the law”— and the true goal of assessment—is to help improve teaching and learning. 

Best regards, 

Stuart R. Kahl, Ph.D.

President and Chief Executive Officer