More Information on PARCC and Smarter Balanced Assessments 

In this interview with Nancy Walser in Harvard Education Letter, assessment expert Joan Herman (currently with CRESST), shares important information on the upcoming Common Core tests:

• A more-transparent process – Both consortia are using evidence-centered design, starting with 4-5 claims about student learning in each subject (for example, that students can read and understand increasingly complex texts), defining assessment targets, and creating item specifications. 

• Difficulty – The new tests “will definitely be harder, but they should be more interesting,” says Herman. “There should be a lot of use of authentic texts and authentic, real-world problems.” Students will be asked to synthesize information, reason mathematically, conduct research, integrate multiple sources, write coherent explanations, communicate effectively, and make reasoned arguments. Educators need to prepare the public for a drop in scores, arguing that the tests measure what is needed to be ready for college, work, and life and the results can help students meet those goals.

• Comparisons – Herman says the two tests will be quite similar. Both will be taken on computers (with multimedia features and accommodations) and scored almost immediately. Students will answer multiple-choice, short-answer, and other types of questions. Both end-of-year tests will include a performance task, and so will the PARCC mid-year tests. Smarter Balanced tests will be adaptive (adjusting questions based on students’ answers) while PARCC tests will all have common questions. 

• On schedule – Herman believes the tests will be ready for implementation in the spring of 2014. They won’t be perfect and there will undoubtedly be glitches, she says, but “Let’s not let the perfect be the enemy of the good.”

• Challenges – Crafting performance tasks is difficult, says Herman. Students’ prior knowledge varies, and year-to-year comparability is tough. 

• Misuse – “I worry about results being used in value-added models for teacher evaluation,” she says. Results can give useful information on the achievement of schools and individual students, but she doesn’t think they should be used to make fine-grained distinctions between schools, teachers, and individual students. “Any score is an estimate.” 

“Assessing the New Common Core Tests: An Interview with Joan L. Herman” by Nancy Walser in Harvard Education Letter, July/August 2013 (Vol. 29, #4, p. 8, 6-7), www.edletter.org. 

From the Marshall Memo #492