全球智库信息集成服务系统

Six Nationally and Internationally Administered Tests Had Greater Cognitive Demand Than Previously Studied State Tests

• On average, six benchmark tests — Advanced Placement, International Baccalaureate, the National Assessment of Educational Progress, the Programme for International Student Assessment, the Progress in International Reading Literacy Study, and the Trends in International Mathematics and Science Study — were more cognitively demanding than previously studied state achievement tests in both mathematics and English language arts.

The Cognitive Demand of Test Items Varied by Subject and Item Format

• In general, the cognitive demand on the six tests was greater for English language arts than for mathematics, and it was greater for open-ended items than for multiple-choice items.

The Six Tests Varied in the Extent to Which They Assessed Students' Deeper Learning Skills

• Only two of the six tests met both criteria proposed by a panel of education researchers for high-quality measures of deeper learning skills.
• The cognitive demand level of the tests varied with test purpose and the characteristics of the targeted students.

In 2010, the William and Flora Hewlett Foundation's Education Program has established the Deeper Learning Initiative, which focuses on students' development of deeper learning skills (i.e., the mastery of core academic content, critical-thinking, problem-solving, collaboration, communication, and "learn-how-to-learn" skills). Two test consortia are developing the next generation of tests to measure students' attainment of the Common Core State Standards. These tests are expected to assess deeper learning skills to a greater extent than existing large-scale tests. A RAND study rated the cognitive demand of mathematics and English language arts items on six nationally and internationally administered exams: Advanced Placement, International Baccalaureate, the National Assessment of Educational Progress, the Programme for International Student Assessment, the Progress in International Reading Literacy Study, and the Trends in International Mathematics and Science Study, using Norman Webb's Depth of Knowledge framework and the Partnership for Assessment of Readiness for College and Career's self-developed frameworks. It found that these tests were more cognitively demanding than previously studied state achievement tests in both subjects, on average. The test items' level of cognitive demand varied by subject and format. The six tests varied in their percentages of cognitively demanding items, with only two tests meeting both criteria proposed by a panel of education researchers for high-quality measures of deeper learning. Moreover, the tests' cognitive demand levels varied with test purpose and the characteristics of the targeted students. The findings establish a benchmark for comparing how well the new generation of tests performs in assessing deeper learning.

• Chapter One

  Introduction

• Chapter Two

  Tests Included in This Study

• Chapter Three

  Cognitive Demand Frameworks and Ratings for the Benchmark Tests

• Chapter Four

  Findings

• Chapter Five

  Discussion and Implications

• Appendix A

  Distributions of NAEP Items, by Grade and Specific Framework Dimension

• Appendix B

  Distributions of TIMSS Items, by Content and Cognitive Domain

• Appendix C

  Exemplary Test Items at Each DOK Level

• Appendix D

  Exemplary Test Items at Each PARCC Level, by Subject and Dimension

• Appendix E

  Distribution of Modified PARCC Ratings

• Appendix F

  Results for the Original PARCC Dimensions and Levels