Exploring the Feasibility of Using NAEP as a Common Benchmark to Measure Differential Course Difficulty

• In Event: Friday Roundtable Session 11:25 am
  In Roundtable Session: Understanding Student Achievement Using the NAEP Data (Table 13)

Abstract

Research has long shown that course grades may be subject to systematic variability due to factors unrelated to student ability, such as course difficulty, teacher discretion, school policy differences, and grade inflation. This study uses an "overlap sample" of students from the HSLS:09 who were selected to take the grade 12 NAEP mathematics. The study investigates the feasibility of using NAEP as a common benchmark in a graded-response model (GRM) to measure differential course difficulty in five high school mathematics courses. Systematically lower or higher thresholds for some student subgroups indicate they are likely to receive higher or lower grades than others with the same NAEP-measured proficiency.

Author

• Darrick S. Yee, American Institutes for Research