Development and calibration of an item bank for PE metrics assessments: Standard 1

The purpose of this study was to develop and calibrate an assessment system, or bank, using the latest measurement theories and methods to promote valid and reliable student assessment in physical education. Using an anchor-test equating design, a total of 30 items or assessments were administered to 5,021 (2,568 boys and 2,453 girls) students in Kindergarten and Grades 2 and 5. Their performances were video-taped and rated using 4-point scoring rubrics. The coded data were analyzed using the many-faceted Rasch model, with three defined facets: items, rubrics, and students. Data from Grade 2 were analyzed first, followed by Kindergarten and Grade 5, with the later two anchored on the scale of Grade 2. It was found that the model-data fit well and item difficulties were well spread (Kindergarten: –1.52 to –0.13, Grade 2: –1.22 to 1.29, and Grade 5: –1.26 to 0.44 logits). As a result, an assessment bank was established. An anchor-test equating design allows comparison of students' performance across grades ...