A set of test questions is calibrated on a sample of high-pe…
A set of test questions is calibrated on a sample of high-performing students and re-estimated on a sample of low-performing students. Under Item Response Theory, how should the item difficulty parameters (δj) compare across samples?