Impact of Different Performance Assessment Cut Scores on Student Promotion

This cut score modeling study investigated whether changing the cut scores of language arts performance assessments (PAs) would differentially affect subgroups of students. Based on a data set of 227,817 students, hierarchical logistic regression analysis revealed that no sub?groups of students were found to have a higher or lower possibility of passing the PAs based on their background variables when we switched between the two different cut scores, with one exception. Immigrant students were found to have a higher possibility of passing than the native students when the cut score was set at 3, and we found no differences between them with the cut score at 2. This result contributes to evidence on the relative fairness of the Pas.