Gender fairness in measuring student agency: A Rasch DIF analysis

Abstrak

As student agency increasingly shapes educational policy and reform, ensuring that its assessment practices are fair across genders has become a pressing equity concern. This study examines the gender-based measurement fairness of a multidimensional student agency instrument using Rasch-based differential item functioning (DIF) analysis in an Islamic secondary school. The study involved 601 students in Grades 10–12 (301 male; 300 female) selected through a school-based total accessible sampling approach, comprising 96 male and 101 female students in Grade 10, 138 male and 124 female students in Grade 11, and 67 male and 75 female students in Grade 12. Data were collected using a 60-item student agency questionnaire measured on a four-point Likert scale covering eight regulatory, motivational, and future-oriented dimensions. Results show that the instrument operates largely equivalently for male and female students. Although several items exhibited statistically significant DIF, substantively meaningful DIF was limited, localised, and bidirectional, and did not accumulate at the dimensional level. Core regulatory dimensions demonstrated strong invariance across gender. These findings indicate that observed gender differences in agency scores are unlikely to reflect measurement bias but rather reflect contextual variations in the expression of agency within this educational setting.