Abstract

This study was conducted using paper-based test (PBT) and computer-based test (CBT) data collected from the 2020 national level academic achievement evaluation for high school sophomores. First, according to the reliability analysis results, stable reliability indices were calculated for the two implementation methods in Korean, math, and English subjects. As a result of examining construct equivalence, it was found that the construct equivalence between the two modes in Korean and English subjects was suspected. In order to control the school-level characteristics, a multi-level model was applied to the schools where both PBT and CBT were implemented, and the regression coefficients of the random intercept model showing greater model-fit were interpreted from the analysis results. As a result, a clear mode effect was found in the type of construct item in Korean and English subjects, and the CBT score was significantly higher. Next, differences in mode effects for each subgroup such as gender, regional size, and school gender were analyzed. As a result, it was found that male students tend to score favorable on CBT in mathematics and English subjects, and female students

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call