The estimation of test item difficulty using focus group discussion approach on the semantic differential scale

Abstract

This study aimed to estimate the test item difficulty by comparing the approach to the semantic differential scale and the item and test analysis program. The research used a quantitative-comparative approach, involving 35 samples of 14 teachers and 21 students. Four group pairs were compared. The comparative used the Kruskal-Wallis test then used the Mann Whitney test, with the .01 significance level. The results showed as follows. First, students’ involvement in assessing the test item difficulty is not just providing true or false information. However, the teacher can provide feedback to diagnose student difficulties with the material. Second, student groups are an alternative approach to estimating the difficulty of test items in schools as an option to replace the program. Third, teacher-student groups are an alternative approach to estimating the difficulty of test items in schools as an option to replace the program. The combination of teacher and student was estimated difficulty of the closest item test to the program’s output. It refers to the approach of assessment for learning.