Evaluating item difficulty patterns for assessing student misconceptions in science across physics, chemistry, and biology concepts

Abstract

Understanding item difficulty in science concepts is essential for teachers in teaching and learning to avoid student misconceptions. This study aims to evaluate the patterns of item difficulty estimates in science concepts exploring student misconceptions across physics, biology, and chemistry and to explore differential item functioning (DIF) items in the developed diagnostic test on the basis of gender and grade. Participants were drawn from 856 students (52.3% females and 47.7% males) comprising senior high school students from 11 ^th to 12 ^th grades and pre-service science teachers in the West Kalimantan province, Indonesia. Out of 16 science concepts categorized, the common science concepts causing misconceptions among students were investigated to understand item difficulty patterns using Rasch measurement. The findings of this study evaluated that 32 developed items are valid and reliable whereby the item difficulty estimates ranged from −5.13 logits to 5.06 logits. Chemistry is the scientific discipline with the highest mean logits than other disciplines. There is no significant item difficulty estimate across the science disciplines. We also found DIF issues in one item based on gender and four items based on grade. This study contributes a significant role in mapping and informing item difficulty patterns in science concepts to tackle teachers' problems in assessing and teaching science concepts to improve the students’ science performance. Future studies and limitations are also discussed.

Abstract

Item difficulty; Rasch measurement; Misconception; Science; Diagnostic test.