Reading comprehension test fairness across gender and mode of learning: insights from IRT-based differential item functioning analysis

Abstract

Abstract Given the significance of the test fairness, this study aimed to investigate a reading comprehension test for evidence of differential item functioning (DIF) based on English as a Foreign Language (EFL) learners’ gender and their mode of learning (conventional vs. distance learning). To this end, 514 EFL learners were asked to take a 30-item multiple-choice reading comprehension test. After establishing the unidimensionality and local independence of the data as prerequisites to DIF analyses, Rasch model-based DIF analysis was conducted across learners’ gender and their mode of learning. The results showed that there were two gender-DIF items which functioned differentially in favor of female respondents. Also, DIF analysis in terms of EFL learners’ mode of learning revealed that there were no DIF items across the two target groups indicating that the reading comprehension test functioned the same way for learners who enjoyed conventional learning compared to those who experienced distance and self-directed learning. In the end, the findings were discussed and implications were provided.

Keywords

• Differential item functioning (DIF)
• Gender
• Mode of learning
• Rasch model
• Reading comprehension test