Differential item functioning (DIF) has been studied as a statistical method of finding item bias, and DIF studies have been conducted on tests of verbal abilities such as reading comprehension and vocabulary knowledge. Recently, researchers of language testing sought to understand the causes of DIF. The purpose of this study was to find possible sources of explaining DIF on an EFL vocabulary test, which can then benefit EFL learners and teachers as well. This study thus applied three DIF detection methods (the likelihood ratio test, SIBTEST, and the Mantel-Haenszel test), followed by a regression analysis using item categories and difficulties as predictors of explaining DIF. As a result, only the “Academic” category was statistically significant, positively contributing to DIF in favor of male examinees. Male examinees were better at academic lexical words than female learners of EFL at the same vocabulary knowledge level. This also means that female learners were relatively strong at nonacademic words in comparison to male EFL learners. This finding may lead to more effective curriculum of teaching English. The need for more evidence from replicated studies with better item categorization is also warranted.

Ⅰ. INTRODUCTION
Ⅱ. LlTERA TURE REVIEW
Ⅲ. METHOD
Ⅳ. RESULTS
Ⅴ. DISCUSSION AND CONCLUSION

• Chanho Park(Korea Institute for Curriculum and Evaluation)