This study compares the differences between a multiple-choice (MC) and an open-ended (OE) reading comprehension test in a within-subject design to investigate test validity by comparing the two methods. A total of 70 Korean undergraduates participated in the present study. After taking a cloze test, participants read a passage, returned it and took two types of reading comprehension tests. Results showed that there was a statistical difference between the two test methods in a within-subjects design, which agrees with previous between-subjects studies. The higher scores on the MC format were attributed to three reasons: memory constraint, the nature of the task, and test-taking strategy. Further analysis indicated that both proficiency level groups were significantly affected by the test methods. Moreover, both the MC and OE tests significantly separated the two proficiency levels. Information from a survey provides more information regarding differences between the two test methods.

Abstract
I. INTRODUCTION
II. REVIEW OF LITERATURE
III. METHOD
IV. RESULTS
V. DISCUSSION
VI. CONCLUSION
ACKNOWLEDGMENT
REFERENCES