Item analysis of a multiple-choice test of incoming EFL students at a public university in Guanajuato State

Résumé

The importance of multiple-choice (MC) exams in Mexico can hardly be overstated. Despite their ubiquity in Mexican EFL classrooms, little research on their quality or impact has been carried out. The current paper reports on an item analysis conducted on a multiple-choice placement exam administered to 50 incoming English language students at a public university in Guanajuato State. All the items contained in this test were subjected to four different types of statistical analyses: item facility (IF), distractor analysis (DA), item discrimination (ID), and response frequency distribution (RF). The results showed that the exam was critically deficient in each of these areas. While the findings of the current analysis cannot be generalized beyond the exam in question, the results highlight the extreme difficulty of creating MC exams and the indispensability of piloting tests and conducting item analyses before applying them.