An Experimental study of the multiple-choice test technique for measuring reading comprehension of EFL students

Abstract

I ABSTRACT Since reading is an integrative and difficult skill to measure no single test adequately measures reading ability. Testing reading comprehension is problematic, especially in a foreign or second language setting, because test results may be invalid; a reading comprehension test score may be measuring something other than the reading comprehension ability of the student. The purpose of this study is to test two reading comprehension testing techniques, multiple-choice and retelling by conducting an experiment using intermediate level EFL students from Bilkent University School of English Language (BUSED as subjects. The goal of the experiment is to correlate these two testing techniques with two other measures of reading comprehension ability, '.^teacher ratings and the reading section of a BUSEL achievement exam. The correlations of these four measures were done using the Pearson Product Moment Correlation and the Tetrachoric Correlation. The results showed that the multiple-choice reading test did not correlate significantly with the other three measures, implying that the multiple-choice test did not successfully predict the reading abilities of students. On the other hand, the correlation of the retelling procedure with the other measures showed significance,suggesting that the retelling is a more valid way to test reading comprehension. Because the total number o-F subjects in this study was small, only 18, we should hesitate be-Fore generalizing and applying these results to other EFL situations. However, the -Findings of the experiment should be o-F interest to all those involved in testing the reading comprehension ability of foreign or second language students.