Two types of translation tests: their reliability and validity

Abstract This study examines the reliability and validity of translation tests as a reading comprehension measure. The following tests were administered to 70 Japanese high school students: (1) open-ended translation (OE-TRAN) test from English to Japanese; (2) multiple-choice translation (MC-TRAN) test; (3) cloze test (CLOZE-T); and (4) short-answer question test (SHOAQ-T). The results indicate the following: (1) OE-TRAN test might be a reliable and valid method for measuring reading comprehension; (2) MC-TRAN test also has a potential for being a reliable and somewhat valid method for measuring reading comprehension; (3) There was a moderate correlation between the two translation tests; (4) The participants performed similarly across the two translation tests; and (5) The examinees' proficiency level affected the level of reliability and validity of the two translation tests as far as the values of correlation coefficients are concerned, though it should be pointed out that we do not have mean-difference evidence on this matter.