Negative washback is important in test validation, and if it is the result of the test or testing procedure, it can invalidate the test even if the interpretations are validated (Kane, 2013; Lee, 2015). This study used an argument-based approach to validation to compare the washback effects of Vocabulary Levels Test (VLT), the monolingual version of Vocabulary Size Test (VST), and the bilingual versions of the VST, which are three widely-used measures of written receptive vocabulary size in L2 vocabulary research. For this purpose, a five-point Likert-scale questionnaire containing 20 items was developed for each test to find the perceptions of the EFL teachers towards their effects on the teaching and learning processes as well as the ELT materials. Moreover, an open-ended question was included to further explore the perceptions of the participants. After checking the content validity of the questionnaire and checking its reliability through piloting it using a group of 62 EFL instructors and MA graduates, it was administered to 30 EFL instructors who were Ph.D. students or holders and had at least five years of teaching experience at both language institutes and universities. Results of one-sample t-test and thematic analysis indicated that the VLT and VST were perceived as vocabulary size tests which have positive washback while the bilingual Persian VST was identified as a test which could have negative washback for vocabulary teaching and learning processes as well as the corresponding ELT materials. The findings have implications for teachers, researchers, and test developers.
