This study aimed at investigating the washback effects of the English module of Iran's National University Entrance Exam (NUEE) on English language education at high schools of privileged and under-privileged areas as perceived by high school teachers and students. To this end, 472 high school students and 260 teachers from Tehran, Qom (privileged), Gharchak, and Varamin (under-privileged) were selected on a convenience sampling technique and were given a washback effects questionnaire to seek and compare their perspectives about NUEE washback effects. Utilizing t-tests on respondents' obtained scores, it was shown that the overall mean score obtained by teachers from privileged areas is significantly higher than that obtained by teachers from the under-privileged areas. In contrast, it was revealed that students from the under-privileged areas obtained a significantly higher mean score than those from the privileged areas. Further frequency counts and detailed content analyses revealed similarities and differences among the participants' perceptions regarding the diverse aspects of the washback effect. Manuscript profile

The current study was an attempt to develop a local model and scale of washback effects of the English language module of the National University Entrance Exam on English language education at Iran's High Schools. To this aim, a mixed methods study was adopted. The qualitative phase involved data-driven from interviews with twenty experts in English education concerning NUEE washback effects which led to the formation of a tentative scale. Then, the quantitative phase included piloting and reliability estimation as well as factor analyses and Structural Equation Modeling for validation of the developed questionnaire. To this purpose, 570 participants were selected through convenience sampling from Tehran, Qom, Varamin, and Gharchack amongst teachers and students as respondents to the questionnaire. Data analyses showed that 58 items were loaded under five dimensions: educational process, attitudes and perceptions, educational policies, emotional and consequential, and socio-cultural dimensions. The model was shown to enjoy acceptable fit indices. Manuscript profile