Researching Academic Reading in two contrasting EMI (English as a Medium of Instruction) university contexts, and implications for the design of EAP reading tests

This project examined academic reading in two contrasting English-medium instruction (EMI) university settings in Nepal and Sweden and the unique challenges facing students who study in a language other than their L1. The project was also motivated to consider the suitability of test tasks used in high-stakes tests of English for academic purposes (EAP). We employed a sequential mixed-methods approach in order to gather substantive and authentic data from stakeholders immersed in these EMI settings. A small sample of students (Nepal = 19, Sweden = 9) were asked to complete reading logs over a period of three weeks in order to determine the types of texts and reading load associated with the EMI settings. Additionally, a larger cohort of students from each setting (Nepal = 69, Sweden = 60) completed questionnaires examining academic reading demands, reading skills and practices. Students who completed the questionnaires also completed a research form of the TOEFL iBT® reading section and a post-test questionnaire for participants to reflect on the suitability of the test tasks for their EMI context. Following the test and questionnaires, semi-structured interviews (Nepal = 21, Sweden = 23) focused more closely on students’ perspectives of reading demands in their academic contexts and the suitability of the test tasks for making claims about readiness to enter higher education in EMI contexts. Findings reveal that both EMI contexts are diverse and multilingual. Despite representing a single nationality, Nepali participants display a similar level of multilingualism to the multinational Stockholm cohort. Reading for academic purposes is fundamental to success in both EMI contexts. Students employ a combination of academic reading strategies (Phakiti, 2003) in order to complete assignments and exams. Nepali learners frequently use translanguaging to communicate with peers and staff to make meaning of academic texts. Students in both contexts read a variety of sources in English from authors and publishers both from and outside their academic context. The overwhelming majority of texts encountered in both EMI contexts represented a standard variety of English (McKay and Brown, 2015). Test data revealed that most task types were reliable (α = .93) and discriminated between high and low ability participants in both contexts. Interview data indicated that students in both contexts perceived the tasks to be suitable in terms of register and grammar. However, participants in both contexts questioned the suitability of text topic, preferring subject-specific input. Nepali test data revealed a bi-modal distribution of scores. Interview data revealed that this distribution occurred due to local socioeconomic factors. Students in Nepal exposed to EMI at an early age performed significantly better than those who pursued Nepali-medium schooling. We conclude that to make valid claims about preparedness to study in EMI higher education contexts requires careful consideration of the relationship between test design, content and use. Validation arguments must address the impact of using standard versus local varieties of English, attitudes of stakeholders towards Englishes contained within test tasks (Taylor, 2006) and how context-specific socioeconomic variables impact on test performance