Stable EMI research frameworks are needed to ensure that different areas in this multidimensional phenomenon are equally investigated (Dafouz & Smit, 2016). In order to develop such a research framework, we have examined the EMI research (1999-2018) at five European countries that are at different EMI implementation stages.
The fast growth of English medium instruction (EMI) at non-Anglophone universities has led to an increased research interest among applied linguistics scholars, as the escalating literature on the topic demonstrates (Macaro, Seiter, Pun, An, & Dearden, 2017). Researchers have experimented with different theoretical frameworks and research methods and have applied findings for development of local educational and language policies and planning teaching practices. However, more stable EMI research frameworks are needed to ensure that different areas in this multidimensional phenomenon are equally investigated (Dafouz & Smit, 2016). In order to develop such a research framework, we have examined the EMI research (1999-2018) at five European countries at different EMI implementation stages -- from well-established (EMI programs across different universities in the country) to emerging (a few departments at some universities). The resources for the analysis (N=200), articles, books, and reports, were double-coded and entered in an annotated database. Findings suggest an apparent evolution of EMI research foci based on EMI implementation stages due to the pragmatic nature of the research. In the initial stages of EMI implementation, research foci rest primarily on stakeholder attitudes (25% of all sources) and national or institutional language policies (20% of all sources). Therefore, surveys, interviews, and policy analyses (text, discourse) are the most commonly utilized data collection and analysis methods. In the later stages of EMI implementation, the focus shifts to students’ and teachers’ language use through text, discourse, and small corpus analyses (14% of sources). Other EMI areas, such as assessment (2%), intercultural communication (1%), teacher identity (1%), learning outcomes (6%) and training (9%), receive much less attention across the five countries. In order to differentiate between contextual variables and trans-institutional variables related to EMI, we propose a more versatile research framework that includes comprehensive ethnographic analyses of contexts and comparative large-scale research.
