AILA Review - Volume 38, Issue 1, 2025

English-medium instruction (EMI) is rising in importance in higher education, especially for fields such as medicine. However, most students have some degree of linguistic barriers when studying courses in EMI. This study explored the challenges medical students face in studying medicine taught in English; whether these challenges differ by gender, academic discipline or academic level; and students’ perceptions of the relationship between English proficiency and academic performance in EMI. It drew upon cognitive load theory to explain how these linguistic difficulties influence students’ ability to manage the cognitive load of EMI in medical studies. The study included 264 Saudi EFL medical students who had nine years of prior English experience. The English-Medium Instruction Challenges Scale, adapted from Aizawa et al. (2020), measured perceived challenges in writing, speaking, reading, and listening. The participants’ International English Language Testing System (IELTS) scores represented their English language proficiency, their specialized-English competence was gauged through English for Specific Purposes (ESP) scores, and their grade point averages (GPA) represented academic success. A multivariate analysis of variance (MANOVA) showed significant differences in challenges according to discipline and academic year, and an analysis of variance (ANOVA) showed differences that were not significant with regard to gender. It was found in the regression analysis that students with lower English proficiency (IELTS scores) reported more language challenges in EMI contexts. However, ESP scores did not significantly affect perceived challenges, becoming insignificant once IELTS scores were obtained. This study emphasises the need for language training to be incorporated into medical EMI settings and calls for policy reforms to include EMP to empower future health care professionals with the needed language skills that clinical settings require.

The growth of and attention given to English-medium instruction (EMI) and related topics in contemporary higher education research was well-represented at the AILA 2024 World Congress in Kuala Lumpur, Malaysia. Research reports and symposia from around the globe attest to the fact that EMI remains a consistent area of interest and that many questions and research agendas are being engaged in a number of ways and at a range of levels. This paper takes the opportunity offered by one of the AILA 2024 workshops, namely “Comparative guidelines for cross-contextual EMI research: Policies, people, and practices” (Siegel, Kumazawa & Zuaro, 2024), to reflect on some of the key interrogatives that still accompany EMI as an educational approach, regardless of its context of implementation. Set against the backdrop of multiple EMI sessions at the conference, the workshop is foregrounded as an example of successful cross-contextual EMI description, shedding light on the advantages of adopting similar analytical categories in — sometimes very — different contexts. This is done to showcase how structured comparison can facilitate the identification of areas of divergence. In the present paper, then, we argue in favor of clarity and efficiency in EMI contextual description and suggest avenues for future research that could benefit from cross-contextual studies. We do this by recounting our collaborative experiences and articulating when and in what ways aspects of the road-mapping framework (Dafouz & Smit, 2020), deployed in the workshop itself, emerged naturally since the beginning of our joint work.

In investigating second language (L2) listening, verbal report (VR) can provide direct insight into listeners’ behaviors. However, a concern over this method is its potential for reactivity — that the VR may change (either positively or negatively) the individual’s cognitive processes involved in the task being investigated. Thus far, few studies have examined for reactivity in L2 listener VRs, and none have investigated it from a quantitative perspective, a gap which is addressed in this study. The study involved lower-intermediate to intermediate-level Mandarin-speaking English learners, and it compared the listening comprehension of an experimental group undertaking VRs with that of control groups assessed through recall protocols. No significant impact on comprehension was found for the VRs indicating, by extension, no interference with the listeners’ thought processes, and thus no apparent reactive effect. The study additionally examined for reactivity: (1) among three different VR styles (styles 1 and 2 were mediated styles, with each of these styles mediated in differing ways by researcher prompts, and style 3 was an unmediated VR); and (2) for mediated VRs, in general (the combination of styles 1 and 2). None of the analyses found a significant effect on comprehension. One further analysis, which lacked a control group, though, concluded there was a slight comprehension advantage for one of the mediated VR styles over the other two styles. The study had its limitations, but the findings suggest L2 listener VRs do not generally appear to be reactive, suggesting they are suitable for use by researchers investigating L2 listeners.

Most reading technologies claim to provide experiences similar to reading on print paper. This study compared reading across different digital platforms and print books. Digital reading mediums used in this study were reading on a PC screen, a handheld e-reader and an iPad. A total of eighty participants enrolled in various university courses took part in this experiment. Each participant was randomly assigned to one of the reading conditions, and subsequent eye-tracking measurements were compared. Results indicated that the mean fixation duration was statistically different only for the reading from the PC condition, whereas it was similar across all other conditions. The reading time and the total number of fixations were comparable for the dedicated e-reader, iPad and printed book. This study provides empirical evidence that e-reading tools closely mimic book-reading conditions.