English medium instruction lecturer within-course linguistic evolution: monitoring changes between STEM lectures

Abstract Research in English Medium Instruction (EMI) has often focused on broad policy levels or disciplinary content rather than teachers’ linguistic behavior in EMI classrooms. In order to explore linguistic behaviors in one Science Technology Engineering and Mathematics (STEM) subject, this study investigated changes in linguistic behavior over a series of three lectures in an EMI robotics course. To monitor speech and articulation rates, five evenly-distributed one-minute excerpts from each of three lectures were analyzed and compared using ELAN software. AntWordProfiler, in combination with various word lists (e.g., the Academic Word List and the Academic Spoken Word List), was used to study vocabulary choice and keyword aspects throughout the entire lectures. Findings showed increases in both speech and articulation rate from the first to the third lecture with an accompanying reduction in total pause time. Vocabulary coverage remained largely consistent, with the most common 2000 words being used around 90 % of the time. Regarding keyword appearances, frequency decreased over time, with noticeably fewer uses evident in the third compared to the first lecture. Based on the findings, implications for EMI teachers, teacher trainers, and students are outlined with respect to teachers’ rate of speech and vocabulary use.