Exploration of performance indicators and predictors in a blended Moodle-based course within undergraduate EFL studies

Xenia Liahuk, University of Trnava, Slovakia

Abstract: The paper presents the results of a correlation study aimed at exploring factors that can work as indicators of low, moderate, or high level of performance in a blended course designed to educate second year students about the contrastive features of British English and American English within undergraduate EFL studies. The largest group of factors was compiled on the basis of Moodle course statistics, processed using educational data mining techniques. This group includes such indicators as total number of logs, number of quiz attempts, time spend on quizzes and procrastination time. The second group of factors concerned students’ academic background and included the type of secondary school finished and the grade point average (GPA) in all EFL courses taken during the first 1.5 years of bachelor studies, i.e., till the semester in which the target course was offered. The last group of factors comprised competence-related factors, such as EFL proficiency measured through a placement test and plurilingual/pluricultural competence. The research findings suggest that Moodle statistics on students’ performance can give sufficient evidence to identify students that are likely to fail or to perform poorly in the final evaluation, and this evidence can be accumulated after as early as the first four weeks of the semester. On the contrary, students’ EFL proficiency does not seem to correlate with their performance in an academic course that combines theoretical linguistic and cultural premises with their practical application.

Key words: EFL, blended learning, performance indicators, Moodle course statistics, language proficiency.