Teaching and learning languages remotely: Pandemic lessons for a post-pandemic future

Building on MacIntyre et al. (2020) and Mercer & Gregersen (2020), this exploratory study examines the transition to emergency remote instruction (ERT) during the COVID-19 pandemic based on questionnaire responses from over 6,000 language learners and teachers from 118 countries.

To examine how the stakeholders adapted to ERT, we had constructed an original online survey composed of 441 items. Constructs of interest were derived from existing validated scales as well as ones developed specifically for this project. Exploratory and confirmatory factor analyses with direct oblimin, excluding items exhibiting cross- or low loadings (<.30), led to the development of 30+ new scales exhibiting acceptable to very good reliability indices (Cronbach's α:.72–.92, McDonald's ωₕ:.75–.92) and adequate fit as indicated by TLI and RMSEA values.

Responding to specific research questions, regression and other inferential analyses of instructors' responses revealed that:

1. teachers coped better when they worked in higher education: F=9.31,p<.001,η_p²=.02[.01;.04] and used synchronous delivery: t=−6.2,p<.001,d=.33[.22;.43];
2. educators were more engaged in developing than economically developed countries: t=3.59,p<.05,d=.31[.14;.48];
3. psychological overload was mediated by perception of student coping;
4. instructors' stress levels were affected by anxiety about the future, living conditions, self-acceptance, appraisal of situational impact, course optionality, and perceived effectiveness of virtual delivery;
5. on average, teachers felt that remote instruction depressed students' language progress by around 64% compared with in-person classes;
6. future learning outcomes were the biggest cause for concern in A1-level classes:β=.09,R²=.51,p=.026;
7. the breakup of some constructs in clusters of naturally correlating variables suggests that in crisis situations these may function differently than during 'business as usual', supporting the Strong Situation Hypothesis (Meyer et al., 2010) and in line with Resnik and Dewaele (2021);
8. participants' coping behaviour and attitudes were moderated by multilingualism operationalized as weighted proficiency in languages spoken. Although 'more polyglot' teachers found remote teaching harder than initially expected (r=.223), they demonstrated more stability in their lives (−.278) and instruction (−.373) and were more likely to believe that they would come out unscathed (.252). They felt their students were coping well (.302), and their classes were longer (.271).

Subsequently, we shift attention to factors distinguishing better- and worse-coping language learners. We then compare the teacher and student subsamples side-by-side to see whether Extraversion similarly or differentially affected Anxiety in the two cohorts. A two-step cluster analysis, combining the hierarchical and k-means clustering algorithms, revealed that Extraversion influenced Anxiety only in the instructor cohort (t=6.02, p<.001, d=.28[.16,.42]).