Sampson, Andrew and Potts, Diane (2019) Languaging and learning compared across EFL course delivery modes. PhD thesis, Lancaster University.
2019sampsonphd.pdf - Published Version
Available under License Creative Commons Attribution-NoDerivs.
Download (4MB)
Abstract
The present study investigates the effects of delivery mode in adult EFL courses by comparing a) languaging, the “process of making meaning and shaping knowledge and experience through language” (Swain 2006: 98) and b) learning in three settings that are representative of common course delivery modes: i) face-to-face group classes; ii) one-to-one private tutoring sessions; and iii) individual online courses. Conducted within a Vygotskian (1978, 1987) sociocultural framework, data for this quasi-experimental study include recordings of pairwork in group classes and of tutor-learner interaction in one-to-one classes, and also think-alouds of individuals in online courses, thus drawing on Vygotsky’s concept of inner speech and more recent work on self-scaffolding (Holton & Clark 2006; Knouzi, Swain, Lapkin, & Brooks 2009). The analysis is a mixed-methods approach of microgenetic qualitative analysis of learner talk and quantitative analysis of LRE number, focus, resolution and engagement, and includes questionnaire responses and scores from post-tests. Findings indicate that languaging, evidenced in Language-Related Episodes (LREs), occurred in all three modes. While individuals in the online mode produced significantly fewer LREs than learner-learner dyads in group classes or learner-teacher dyads in one-to-one tuition, online individual numbers were similar to LREs initiated by each learner in learner-learner dyads, suggesting individuals identified language problems with a similar frequency as their group counterparts. Learner-learner dyads in group classes and one-to-one learner-teacher dyads produced similar numbers of LREs, but one-to-one episodes were more closely associated with learning than group or individual LREs, as observed in instances of microgenetic development and post-test responses. This may be because one-to-one episodes were better quality in terms of correct resolution and greater resolution by the learner, rather than the teacher. In one-to-one, resolutions followed scaffolding in the form of elicitations and prompts contingent on learners’ tentative responses and teachers’ perceptions of learners’ current knowledge. Such teacher guidance towards learner resolution may have made outcomes more memorable for subsequent post-test recall. Regarding LRE focus, dyads produced more grammar LREs than online individuals, which may relate to habitual grammar-focussed learning practices of face-to-face classrooms, whereas one-to-one dyads focussed more on spelling, suggesting teachers sensed their role was to correct learners’ written language. Regarding LRE resolution, proportions of correctly resolved episodes were similar between group and online individual modes, although the individual proportion was based on fewer LREs, suggesting individual learners did not initiate episodes they would be unable to resolve. The extent to which LREs were characterised by limited engagement (linguistic preferences were stated without further deliberation) or elaborate engagement (there was evidence of a cognitive self-regulation strategy) did not differ significantly between modes. In learner-learner interaction in group mode, the prominence of LREs characterised by limited engagement in one learner and elaborate engagement in the other suggested it was unnecessary for both participants to be elaborately engaged for episodes to be languaged and resolved. Learners across modes averaged post-test scores of 70% to 80% of items resolved in agreement with LRE resolution in the task, suggesting associations between languaging and learning. However, group learners attempted significantly fewer test items relating to their LREs than individual or one-to-one learners, which suggests that forms languaged individually or with a tutor are more memorable. Methodological implications of this research include the limitation of think aloud protocols in making microgenetic development visible to the researcher, while pedagogical recommendations include: i) that learners in all modes be exposed to tasks that focus on form and provide languaging opportunities; ii) that group learners be encouraged to seek gaps in their interlocutor’s knowledge by asking the kinds of questions that teachers ask; iii) that teachers in one-to-one contexts scaffold learners towards resolving LREs; and iv) that online learners seek out an interlocutor – whether a teacher or another learner – the presence of which appears to be positively associated with LRE numbers, correct resolution and microgenetic development. Directions for future research are discussed.