L2 revision and post-task anticipation during text-based synchronous computer-mediated communication (SCMC) tasks

Charoenchaikorn, Vararin and Kormos, Judit (2019) L2 revision and post-task anticipation during text-based synchronous computer-mediated communication (SCMC) tasks. PhD thesis, Lancaster University.

Text (2019CharoenchaikornPhd)
2019CharoenchaikornPhd.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (2MB)

Abstract

The current research investigates L2 revision during text-based synchronous computer-mediated communication (SCMC) and its relationships with the accuracy of text in chat logs and typing ability. Another main aim of this study is to explore how tasks can be implemented to facilitate learning in this medium. In particular, the effects of post-task anticipation (± post-task anticipation) and its type (anticipation of an individual vs a collaborative language correction post-task) on learners' main task performance in terms of revision, speed fluency and accuracy are examined. The study is primarily motivated by the methodological shortcomings of previous research exploring L2 changes during text-based SCMC, and the scarcity and limited scope of investigation of post-task anticipation studies. Various data collection methods were utilized to gain rich research data. Performance data were obtained from computer screen recordings, keystroke logs and chat logs by means of two text-based SCMC tasks involving picture description and decision-making. Stimulated recall interviews were carried out to gauge participants' thoughts during revision in order to ensure the reliability of the coding of revisions; and an exit questionnaire and a follow-up interview were administered to elicit their responses pertaining to different aspects of the research including their experience of post-task anticipation. This study manipulated both between- (± post-task anticipation) and within-participant (two types of post-task anticipation; anticipation of an individual and a collaborative language correction post-task) factors. Eighty-four Thai learners of English were randomly assigned to either a control (N = 28) or an experimental (N = 56) condition. While the control group carried out two main tasks without any post-task anticipation, the experimental group was informed about a post-task before each of the main tasks. Keystroke logs were examined for linguistic errors and evidence of revisions made during drafting or to the already-sent text. Revisions were coded based on criteria adapted from the revision taxonomies of previous writing research and aided by the data from computer screen recordings and stimulated recall interviews. The variables investigated included quantity, linguistic units, focus and triggers of revision, and rates of error revision success and error corrections. Accuracy was gauged in terms of both accuracy during writing and final text accuracy. Speed fluency was assessed by process-based measures, and typing ability was operationalized as typing speed adjusted for typing accuracy during a typing test. Qualitative data from the exit questionnaire and follow-up interviews were used in conjunction with quantitative data during the analysis. The results showed a high total revision frequency and a high rate of error revision success, suggesting that learners paid close attention to their L2 output and could successfully draw on their L2 knowledge to improve form-related errors in this medium. There was evidence that participants attended more to grammatical features than lexical ones, noticed and corrected more grammatical mistakes compared to lexical ones, and tended to correct grammatical errors more successfully than lexical ones. However, although students attended to grammatical items and revised frequently, the observed dominance of content revisions over form-related revisions indicated that their attention was primarily devoted to the meaning-related aspects of language, rather than to form. This finding does not support previous claims regarding the benefit of text-based SCMC, which argue that this medium is suitable for promoting learners' attention to form. In addition, local revisions occurred very frequently, suggesting that learners' attention might be restricted to short stretches of text at the letter, word or phrase level. Regarding the relationship between revision and final text accuracy, error correction rates were found to be the best predictors of final text accuracy out of all the revision measures. The results of follow-up analyses showed that proficiency potentially influenced final text accuracy and error correction rates; higher proficiency was significantly correlated with increased error correction rates and final text accuracy. Although the correlations between typing ability and most L2 revision measures were not significant, significant relationships were observed between typing ability and 1) error correction rates and 2) accuracy. These findings indicate that learners with better L2 typing ability may have more attentional resources available for attending to L2 output, resulting in increased detection and correction of their linguistic errors and increased internal L2 monitoring. As far as post-task anticipation is concerned, the findings do not support Skehan's (1998) hypothesis about the potential of post-task anticipation for enhancing attention to form and accuracy during the main task performance. No significant effect of post-task anticipation was detected on revision, accuracy or speed fluency. The non-significant effect found on fluency is consistent with the findings of previous post-task anticipation research which did not detect a clear influence of post-task anticipation on this performance aspect. In addition, no significant effect of type of post-task anticipation was observed.