2024 Theses Doctoral

Interactional Feedback in Text-based and Voice-based Synchronous Computer-Mediated Communication

Akbar, Farah Sultana

In the ever-changing digital era, the surge of online informal language learning platforms has revolutionized the way we learn new languages. This dissertation delves into the critical role of interactional feedback (IF) in synchronous computer-mediated communication (SCMC), a cornerstone for optimizing these digital learning spaces. Anchored in the interactionist perspective and usage-based linguistics, this research seeks to illuminate the pathways through which IF can enhance language acquisition in virtual settings.

The proliferation of online language learning tools has not been matched by a comprehensive understanding of feedback dynamics. This study, therefore, sets out to bridge this knowledge gap by examining the impact of IF on language learners' proficiency, particularly in article usage - a notorious challenge for language learners. Over a span of ten weeks, ten pairs of native speakers (NS) and non-native speakers (NNS) engaged in SCMC sessions, generating a rich dataset of 49 hours and 197 transcripts. This corpus served as the foundation for a mixed-method analysis, scrutinizing the types of NS feedback, the NNS errors that prompted such feedback, and the subsequent NNS responses, with a focus on article usage across varying proficiency levels and discussion topics.

The investigation uncovered a spectrum of eleven distinct feedback types, with embedded corrections and interrogative recasts emerging as the most potent in eliciting learner uptake. The comparative analysis of text-based and voice-based feedback revealed a notable disparity in effectiveness, with text-based feedback succeeding in 33% of instances, while voice-based feedback achieved a higher success rate of 43%. The study also highlighted article errors as the predominant trigger for NS corrective feedback, with a remarkable 69% of NNSs adjusting their article usage in response to such feedback.

A qualitative dive into the data exposed personal subjects as the most frequent topics of conversation, underscoring the significance of contextually relevant interactions in language learning. Employing a dynamic coding approach, complemented by rigorous statistical analyses such as the Friedman test and subsequent post hoc comparisons, the study meticulously charted the evolution of article usage over the course of the ten-week period. The post hoc comparisons, in particular, shed light on the intricate relationship between learners' proficiency levels and their mastery of article usage. These analyses confirmed a positive correlation between proficiency and article accuracy, with higher proficiency learners consistently demonstrating more precise article usage in both text and voice-based SCMC.

The implications of these findings are far-reaching for the field of language education. They underscore the necessity for educators to cultivate a nuanced understanding of the diverse feedback mechanisms at play in different communicative contexts. The study advocates for the strategic use of varied feedback techniques, such as integrated corrections and clarification requests, and the design of activities that foster target-like language production. In crafting SCMC tasks, educators are encouraged to prioritize feedback that bridges communicative gaps, ensuring that such activities are in harmony with pedagogical objectives and learner preferences.

In conclusion, this research contributes to the body of knowledge on digital language learning by providing empirical evidence on the efficacy of IF in SCMC. It offers actionable insights for language educators seeking to refine online instructional practices and curriculum development, with the ultimate goal of enhancing learners' linguistic competence in an increasingly digitalized world.