Learner beliefs and corrective feedback in telecollaboration: A longitudinal investigation

While many studies highlight the role of telecollaboration for language learning, there is a lack of research investigating focus on form practices in synchronous, video-based exchange, let alone studies that do so longitudinally. Furthermore, none of the previous studies on telecollaboration examined beliefs as one of the individual differences that may mediate the effectiveness of corrective feedback (CF). This study examined the dynamicity of learner beliefs about error correction and its relationship with actual CF practices regarding provision and uptake of CF in a 14-week, Skype-based eTandem project between American and Japanese universities. The study collected mixed-method, longitudinal evidence from survey (N = 24), interaction (n = 6 pairs), and interview data (n = 5). The analysis of survey data revealed that by the end of the semester, the majority of participants chose recasts as the most preferred way to give and receive CF, since recasts were considered immediate, time-saving, unintrusive, and easy to provide. The interaction data revealed that successful uptake increased when CF was provided in the way learners preferred to receive it. However, no relationship was found between native speakers' preferred method of CF provision and actual practices, suggesting a complex picture of face and identity negotiation as a friend vs. language tutor. (C) 2017 Elsevier Ltd. All rights reserved.