Session 4E

In foreign language (FL) contexts, exposure to the target language is extremely limited (Muñoz, 2008) and additional sources of input are often very much needed to assist the learning process. Subtitled television can be used as a resource to enhance FL learning (Vanderplank, 2019); so far, research exploring its potential has mostly focused on adult learners viewing short clips (e.g., Hsieh, 2019) and on the usefulness of different subtitling modes (e.g., Peters, 2019). Yet little is known about the possible benefits of extended exposure to subtitled FL television in younger learners, particularly in addition to formal curricular instruction. This longitudinal study investigates whether a pedagogical intervention on TV viewing facilitated the learning of a set of target words (TWs) to Catalan / Spanish beginner (n=40) and intermediate (n=57) EFL learners in primary and high school. At each level, participants were allocated to an experimental (EG) or control group (CG). Weekly, and for a whole academic year, all the groups were pre-taught the corresponding TWs and performed several vocabulary learning tasks. However, only the EGs were additionally exposed to subtitled episodes of a TV series where the TWs appeared, while the CGs carried out other classroom activities unrelated to the target vocabulary. To measure lexical growth, all students took a pre- and a post-test evaluating both TW form and meaning recall at the beginning and end of each of the three terms the academic year was divided into. To measure vocabulary retention, a delayed test was administered eight months after the end of the intervention. Results revealed that all participants (EGs and CGs) learned new vocabulary, but intermediate learners benefitted more from the experiment than beginners. Furthermore, significant differences between conditions were more remarkable at the end of the academic year, suggesting a positive effect of sustained exposure to subtitled TV series at both proficiency levels. However, similar vocabulary retention rates were observed irrespective of the experimental condition. Conclusions will be drawn in relation to multimodal input theories (Mayer, 2009; Paivio, 1986) and the use of audiovisual materials in the language classroom (Webb, 2015). References Hsieh, Y. (2019). Effects of video captioning on EFL vocabulary learning and listening comprehension. Computer Assisted Language Learning. Mayer, R. E. (2009). Multimedia learning (second edition). Cambridge: Cambridge University Press. Muñoz, C. (2008). Symmetries and asymmetries of age effects in naturalistic and instructed L2 learning. Applied Linguistics, 29(4), 578-596. Paivio, A. (1986). Mental representations: A dual coding approach. New York: Oxford University Press. Peters, E. (2019). The effect of imagery and on-screen text on foreign language vocabulary learning from audiovisual input. TESOL Quarterly, 53(4), 1008-1032. Vanderplank, R. (2019). ‘Gist watching can only take you so far’: attitudes, strategies and changes in behaviour in watching films with captions. The Language Learning Journal, 47(4), 407-423. Webb, S. (2015). Extensive viewing: Language learning through watching television. In D. Nunan, & J. C. Richards (Eds.), Language learning beyond the classroom (pp. 159-168). New York: Routledge.

According to the cognitive-interactionist paradigm (Doughty, 2001; Doughty & Long, 2003), corrective feedback (CF) in second language (L2) learning is most effective when provided immediately after learners’ non-target-like productions. This is because the short time lapse between feedback and error may help learners make the mental comparison between their intention, their output, and the input conveyed through feedback (Doughty, 2001). However, in online, or computer-mediated, language learning contexts, teacher-learner communication cannot always take place synchronously. As a result, corrective feedback is typically delayed and provided, for example, at the end of a task, at the end of a lesson, or several days after a lesson. If the contingency between error and feedback contributes to the effectiveness of feedback, this means that recreating immediacy conditions as part of delayed feedback could make delayed feedback as effective as immediate feedback. This expectation was the main motivation behind the present study, together with the need to investigate feedback timing in voice-based, rather than text-based, interaction, which has been the main modality investigated in computer-mediated communication studies in the field of second language acquisition (SLA) (e.g., Shintani & Aubrey, 2016). The study we report examined the relative effects of feedback timing on the acquisition of two English structures (third-person possessive determiners and –ing/-ed adjectives). Fifty Spanish learners of English as a foreign language were randomly assigned to one of three groups (immediate, delayed immediate, control) and carried out two communicative tasks with a native speaker via a video-conferencing tool. The immediate group received immediate explicit corrections involving the overt rejection of the erroneous segment of their utterance followed by a target-like reformulation. The delayed immediate group received explicit corrections 24 hours after the interaction by means of a video replay with inserted oral corrections by the same native speaker. Insertions kept the “spaced” feature of immediate feedback constant. The additional exposure to positive evidence in the delayed immediate group was controlled in the immediate group by a replay of the interaction with feedback removed. By providing feedback in the same modality that the communicative task was carried out (i.e., oral) and in context (i.e., not in isolation, but as insertions in the interaction), we aimed at maximizing the chances of the learner linking form and meaning, and, thus, at making delayed feedback more effective. Oral production and untimed grammaticality judgments were used to measure learners’ pretest-posttest development. Results showed that feedback had a positive effect on learners’ development of English –ing/-ed adjectives and that there were no significant differences between the two experimental groups on either outcome measure, even though the magnitude of the effect according to Cohen’s d was large in the case of the immediate group and medium in the case of the delayed immediate group. These findings can inform pedagogical contexts where the provision of immediate feedback is not possible or contexts where an instructor might prefer not to provide immediate feedback in order not to disrupt the communicative flow. Theoretically, this type of research can help us identify the crucial element that makes immediate feedback more effective than feedback under other timing conditions.

Collaborative writing as a means to foster language learning has received a lot of interest in the past years (Storch, 2016). Previous research has shown that collaborative writing is not always beneficial, due to the fact that language learning potential and writing development depend largely on how interaction between peers enfolds (Storch, 2005). Similarly, earlier studies demonstrated that individual differences (e.g., proficiency) affect interaction, indicating that individuals engage differently with a task and each other when paired with different partners (Kim & McDonough, 2008). Most collaborative writing nowadays is facilitated by web 2.0 technologies, such as shared online spaces (e.g. Google Docs). Earlier work comparing face-to-face and digitally mediated collaborative writing demonstrated that the mode of delivery influences the quality and quantity of interaction between learners (Rouhshad & Storch, 2016). Little is known, however, about how working with different peers and tasks affects interaction in digitally mediated contexts. The present study investigates the variability of interaction shown by students when working with different partners on different tasks during online collaborative writing. We use a mixed-methods design to explore how affective, social and cognitive dimensions of learners’ engagement (Svalberg, 2009) mediate learners’ dynamics of interaction. Eight international learners of English (CEFR level B2/C1) at a tertiary institution in the North of England completed four paired collaborative writing tasks in Google Docs and communicated via written Google Chat. During three sessions, pairs had 30 minutes each to complete different integrated writing tasks that required them to understand, describe and discuss data presented in a scientific graph. Session 1 took place in-class during a pre-sessional English for Academic Purposes (EAP) course and students worked with a randomly allocated partner. For session 2 and 3, pairs of students were invited to the lab where they completed the tasks sitting in separate rooms, working with a different partner for each task. Partners were allocated based on their initial collaborative pattern (Storch, 2002) exhibited during session 1 as retrieved by the visual history charts created by the text mining tool DocuViz (Wang, 2016). During lab-sessions 2 and 3 each participants’ eye-movements were recorded using a Tobii eye-tracking system. After session 3, individual participants reviewed a video replay of their performance, and were asked to comment on their thoughts during their last task performance, following stimulated recall (SR) procedure (Gass & Mackey, 2017). To investigate the dynamics of interaction three different data sources were analysed. First, learners’ contribution to composing and editing, as well as the chat interaction was analysed. This analysis aimed to determine the equality of the interaction among partners and to explore affective and social dimensions of engagement. Second, SR interviews were coded to tap the learners’ attitudes towards the task and their interlocutor as additional indicators of affective engagement. Third, eye-tracking data indicated the level of cognitive engagement with both the task and the partner’s input. Together these data sources reveal how engagement is mediated by different task partners at the affective, social and cognitive level. This becomes apparent in the writing and editing process as well as in the chat conversations. We will discuss how interaction patterns evolve dynamically during collaborative writing tasks as a result of the interrelation between learners and tasks (Storch, 2017; Zhang, 2019). We provide implications of these findings on the design and implementation of collaborative writing tasks in online contexts.