Effects of interactivity of written practice on incidental vocabulary acquisition

Autor: Bibauw, Serge, François, Thomas, Desmet, Piet
Přispěvatelé: UCL - SSH/ILC/PLIN - Pôle de recherche en linguistique, UCL - SSH/TALN - Centre de traitement automatique du langage
Jazyk: angličtina
Rok vydání: 2019
Popis: Dialogue-based computer-assisted language learning (CALL) allows a learner to practice an L2 by discussing with a computer. By offering a fully controllable meaning-focused learning environment, it also allows to conduct research on the effectiveness of different modalities of interaction. In this study, we compare the effects of the interactivity of the writing activity on the incidental vocabulary acquisition. Interactive learning activities are often assumed to generate more learning outcomes than non-interactive ones (Chi, 2009). We know that negotiation of meaning and interactionally modified output are beneficial for incidental vocabulary learning (Newton, 1995; Ellis & He, 1999). Is it possible however that, at identical input and output opportunities, the level of interactivity offered by a dialogue activity affects incidental vocabulary learning? In a randomized controlled experiment, 160 learners used two versions of a dialogue-based CALL game. Their receptive (form-meaning mapping) and productive (collocational use) knowledge of specific French lexical items was tested before and after these sessions. In the treatment, participants had to maintain several task-based conversations with in-game characters. Both conditions presented the same input and had the same output opportunities, but the difference lied in the interactivity of the conversational practice: in the “dialogue system” group, the system-controlled interlocutor answered dynamically to each participant, while in the “dialogue completion” group, all answers from the interlocutor were visible at the start of the activity. Preliminary results show that the “dialogue system” group (Mgain = +10.9%, d = .55) significantly outperforms the “dialogue completion” group (Mgain = +8.8%, d = .35). These findings are consistent with previous studies (e.g., Newton, 1995; Ellis & He, 1999) and with the Task Involvement Hypothesis (Laufer & Hulstijn, 2001). They put in evidence the importance of spontaneous interactive production activities and the inherent limitations of dialogue completion exercises. For tutorial CALL, they also show the necessity of developing interactive dialogue systems to allow for autonomous conversational practice of the L2.
Databáze: OpenAIRE