Task-based language teaching in VR versus traditional settings

This study compared Task-Based Language Teaching (TBLT) delivered in virtual reality (VR) and in traditional settings, examining learning gains, transfer, retention, and learner perceptions. 22 participants completed pretests, immediate posttests, transfer tests, and a questionnaire, while 10 of them, selected to balance proficiency across groups based on posttest scores, additionally completed a two-month delayed posttest and follow-up interviews. Wilcoxon signed-rank tests showed significant learning gains in both VR and traditional groups (p < .001; p = .006, in target discourse and p = .004; p = .008 in listening tests). ANCOVA indicated no immediate group difference in posttest scores (p = .205; p = .322) or in transfer measures (p = .608). After two months, a Mann–Whitney U test showed the VR group preserved target discourse significantly better than the traditional group (p = .036). Mann–Whitney U analyses of Likert-scale responses revealed greater enjoyment (p = .008), perceived retention (p = .009), reported difficulty (p = .01), and motivation (p = .033) in the VR group. Interview analyses (in vivo coding) highlighted VR affordances that promote contextualized practice and positive emotional engagement. These results suggest VR-TBLT may enhance long-term retention and learner engagement compared with traditional TBLT.