Feedback provision through WhatsApp: The effects of feedback type on EFL grammar learning

Abstract

Mobile-assisted language learning (MALL) tools like WhatsApp have opened new avenues for delivering timely personalized feedback in language education (Murphy et al., 2023). On a different note, feedback type is a critical factor in language acquisition, influencing both implicit and explicit knowledge development (Ellis, 2009; Nassaji, 2020). While different types of feedback have shown potential in traditional settings, their impact within MALL contexts remains underexplored. Therefore, this study examines the effects of different almost immediate feedback types on English as a foreign language (EFL) grammar learning via extramural practice on WhatsApp. Four intact groups (N=91) of first-year pre-intermediate EFL learners from the Early Years and Primary Education degrees at a Spanish university participated in the study (M age = 19.55 years). For an academic semester, they were taught three grammar structures following the Presentation-Practice-Production approach (Larsen-Freeman, 2014) and received extra exposure through WhatsApp. Participants from three of the groups completed six WhatsApp tasks (two per structure), plus an icebreaker, and received almost immediate feedback from their teacher through this platform. One group was exposed to metalinguistic explanations (n=25), another to reformulations (n=26), and the last one to repetition prompts (n=15) (Ellis, 2009). Additionally, a control group (CG; n=25), which underwent grammar teaching but was not provided feedback through WhatsApp, was also included in the design. One week before and after the intervention, the four groups completed two grammaticality judgement tests (GJT): a timed, aimed at measuring implicit knowledge, and an untimed version, measuring explicit knowledge (Nassaji, 2020). In both tests, participants had to decide whether 72 sentences (48 items with the target structures plus 24 distractors) were grammatically accurate or not. Descriptive statistics revealed that all groups scored higher in the post-tests than in the pre-tests, obtaining higher scores in the explicit test. Paired samples t-tests and Wilcoxon signed-ranked tests also revealed that those exposed to metalinguistic explanations and the CG made significant progress from pre- to post-test in both timed and untimed GJTs. One-way ANCOVAs further yielded significant differences (p=.037) between groups in the untimed GJT, with the metalinguistic explanations group scoring significantly higher than the CG (p=.045). However, the group factor did not yield significant differences in the timed GJT (p=.226). Thus, the study not only highlights the need for more explicit, continuous, and almost immediate feedback at the pre-intermediate level, but also offers valuable pedagogical implications for the integration of MALL into the EFL classroom.