Comparing self-regulation, anxiety, and willingness to communicate across blended, face-to-face, and online instruction in female EFL learners

Abstract Technology-enhanced language instruction has expanded the range of instructional options available to EFL learners, yet evidence comparing blended, online, and face-to-face modes across multiple cognitive and affective outcomes remains limited, particularly among advanced EFL learners. This study compared self-regulated learning strategies (SRLSs), foreign language classroom anxiety (FLCA), and willingness to communicate (WTC) in English writing across these three instructional modes. A quasi-experimental pretest–posttest design was conducted with 360 advanced female EFL learners in Iran (120 per group), selected through two-stage cluster sampling from language institutes. Learners participated in 10 instructional sessions delivered through blended, online, or face-to-face modes. Pre-test and post-test measures of SRLS, FLCA, and WTC were collected using self-report questionnaires. One-way ANCOVAs were conducted for each outcome, controlling for the corresponding pre-test score, followed by Bonferroni-adjusted pairwise comparisons. After adjustment for baseline scores, statistically significant between-group differences were observed for SRLS, FLCA, and WTC. The blended group had higher adjusted SRLS scores than the online and face-to-face groups (partial η² = 0.411) and higher adjusted WTC scores in English writing (partial η² = 0.266), but lower adjusted FLCA scores than both comparison groups (partial η² = 0.136). No statistically significant differences were observed between the online and face-to-face groups for any outcome. These findings identify a consistent pattern of adjusted between-group differences associated with instructional mode among advanced female Iranian EFL learners. They do not establish comparative efficacy or causal superiority of blended instruction. Pending replication using randomized designs, objective measures, and more diverse samples, the findings may inform further investigation of blended instructional arrangements and the specific pedagogical features associated with learner outcomes.