<![CDATA[Although the Gallery Walk learning model and self-confidence are recognized as important factors in developing communicative competence, limited research has examined their combined influence on students’ English speaking skills in English as a Foreign Language (EFL) classrooms. This study aimed to investigate the effects of the Gallery Walk learning model, students’ self-confidence, and their interaction on the speaking skills of eleventh-grade students at SMAN 1 Masbagik. A quantitative quasi-experimental design with a 2 × 2 factorial structure was employed. The sample comprised 80 students selected through simple random sampling, with 40 students assigned to the experimental group and 40 to the control group. Data were collected using a speaking performance test and a self-confidence questionnaire and analyzed through descriptive statistics, normality and homogeneity tests, an independent-samples t test, and two-way analysis of variance. The findings showed that students taught through the Gallery Walk learning model achieved significantly higher speaking scores than those receiving conventional instruction. The two-way analysis of variance further revealed significant effects of the learning model, F = 19.488, p < .001, self-confidence, F = 74.041, p < .001, and their interaction, F = 76.059, p < .001, on students’ speaking achievement. These findings demonstrate that English speaking performance is shaped by both instructional strategy and learners’ psychological readiness. The study concludes that integrating collaborative, learner-centered instruction through the Gallery Walk model with efforts to strengthen students’ self-confidence can effectively enhance speaking competence. This study contributes to communicative language teaching by providing empirical evidence of the interaction between instructional and affective factors and offers practical guidance for designing EFL speaking activities that promote active participation, confidence, and oral communication. Future research should examine the model across broader educational contexts and incorporate additional affective variables.]]>
