<![CDATA[This study aimed to evaluate the effectiveness of a deep learning-based instructional model in improving critical listening skills in higher education. A quantitative approach was employed using a quasi-experimental design with a nonequivalent control group. Two classes from Indonesian Language and Literature Education programs at two state universities with “A” accreditation were purposively selected. The instrument was a seven-item essay test that had been validated for reliability, item difficulty, and discrimination power. Assumption checks (Shapiro–Wilk and Levene tests) confirmed data normality and homogeneity of variance. Baseline differences were analyzed using independent-samples t-tests; within-group gains were tested with paired-samples t-tests; between-group posttest differences with independent-samples t-tests; and learning efficiency was summarized using the N-gain index. Both groups demonstrated significant improvement from the pretest to the posttest (p < 0.001). The experimental group achieved a higher mean posttest score of 83.54 ± 7.33 compared to the control group, which was 79.63 ± 5.31, representing a statistically significant difference (p = 0.012). The experimental group also demonstrated a significantly higher N-gain (p = 0.020), with a high-gain category (13.51%) observed only in this group. Deep learning-based instruction proved more effective than conventional teaching in enhancing critical listening skills, both in absolute achievement and learning efficiency. These findings support the adoption of deep learning pedagogy as an instructional strategy to strengthen students’ critical literacy and 21st-century competencies in higher education.]]>
Copyrights © 2026
