<![CDATA[This study aims to describe students' oral argumentation skills in expository text discussions without using the Learning by Debate model based on the Deep Learning approach, and to determine the effect of implementing this model on the oral argumentation skills of Grade XI students at SMA Negeri 5 Medan. The method used is a true experimental design with a Two Group Posttest-Only Control Design. The research population consisted of all Grade XI students at SMA Negeri 5 Medan (429 students), while the sample comprised two classes selected through cluster random sampling: class XI-7 as the experimental group and class XI-2 as the control group (36 students each). The experimental class received treatment using the Learning by Debate model integrated with Deep Learning principles (mindful, meaningful, and joyful), while the control class used conventional learning. Oral argumentation ability was measured using a performance test based on four indicators by Macagno and Walton which include clarity and coherence of arguments, relevance and sufficiency of reasons, critical response ability, and persuasive delivery skills. The results showed that the experimental class obtained a mean score of 76.83 (highest score 95, lowest 60), while the control class obtained a mean score of 68.42 (highest score 85, lowest 45). Based on an independent t-test at a significance level of α = 0.05 with df = 70, the t-count value of 3.77 exceeded the t-table value of 1.994, so H₀ was rejected and Hₐ was accepted. Thus, there is a significant effect of implementing the Learning by Debate model based on the Deep Learning approach on students' oral argumentation skills in expository text discussions. ]]>
Copyrights © 2026
