The Effect of the EnglishScore Media-Assisted Deep Learning Approach on TOEFL Scores in the “Structure and Written Expression” Module of the Chemistry English Course

Abstract

Low TOEFL scores among students in the Structure and Written Expression section are often caused by conventional, superficial learning methods. This study aims to analyze the effect of a deep learning approach supported by the EnglishScore media on students’ TOEFL scores in the Chemistry English course. The research method used was a quasi-experimental study with a non-equivalent control group design. The research sample consisted of two classes of second-semester Chemistry Department students at the State University of Medan, selected through purposive sampling. Data were collected through pretests and posttests, which were analyzed using descriptive statistics and the independent samples t-test via SPSS software. The results showed that the average posttest score of the experimental class (28.829) was significantly higher than that of the control class (23.229). The results of the hypothesis test showed a significance value of 0.001 (p < 0.05), indicating a significant difference in achievement between the two groups. The deep learning approach proved effective in creating a more homogeneous distribution of scores and enhancing students’ conceptual understanding through reflective and analytical activities. The integration of the EnglishScore platform played a crucial role in providing an adaptive learning environment and enhancing cognitive engagement. This study highlights the importance of transforming technology-based learning strategies to strengthen students’ English language competencies in the field of science.