ENHANCING DESCRIPTIVE WRITING VIA PROJECT-BASED LEARNING: A QUASI-EXPERIMENTAL STUDY WITH EIGHTH-GRADE STUDENTS

Abstract

Descriptive writing remains one of the most challenging text types for junior high school EFL students, requiring the simultaneous development of content, organization, grammar, vocabulary, and mechanics — demands that conventional, teacher-centered instruction often fails to fully address. This study investigated the effectiveness of Project-Based Learning (PjBL) in improving descriptive writing achievement among eighth-grade students, using a quantitative, quasi-experimental two-group pretest-posttest design. From a population of 351 eighth-grade students, 70 were selected through purposive sampling and assigned to an experimental group (n = 35), taught using PjBL, and a control group (n = 35), taught using conventional methods. Writing achievement was measured using an expert-validated, highly reliable descriptive writing test scored across five weighted aspects (content, organization, grammar, vocabulary, mechanics), and data were analyzed through normality and homogeneity tests, N-Gain analysis, and an independent samples t-test. The experimental group's mean score rose from 66.54 (pretest) to 86.09 (posttest), a gain of 19.54 points and an N-Gain of 56.47% (medium category), compared to the control group's rise from 58.89 to 71.63, a gain of 12.74 points and an N-Gain of only 30.12% (low category) — a difference of more than 26 percentage points. The independent samples t-test confirmed this gap was statistically significant (p = 0.000 < 0.05). These findings demonstrate that PjBL's collaborative, project-based structure produces substantially greater gains in descriptive writing achievement than conventional instruction, offering empirical support for its adoption as a more effective pedagogical alternative in junior high school EFL writing classrooms, particularly within the framework of Indonesia's Merdeka Curriculum.