This study was conducted with the aim of determining (1) the ability to write procedural texts using the direct learning model, (2) the ability to write procedural texts using the time token model, (3) determining the effectiveness of the time token model as a model that supports procedural text writing skills in seventh-grade students of Johannes Prodromos Christian Middle School. The subjects in this study were seventh-grade students, with a sample of 68 students. The class using the time token model had 37 students and the class using the direct learning model had 31 students. The method used in this study was an experimental method with a two-group post-test design, with the research instrument being a procedural text writing assignment. The average score of the control class was 66.7, while the experimental class was 77. Thus, it can be concluded that the average score of the experimental class was higher than the control class. The results of the Mann-Whitney U hypothesis test showed a sig value (2-tailed) <0.001 (H1) was accepted. Therefore, it can be concluded that the time token model is indeed effective in improving procedural text writing skills.
Copyrights © 2026