Page 219 - Jurnal Penelitian MTsN 6 Jakarta

ABSTRACT

This study aims to develop and analyze the effectiveness of an artificial
intelligence (AI)-based educational chatbot called CEMARA (Chatbot
Edukasi untuk Meningkatkan Literasi Sains or Educational Chatbot to
Enhance Scientific Literacy) in improving the scientific literacy of Madrasah
Tsanawiyah students. The research background stems from the low level
of scientific literacy among Indonesian students, as reported by PISA 2022,
and the limited availability of interactive learning media capable of
stimulating scientific thinking skills. Through conversation-based
interaction, CEMARA is designed to help students understand scientific
concepts contextually, facilitate self-directed learning, and provide
adaptive feedback to students’ inquiries. The research employed a
quantitative approach using a quasi-experimental design, specifically the
Nonequivalent Control Group Design. The sample consisted of two eighth-
grade classes at MTs 6 East Jakarta: one experimental class using the
CEMARA chatbot and one control class using conventional teaching
methods. Research instruments included a scientific literacy test (pre-test
and post-test), a student perception questionnaire, and classroom
observations. Data were analyzed using normality and homogeneity tests,
independent sample t-test, and N-Gain calculations. The results
indicated that the use of the CEMARA chatbot had a significant effect on
improving students’ scientific literacy. The t-test results for all aspects ease
of use, learning engagement, benefits for scientific literacy comprehension,
and response quality and learning satisfaction showed p-values < 0.05.
The average student perception score ranged from “agree” to “strongly
agree” (M = 4.13), while the mean N-Gain score was 0.47, categorized as
moderate, indicating consistent improvement in scientific literacy skills.
Furthermore, students reported that the chatbot was easy to use,
interactive, and capable of providing an enjoyable and meaningful learning
experience. In conclusion, the implementation of the CEMARA chatbot
proved effective in enhancing students’ understanding and engagement in
science learning. This medium serves as an innovative alternative for
developing AI-based digital learning that is adaptive and aligned with the
needs of 21st-century education. The study also provides empirical
evidence for the use of AI technology as an interactive learning partner to
strengthen scientific literacy among madrasah students.

Keywords: educational chatbot, scientific literacy, interactive learning,
artificial intelligence.

ix