The Influence of Deep Learning on Dialogic Communication in the Learning Process at Community Learning Centers

Abstract

Background: The rapid development of 21st-century education requires a paradigm shift from teacher-centered learning toward learner-centered approaches that emphasize deep learning. In the context of non-formal education, Community Learning Centers (Pusat Kegiatan Belajar Masyarakat, PKBM) serve as strategic spaces for lifelong learning. However, many still rely on conventional, one-way instructional practices that limit reflective dialogue and learner engagement. Deep learning, which focuses on critical thinking, collaboration, and knowledge transfer to real-life contexts, has the potential to transform these interactions into more dialogic and meaningful communication processes.

Research Objectives: This study aims to analyze the influence of the deep learning model on enhancing dialogic communication during the learning process at community learning center. Additionally, it seeks to examine the extent to which deep learning contributes to the achievement of UNESCO’s four pillars of learning—learning to know, learning to do, learning to be, and learning to live together.

Research Method: This research employed a mixed-methods design with an explanatory sequential approach. The quantitative phase used a quasi-experimental method involving 120 learners across six community learning center in Jakarta, divided into experimental and control groups. The experimental group implemented a deep learning model emphasizing reflective questioning and collaborative dialogue, while the control group applied conventional methods. Quantitative data were analyzed using t-tests and multiple linear regression, while qualitative data were gathered through interviews and observations, then analyzed thematically to provide contextual understanding.

Research Findings: The results showed that learners in the experimental group experienced a significant improvement in dialogic communication compared to the control group (t = 6.42; p = 0.000). Their mean score increased from 61.2 to 84.3, demonstrating enhanced participation, reflection, and critical dialogue. Multiple regression analysis revealed that deep learning was a dominant predictor of achieving the four pillars of learning (β = 0.68; p = 0.000), explaining 53% of the variance in holistic learning outcomes. Qualitative findings supported these results, illustrating that learners perceived the deep learning environment as challenging yet meaningful, fostering curiosity, collaboration, and a sense of empowerment.

Conclusion: The study concludes that deep learning significantly strengthens dialogic communication and holistic learning in non-formal education contexts. By integrating reflection, collaboration, and authentic problem-solving, this approach transforms community learning center classrooms into inclusive and participatory learning spaces. It not only enhances cognitive and social skills but also aligns with the broader mission of community empowerment through lifelong learning.

Novelty/Originality/Value:   This study provides one of the first empirical investigations linking deep learning with dialogic communication in Indonesian non-formal education settings. It offers a practical model for community learning center tutors to foster transformative learning interactions and contributes to policy discussions on enhancing pedagogical capacity in community-based education.