The article, titled “Development of A Deep Learning Model Based on Augmented Reality to Improve Self-Efficacy and Physics Learning Outcomes of Senior High School Students in Makassar City,” was published in the Jurnal Ilmiah Pendidikan Holistik (JIPH), Vol. 5 No. 3, 2026, pages 405–410. The article is registered with DOI 10.55927/jiph.v5i3.28.
Physics Learning Still Faces Challenges with Abstract Concepts
Physics learning at the senior high school level often faces difficulties when students are required to understand phenomena that cannot easily be observed directly. Learning that focuses too heavily on memorizing formulas can also make it difficult for students to understand the connection between physics concepts and everyday life.
The article notes that Indonesia’s science literacy score in PISA 2022 was 383, considerably lower than the OECD average. In Makassar City, students’ self-efficacy in physics learning was also reported to be in the moderate-to-low category, with an average score of 3.15 on a five-point scale.
These conditions provided the basis for developing AR-Kolb. Augmented Reality technology allows three-dimensional objects and simulations of physics phenomena to be displayed through digital devices, enabling students to interact with concepts that are otherwise difficult to visualize.
However, Zia-ulHaq and his colleagues did not simply develop AR-based learning media. They combined AR technology with a learning model based on Kolb’s experiential learning cycle and a deep learning approach to encourage deeper conceptual understanding.
Combining AR, Kolb’s Experiential Learning, and Makassar’s Maritime Culture
The AR-Kolb model integrates the four stages of David Kolb’s experiential learning cycle: concrete experience, reflective observation, abstract conceptualization, and active experimentation.
In practice, students first gain experience through AR simulations. They then observe and reflect on the phenomena presented, connect them with physics concepts and formulas, and finally apply their knowledge to new situations.
Another distinctive feature is the use of Makassar’s local maritime context. The simulations include objects and phenomena such as pinisi boats, ocean waves, force fields, and traditional games. Through this approach, physics is presented not merely as theoretical knowledge but as something connected to the environment and culture familiar to students.
According to the findings reported by Muhammad Zia-ulHaq and his team from Universitas Negeri Makassar, this approach makes the learning experience more relevant while giving students opportunities to develop confidence in their own abilities.
Tested on 144 Senior High School Students in Makassar
The development of AR-Kolb employed a Research and Development approach that combined the Borg and Gall framework, ADDIE, and the Bergman and Moore model. The stages included needs analysis, design, prototype development, expert validation, limited trials, field testing, revision, and product finalization.
A total of 144 Grade 11 science students from four senior high schools in Makassar participated in the testing. The schools were SMA Negeri 9, SMAS Zion, SMA Negeri 2, and SMA Katolik Rajawali.
A total of 72 students were assigned to the experimental group and received instruction using the AR-Kolb model, while another 72 students were placed in the control group and received conventional instruction.
The researchers measured students’ self-efficacy, physics learning outcomes, the practicality of the model, and the implementation of each stage of the learning cycle. The self-efficacy instrument was adapted from Bandura’s scale, with a Cronbach’s alpha reliability coefficient of at least 0.87.
Students’ Self-Efficacy Improved More Significantly
The differences between the AR-Kolb group and the control group were evident in the scores obtained before and after instruction.
For self-efficacy, the experimental group’s average score increased from 3.15 to 4.28. Its N-Gain score reached 0.68. Meanwhile, the control group increased from 3.12 to 3.48, with an N-Gain score of 0.19.
For physics learning outcomes, the AR-Kolb group’s average score increased from 52.4 to 84.7, with an N-Gain score of 0.68. The control group increased from 51.9 to 66.8, with an N-Gain score of 0.31.
ANCOVA analysis also showed a significant difference between the experimental and control groups in the post-test scores for both variables, with p < 0.001 after controlling for differences in the initial scores.
Expert validation produced an average score of at least 91 percent, placing the AR-Kolb model in the highly valid category. Practicality testing conducted with teachers and students also reached at least 89 percent, placing the model in the highly practical category.
Simulations Help Students Become More Confident
The improvement in self-efficacy is one of the study’s important findings. In physics learning, students’ belief in their own abilities can influence their motivation, persistence, and willingness to tackle difficult problems.
AR provides students with opportunities to practice simulations safely and repeatedly. When students successfully manipulate three-dimensional objects or understand the relationship between physical phenomena and scientific concepts, these successful experiences can strengthen their belief in their ability to solve problems.
Zia-ulHaq and his colleagues explain that concrete experiences through AR are reinforced by reflection, conceptual development, and active experimentation. This integration allows students to move beyond memorizing formulas and develop an understanding of how physics concepts work and how they can be applied in different situations.
Potential for Schools with Limited Infrastructure
The AR-Kolb model was also designed in an offline format. This feature is considered important for schools that continue to face limitations in internet access or digital infrastructure.
The researchers recommend that senior high school physics teachers use the model, particularly when teaching abstract topics. The instructional package includes an offline AR application, teaching modules, student worksheets, and TPACK guidelines.
Physics Subject Teacher Working Groups (MGMP) and education authorities can also consider adopting the model for broader implementation. Local governments are encouraged to support the provision of basic digital devices and teacher training so that the model can be replicated in other schools.
Nevertheless, the study was limited to four senior high schools in Makassar City and several specific physics topics. Future research is recommended to test the model in more diverse areas of South Sulawesi, including topics such as modern physics and thermodynamics, as well as to examine its long-term effects on knowledge retention and students’ interest in STEM careers.
Author Profile
Muhammad Zia-ulHaq is the lead author of the article and is affiliated with Universitas Negeri Makassar, Indonesia. He co-authored the article with Arismunandar, Abd. Haris, and Suciani Latif, all of whom are affiliated with Universitas Negeri Makassar. The article identifies Arismunandar as the promoter, Abd. Haris as the co-promoter, and Suciani Latif as a member of the promoter team. The article focuses on physics learning development, deep learning, Augmented Reality, experiential learning, and the improvement of students’ self-efficacy.
0 Komentar