International Journal of
Information and Education Technology

Abstract

This research aims to study the characteristics of, and to develop, a digital mentor prototype using generative AI to support the character design process for game design students. Functional requirements were analyzed, followed by the development of the prototype, and finally, the evaluation of the digital mentor’s usage. This research is divided into 3 phases: 1) synthesizing the characteristics of a digital mentor, 2) prototyping a digital mentor using generative general Artificial Intelligence (AI), and 3) conducting a quasi-experimental study with a sample of 120 students enrolled in a game design course. The results were measured by assessing the students’ creativity and comparing the differences between two groups using the independent t-test. The study found that the characteristics of a digital mentor should include 3 aspects: creativity and process acceleration, providing guidance and reinforcing learning, and interaction and credibility. Expert evaluation found the quality to be very high (mean = 4.19 ± 0.28). The highest score was for creativity and for accelerating the process (mean = 4.37 ± 0.20). Nine experts evaluated the prototype and found that it was of very high quality (total mean = 4.79 ± 0.28). Additionally, the experimental results indicated that the group of students using the digital mentor had significantly higher creativity assessment scores in terms of character design than did the control group (p < 0.001), especially in terms of presentation and appearance (mean = 4.42). The research results concluded that the digital mentor powered by generative AI is highly effective and efficient when it comes to enhancing the creativity and quality of character design work done by students in game design courses.