Development of RBL-STEM learning tools to enhance students' metaliteracy in identifying facial skin types using graceful color segmentation and Deep-CNN techniques

At present, possessing literacy skills alone is insufficient to address the challenges of the globalized world. The ability to reflect on one’s cognitive processes when interacting with diverse sources of information is also required; therefore, a new concept known as metaliteracy has emerged. One model and approach that can be applied to enhance metaliteracy is Research-Based Learning integrated with Science, Technology, Engineering, and Mathematics (RBL-STEM). This study aims to identify RBL-STEM activities, to describe the processes and outcomes of developing RBL-STEM learning tools, and to examine improvements in students’ metaliteracy. The study employed a Research and Development (R&D) methodology. The research products consisted of developed learning tools, including student assignment designs, student worksheets, and learning outcome tests. The development process resulted in a validity level of 94%. The trial involved 33 students, and the results indicated that the RBL-STEM approach was effective, achieving an effectiveness score of 94.7%, and practical, with a practicality score of 94.2%. In addition, students demonstrated positive responses to the learning experience and exhibited a high level of engagement. Students’ metaliteracy improved after solving problems related to Deep Convolutional Neural Networks (Deep-CNN), as evidenced by the pretest and posttest results. The study also identified three levels of students’ metaliteracy, namely high, medium, and low. The research findings were validated through statistical analysis using the SPSS application. Therefore, RBL-STEM has the potential to enhance students’ metaliteracy in real-world contexts, such as the application of Deep-CNN with graceful color segmentation for identifying facial skin types.