Use of multimodal data value chain as a contribution to the management of the teaching-learning process in higher education institutions

In education, data collection from students and teachers has occurred in physical spaces and, recently, more frequently in digital spaces. For this reason, the interaction of students and technologies offers an opportunity for multimodal data collection. We present the initial conceptual model of the Multimodal Data Value Chain (M-DVC). It clearly extracts and systematically specifies the raw evidence of learning required for a multimodal learning analytics solution (MMLA) that processes the data and converts it into meaningful information. We followed an educational action research methodology that integrated the researcher-educators into a collective process of producing and reproducing the knowledge necessary to transform the digital post-pandemic educational environment. The qualitative analysis of the MDVC conceptual model's processes made it possible to recognize the institution's characteristics. The analyses occurred in the macro (institution), meso (training program), and micro (subjects) contexts. The results defined the characteristics expected to be crucial for pedagogical decision-making based on results and reliable sources.