To support learning, Learning Analytics Dashboard (LAD) is a tool present in most training programs. By relying on learners' learning traces, the LAD offers many perspectives to support students' success.
This PhD thesis contributes to questions related to the LAD adaptation for higher education students. We propose three contributions to explore LAD that are adapted, adaptable, and adaptive.
First, we explore co-design methods, both in face-to-face and online, to design LAD adapted to the student target, and through this first contribution, the adaptation of tangible tools to digital tools and collaboration.
Then, we focus on the elements that make up LAD, the indicators and their visualizations, to make them adapted and adaptable according to the field of study and the year of study.
Finally, we explore the topic of adaptive LAD over time, based on users' expectations and usage, before proposing an adaptive LAD model. Our work has concretely allowed us to propose the tools PADDLE and ePADDLE and thus allowing us to successfully conduct co-design sessions with n=386 students from different backgrounds.
Based on the data collected, we have identified various needs expressed by students for indicators and visualizations, according to several variables such as the objective of the LAD, the type and level of study, the learning context, and time. Several perspectives are opening up for the continuation of this work, including the implementation of the adaptive LAD model.