Teaching face-to-face is still a major education mode in many universities, yet institutions are increasingly tasked with improving efficient use of teaching spaces. This need to understand space use can be coupled with learning and teaching data to better inform student attendance and subsequently participation. Here, we analyse thermal sensor data used to monitor traffic into classrooms; these data are associated with the timetable to provide knowledge of the course and the teaching mode (such as lecture, tutorial or workshop). Further, we integrate these traffic data with student feedback data to investigate the drivers of student attendance patterns, and aim to also include online activity and behaviour to develop broad models of both room occupancy and student attendance. Combining space utilisation data with information on teaching modality and in-class and out-of-class participation can inform on how to both improve learning and design effective and efficient teaching spaces.