In the ongoing discussion about how learning analytics can effectively support self-regulated student learning and which types of data are most suitable for this purpose, this empirical study aligns with the framework proposed by Buckingham Shum and Deakin Crick (2012) who advocated the inclusion of both behavioural trace data and survey data in learning analytics studies. By incorporating learning dispositions in our learning analytics modelling, this research aims to investigate and understand how students engage with learning tasks, tools, and materials in their academic endeavours. This is achieved by analysing trace data, which captures digital footprints of students' interactions with digital tools, along with survey responses from the Study of Learning Questionnaire (SLQ), to comprehensively examine their preferred learning strategies. Additionally, the study explores the relationship between these strategies and students' learning dispositions measured at the start of the course. An innovative aspect of this investigation lies in its emphasis on understanding how learning dispositions act as antecedents and potentially predict the utilization of specific learning strategies. The data is scrutinized to identify patterns and clusters of such patterns between students' learning disposition and their preferred strategies. Data is gathered from two cohorts of students, comprising 2400 first year students. This analytical approach aims to uncover predictive insights, offering potential indicators to predict and understand students' learning strategy preferences, which holds value for teachers, educational scientists, and educational designers. Understanding students' regulation of their own learning process holds promise to recognize students with less beneficial learning strategies and target interventions aimed to improve these. A crucial takeaway from our research underscores the significance of flexibility, which entails the ability to adjust preferred learning strategies according to the learning environment. While it is imperative to instruct our students in deep learning strategies and encourage autonomous regulation of learning, this should not come at the expense of acknowledging situations where surface strategies and controlled regulation may prove to be more effective.