Gasevic, Dragan

Fan, Yizhou

Pardo, Abelardo

Saint, John

2022-02-22T11:44:01Z

2022-02-22T11:44:01Z

2022-02-23

In educational settings, the increasingly sophisticated use of digital technology has provided students with greater agency over their learning. This has focused educational research on the metacognitive and cognitive activities with which students engage to manage their learning and the achievement of their learning goals. This field of research is articulated as self-regulated learning (SRL) and has seen the development of several key theoretical models. Despite key differences, these models are broadly defined by thematic variations of the same fundamental phases: i) a preparatory phase; ii) a performance phase, and; iii) an appraisal phase. Given the phasic nature of these models, the conceptualisation of SRL as a phenomenon that unfolds in temporal space has gained much traction. In acknowledging this dimension of SRL, researchers are bound to address the methodological demands of process, sequence, and temporality. Learning Analytics research, however, is largely characterised by the use of statistical models for data interrogation and analysis. Despite their value, several researchers posit that the use of statistical methods imposes ontological limitations with respect to the temporal and sequential nature of SRL. Another challenge is that while learner data are mostly collected at the micro level, (e.g., page access, video view, quiz attempt), SRL theory is defined at a macro level (e.g., planning, monitoring, evaluation), highlighting a need to bridge this gap in order to provide meaningful results. This thesis aims to explore the methodological opportunities and address the theoretical challenges presented in the area of temporally focused SRL learning analytics. First, the thesis explores the corpus of research in the area. As such, we present a systematic review of literature that analyses the findings of studies that explore SRL through the lenses of order and sequence, to provide insights into the temporal dynamics of SRL. Second, the thesis demonstrates the use of a novel process mining method to analyse how certain temporal activity traits relate to academic performance. We determined that more strategically minded activity, embodying aspects self-regulation, generally demonstrated to be more successful than less disciplined reactive behaviours. Third, the thesis presents a methodological framework designed to embed our analyses in a model of SRL. It comprises the use of: i) micro-level processing to transform raw trace data into SRL processes; and ii) first order Markov models to explore the temporal associations between SRL processes. We call this the “Trace-SRL” framework. Fourth, using the Trace-SRL framework, the thesis explores the deployment of multiple analytic methods and posits that richer insights can be gained through a combined methodological perspective. Fifth, building on this theme, the thesis presents a systematic analysis of four process mining algorithms, as deployed in the exploration of common SRL event data, concluding that the choice of algorithm and metric is of key importance in temporally-focused SRL research, and that combined metrics can provide deeper insights than those presented individually. Finally, the thesis concludes with a discussion of practical implications, the significance of the results, and future research directions.

https://hdl.handle.net/1842/38614

http://dx.doi.org/10.7488/era/1877

en

The University of Edinburgh

Saint, J., Fan, Y., Gaševi´c, D., & Pardo, A. (2021). Temporally focused Self-regulated Learning: A Systematic Review of Literature

Saint, J., Whitelock-Wainwright, A., Gaševi´c, D., & Pardo, A. (2020). Trace-SRL: A Framework for Analysis of Microlevel Processes of Self-Regulated Learning From Trace Data [Conference Name: IEEE Transactions on Learning Technologies]. IEEE Transactions on Learning Technologies, 13(4), 861–877. https://doi.org/10.1109/TLT.2020.3027496

Saint, J., Gaševi´c, D., & Pardo, A. (2018). Detecting Learning Strategies Through Process Mining. In V. Pammer-Schindler, M. Pérez-Sanagustín, H. Drachsler, R. Elferink, & M. Scheffel (Eds.), Lifelong Technology-Enhanced Learning (pp. 385–398). Springer International Publishing. https://doi.org/10.1007/978-3-319-98572-5_29

Saint, J., Gaševi´c, D., Matcha, W., Ahmad Uzir, N., & Pardo, A. (2020). Combining analytic methods to unlock sequential and temporal patterns of self-regulated learning. Proceedings of the Tenth International Conference on Learning Analytics & Knowledge, 402–411. https://doi.o rg/10.1145/3375462.3375487

. Saint, J., Fan, Y., Singh, S., Gasevic, D., & Pardo, A. (2021). Using process mining to analyse self-regulated learning: A systematic analysis of four algorithms. LAK21: 11th International Learning Analytics and Knowledge Conference, 333–343. https://doi.org/10.1145/3448139.3 448171

learning analytics

self-regulated learning

process mining

temporal data analysis

process analytics

Analytics of self-regulated learning: a temporal and sequential approach

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy