Using Event Sequence Analysis to Uncover Self-Regulatory Behaviour: Towards Design Guidelines
Full recordShow full item record
AbstractThe analysis of behavioural data provides new avenues for investigating the sequential nature of self-regulation. This contribution describes the data-driven approach used in a study that aimed to identify the relationship between learners’ characteristics and their self-regulatory behaviour by performing event sequence analysis on log files extracted from ecologically valid online learning environments. In the first phase of the study, we described the instructional context as an external condition in order to map any aspects that might support self-regulation. We collected data on relevant learner characteristics as internal conditions and extracted timestamped log-file data as indicators of learners’ self-regulatory behaviour. Next, the data was cleaned and recoded using an action library. The third phase focused on discovering event patterns using the TraMineR package in R to identify frequent sub-sequences (Levenshtein distance) (Gabadinho, Ritschard, Studer, & Müller, 2009). Finally, significant discriminant sub-sequences were described in relation to various learner characteristics (Pearson’s Chi-square test). The results demonstrate how the design of an online learning environment triggers different event sequences — and hence different self-regulatory behaviour — in learners with different characteristics. The study offers a starting point for discussing sequence analysis of log files extracted from ecologically valid environments, as well as the conceptual and methodological issues related to this type of research.
TypeDescription (Metadata) only