Explaining trace‐based learner profiles with self‐reports: The added value of psychological networks
- 4 March 2024
- journal article
- research article
- Published by Wiley in Journal of Computer Assisted Learning
- Vol. 40 (4) , 1481-1499
Abstract
Background: Learner profiles detected from digital trace data are typically triangulated with survey data to explain those profiles based on learners' internal conditions (e.g., motivation). However, survey data are often analysed with limited consideration of the interconnected nature of learners' internal conditions.Objectives: Aiming to enable a thorough understanding of trace‐based learner profiles, this paper presents and evaluates a comprehensive approach to analysis of learners' self‐reports, which extends conventional statistical methods with psychological networks analysis.Methods: The study context is a massive open online course (MOOC) aimed at promoting physical activity (PA) for health. Learners' (N = 497) perceptions related to PA, as well as their self‐efficacy and intentions to increase the level of PA were collected before and after the MOOC, while their interactions with the course were logged as digital traces. Learner profiles derived from trace data were further examined and interpreted through a combined use of conventional statistical methods and psychological networks analysis.Results and Conclusions: The inclusion of psychological networks in the analysis of learners' self‐reports collected before the start of the MOOC offers better understanding of trace‐based learner profiles, compared to the conventional statistical analysis only. Likewise, the combined use of conventional statistical methods and psychological networks in the analysis of learners' self‐reports before and after the MOOC provided more comprehensive insights about changes in the constructs measured in each learner profile.Major Takeaways: The combined use of conventional statistical methods and psychological networks presented in this paper sets a path for a comprehensive analysis of survey data. The insights it offers complement the information about learner profiles derived from trace data, thus allowing for a more thorough understanding of learners' course engagement than any individual method or data source would allow.Keywords
This publication has 64 references indexed in Scilit:
- From procrastination to engagement? An experimental exploration of the effects of an adaptive virtual assistant on self-regulation in online learningComputers and Education: Artificial Intelligence, 2022
- Single-Case Learning Analytics to Support Social-Emotional Learning: The Case of Doctoral Education2022
- Towards investigating the validity of measurement of self-regulated learning based on trace dataMetacognition and Learning, 2022
- Learning Analytics to Reveal Links Between Learning Design and Self-Regulated LearningInternational Journal of Artificial Intelligence in Education, 2021
- SENS: Network analytics to combine social and cognitive perspectives of collaborative learningComputers in Human Behavior, 2019
- Exploring the potential of LMS log data as a proxy measure of student engagementJournal of Computing in Higher Education, 2017
- pMineR: An Innovative R Library for Performing Process Mining in Medicine2017
- Learning Strategies, Study Skills, and Self-Regulated Learning in Postsecondary Education2013
- Dimensions of Normal Personality as Networks in Search of Equilibrium: You Can't like Parties if you Don't like PeopleEuropean Journal of Personality, 2012
- The theory of planned behaviorOrganizational Behavior and Human Decision Processes, 1991