University of Oulu

Examining deliberative interactions for socially shared regulation in collaborative learning

Saved in:
Author: Dang, Huong1
Organizations: 1University of Oulu, Faculty of Education, Educational Sciences
Format: ebook
Version: published version
Access: open
Online Access: PDF Full Text (PDF, 2 MB)
Pages: 53
Persistent link:
Language: English
Published: Oulu : H. Dang, 2023
Publish Date: 2023-05-23
Thesis type: Master's thesis
Tutor: Nguyen, Khánh
Lämsä, Joni
Reviewer: Lämsä, Joni
Törmänen, Tiina


Socially shared regulation in learning (SSRL) is essential for collaborative problem-solving and innovation that are required in today’s intricated and interconnected world. Recent advancements in learning analytics (LA) and artificial intelligence (AI) have shown promising potential for delivering a more comprehensive understanding of the temporal and cyclical processes of SSRL. It remains lacking, however, a validated standard for integrating theoretical constructs, methodological assumptions, and data structure in the field, which leads to a misalignment between the theoretical and technical aspects. This thus sparks a pressing need for interdisciplinary efforts to revise and devise theoretical and methodological frameworks that take these factors into consideration. In line with this call, the thesis presents a novel approach to applying AI to advance the field of SSRL. It comprises two empirical studies that employed AI-enabled techniques to (1) record and retain qualitative information from video data of group collaboration and (2) analyse their interaction. In particular, the studies examined the sequences of group-level interactions from the theoretical perspective of SSRL and a more micro-lens of deliberative negotiation. The theoretical framework of these studies is based on the recent conceptualisation of regulation triggering events as specific events (often negative incidents or obstacles) that stimulate regulatory responses and aid in locating them. The pattern of group interactions in response to different triggering events was then examined using processing mining and unsupervised AI machine learning clustering, agglomerative hierarchical clustering (AHC).

The findings suggest that regulation triggering events prompt an immediate shift in group interaction responses, in which they engage in more metacognitive and socioemotional interaction. Two types of deliberation sequences were identified through AHC analysis, with differing regulation and collaboration practices: the plan and implementation approach (PIA) and the trials and failures approach (TFA). A key observation of this study is that the shift in group interaction sequence in response to the regulatory trigger is only temporary. The majority of groups soon revert to or maintain the initial type of deliberation sequence they developed at the beginning and do not adopt it in response to regulatory demands.

Theoretically, the thesis makes contributions to understanding SSRL in collaborative learning, particularly the role played by regulation triggering events and deliberation processes in finding, capturing, and modelling SSRL traces. Methodologically, this thesis demonstrates a novel human-AI collaboration approach to examine regulatory responses to triggering events through group-level deliberation to study SSRL in collaboration. Practically, the findings of this thesis suggest that educators, facilitators, and AIED tool designers need to evaluate the regulatory needs of learners and offer appropriate guidance and support in order to ensure effective collaboration.

see all

Copyright information: © Huong Dang, 2023. This publication is copyrighted. You may download, display and print it for your own personal use. Commercial use is prohibited.