An affordable humanoid robot peer platform for Human Robot Interaction studies 

Get Complete Project Material File(s) Now! »

Socially shared regulation model(SSRL)

Skills and competencies expected out of a learner in the 21st-century Brantsch et al. (2008) have transformed greatly since information has become more accessible, abundant and diverse. The new challenge faced by the educators of current times is in preparing the learners to effectively use and assess this information and transform it into relevant knowledge and higher-order skills. The rapid emergence and obsolescence of knowledge in the 21st-century Luna Scott (2015) demand pedagogical methods that can advance creativity, critical thinking, inquiry, collaboration and lifelong learning skills in the learner. The increased emphasis and advancements in the use of information and communication technology(ICT) Sarkar (2012) to develop Intelligent Tutoring Systems(ITS) Nye (2015) and Computer-supported learning environments (CBLE) Moos and Azevedo (2009) have greatly influenced learning interventions and related research in the past few decades. Collaboration can be defined as a coordinated, synchronous activity that is the result of a continued attempt to construct and maintain a shared conception of a problem Roschelle and Teasley (1995). Various kinds of task or non-task related social interactions such as argumentation, conflict resolution, rapport building, team orientation, mutual regulation involved in collaborative learning interactions are observed to be essential for ensuring effective learning outcomes. The early theories of collaboration in learning Dillenbourg (1999) were grounded in the socio-constructivist approach inspired from Piaget’s theories and considered collaboration as an opportunity for individual cognitive development ini-tiated by socio-cognitive conflicts. The Vygotskian socio-cultural perspective John-Steiner and Mahn (1996) on collaboration identified the development in inter-psychological and intra-psychological levels which are characterised by social speech and inner speech re-spectively. Social speech is a function of group regulation and expression while inner speech serves as a function of self-regulation and internalisation. The shared cognition approach on collaboration relates to the situated learning theory of learning and takes the social and physical contexts of the environment as an integral part of learning. Computer-supported collaborative learning environments have enabled interventions in the social processes during learning by the means of artificial pedagogical agents such as virtual agents or social robots and their roles and behaviours towards the learner. Collaborative groups can be considered as social systems consisting of multiple self-regulating individ-uals who must at the same time regulate together as a social entity Järvelä et al. (2016).
The socially shared regulation model(SSRL) proposed by Hadwin, Järvelä, and Miller Hadwin et al. (2018), brings the element of collaboration into the context by explaining the intertwining of individual and social processes in self-regulation Volet et al. (2009b). For a successful collaboration, the group of learners need to coordinate the efforts, es-tablish common ground through negotiations and share resources, perceptions, strategies and goals effectively Castelli et al. (2008). The SSRL model is in great congruence with the present scenarios of learning which are often mediated by ICT tools and happening in computer-supported collaborative learning (CSCL) settings. The model of socially shared regulation makes certain assumptions about the nature of regulation in learning to oper-ationalize the definitions of phases and modes involved. These are :
1. Regulation is a multi-faceted process containing aspects of metacognition, behaviour, motivation and cognition which are not isolated but influential on each other.
2. Regulation assumes human agency as the learner is associated with the capacity to make choices driven by purpose, intents and goals.
3. Regulation is cyclical and takes place over a period of time being shaped by the knowledge, belief and experiences of the learner.
4. Regulation is linked to personal and social experiences influenced by mental models of self, others, task and knowledge from the acquired experiences.
5. Regulation involves timely adoption to new challenges and situations emerging in the group and environment
6. Regulation is socially situated and comprises a complex dynamic exchange of knowl-edge and beliefs influenced by factors such as personal meaning, outcome utility, task value and past experiences.
The SSRL model Hadwin et al. (2011) identifies three distinct modes of regulation happening in collaborative learning environments namely: (i) self-regulation(SRL), (ii) Co-regulation (CoRL) and (iii) Socially shared regulation(SSRL) (Figure 2.7). Accord-ing to the SSRL model, self-regulation(SRL) in collaboration involves cognitive, metacog-nitive, motivational, behavioural and emotional regulation strategies employed by the learner during the interaction with other participants and the learning environment. Co-regulation in SSRL context refers to a transitional process in a learner’s acquisition of self-regulated learning, appropriated by strategic planning, execution, reflection and adapta-tion during the interactions with other learners or group members Hadwin et al. (2011). Socially shared regulation (SSRL) occurs when “deliberate, strategic and transactive plan-ning, task enactment, reflection and adaptation” are taken within a group Hadwin et al. (2018). SSRL refers to processes by which group members regulate their collective activ-ity, which is different to co-regulation of learning, where individuals’ regulatory activities are guided, supported or prompted by and with others, especially in collaborative learn-ing contexts. Coregulation is an unevenly distributed form of social regulation where the co-regulator mediates the metacognitive and cognitive activities of the co-regulated, thereby influencing the regulation of his/her learning process while SSRL is considered as an evenly distributed social regulation that arises from interactions and exchanges with a learning group Schoor et al. (2015).

Joint cognitive and metacognitive strategies in SSRL

An early case analysis Vauras et al. (2003) that explored how socially shared regulation of motivation emerges in a peer-mediated learning of mathematical problem solving among 4th-grade students, reported mirroring of egalitarian, complementary monitoring and reg-ulation over the task and transfer of learning. The existence of distinct low-level and high-level co-regulation processes was revealed Volet et al. (2009a) from the video analysis of students working on case-based projects where high-level co-regulation was commonly preceded by acts of questioning or explanation. In a comparison of the traditional way of collaboration with a technology-enabled collaboration in medical the decision making context, it was observed that early engagement and co-regulation led to better-shared un-derstanding, effective management and adaptive decision-making behaviours within the highly interactive group Lajoie and Lu (2012). A multiple case study Grau and Whitebread (2012) comprising 8 children organised into two working groups over a period of one aca-demic semester reported an increase in SRL behaviours within learners and relationships with focus on activity and social regulation. Based on an empirical study Järvelä et al. (2013) involving 18 graduate students, working in collaborative teams over an 8-week period, it was observed that supporting fellow team members to successfully regulate their learning was significantly important in achieving team goals. Another study Iiskala et al. (2011) on collaborative mathematical problem solving of dyads of high achieving pupils also suggested the use of socially shared metacognition as a relevant factor for the quality of problem solving and recommends its addition to the conceptual tools of learn-ing research. A case study Hurme et al. (2009) on two groups of pre-service primary teaching with a mathematical problem solving task concluded that emergence of socially shared metacognition as a result of regulation behaviours exhibited by highly metacogni-tive learning partners contributed to a decrease in the individual’s feeling of difficulty. On investigating the motivational aspect of SSRL through a qualitative multi-method study in-volving adaptive instrument, video-recording and group interviews, Järvelä and Järvenoja (2011) identified various activated regulation strategies involved in collaborative learning such as task structuring, social reinforcing, efficacy management, interest enhancement, socially shared goal-oriented talk and handicapping of group functioning.
Interestingly, a variation of co-regulation called directive other-regulation Rogat and Adams-Wiggins (2014) which involves one student taking control of the group’s activity re-sulted in moderate or low-quality regulation and unbalanced participation as compared to facilitative form which yielded higher-quality task contributions and regulation. Through a study of sequences of self and shared regulation activities Zheng et al. (2019) involved in a STEM task on computer-supported collaborative learning (CSCL) environment, it was revealed that higher-achieving groups were more likely to start with self-executing and end with socially shared monitoring, while the less successful group were most likely to start with executing and end with self-executing.

Motivational and emotional regulation in SSRL

Järvenoja and Järvelä (2009) explored the interplay of emotional regulation processes be-tween the individual and group levels as well as the socio-emotional challenges involved in SSRL. The results from this study involving groups of teacher education students in a collaborative learning task indicated existence and sharing of emotional and motiva-tional regulation actively influencing the individuals and group to reach goals. A video based analysis of collaborative groups of university science students reported divergent patterns of engagement related to differences in perception of individual and group goals. Additionally, Rogat and Linnenbrink-Garcia (2011) concluded that negative socioemo-tional interactions would diminish the quality of social regulation in SSRL scenarios. In a study emphasising recognition and response to socio-emotional challenges in SSRL con-text Näykki et al. (2014), it was observed that avoidance-focused strategies for restoring emotional balance in the group came at the cost of compromised learning and group per-formance. The results from study on process discovery Malmberg et al. (2015) indicated that SSRL focus and functions varied temporally from regulating the task and environment towards cognitive and motivational aspects, significantly in higher-performing groups.
On observing regulation and socio-emotional interaction in positive and negative group climate, Bakhtiar et al. (2018) identified four distinct group features where (a) incoming conditions served as a foundation for creating a positive collaborative experience, (b) reg-ulation of emotions during initial planning, (c) negative emotions served as a constraint for shared adaptation in the face of a challenge, and (d) encouragement and motiva-tional statements served as effective strategies for creating a positive climate. Another process-oriented study Isohätälä et al. (2020) to understand socio-emotional layers of SSRL suggested that social interactions promoted more active participation behaviours than cognitive interactions. The behavioural changes in participation were also aligned to the shifts between domain focused and metacognitive activities. In another study us-ing S-REG tool Järvenoja et al. (2020) that traces the emotional and motivational states of the group in different sessions, co-regulation was observed to be occurring more fre-quently while socially shared regulation episodes lasted longer. Also, the emotional and motivational states were associated with the occurrence of co-regulation at the beginning of learning sessions.

READ  Introduction to the thermodynamics of small systems and the spin variable method 

Learning and performance outcomes of SSRL

The study performed by Summers and Volet (2010) observes a positive correlation of group performance with individual high-level contributions during group work but not necessarily equal collaborative learning as an outcome of the activity. Another study Janssen et al. (2012) on students engaging in a CSCL environment observed that reg-ulation of social activities positively affected group performance, while social interaction negatively affected group performance. The relationship between productive engagement in cognitive activity and metacognitive regulation in collaborative learning was explored by Khosa and Volet (2014) and they found differences in learning outcome of groups en-gaging in better social regulation. On investigating the relationship between individual self-regulation, socially shared regulation and group performance, Panadero et al. (2015) reported an influence of SRL on SSRL but no significant effect on the group performance Janssen et al. (2009). On observing the effect of group member familiarity in collabora-tive groups, it was observed that higher familiarity led to more critical and exploratory perceptions and familiar groups had to devote less time for regulating task related activ-ities. Zheng et al. (2017) reported improvements in participant’s learning achievements, group performance and regulation frequency while comparing the experiment group in-volving SSRL approach to a control group based on traditional collaborative tasks. Janssen et al. (2012) studied task-related and social regulation behaviours in online collaborative learning and identified four broad categories of collaborative activities emerging in SSRL which are, (i) discussion of information, (ii) regulation of task-related activities, (iii) reg-ulation of social activities, and (iv) social activities. Also, it was observed that the learners dedicated a good amount of time and effort for task-related and social activities but their regulation did not positively affect group performance. This can be attributed to the fac-tors such as task difficulty and engagement which would demand higher attention for task and social regulation behaviours.

Summary and insights

The field of SRL research and its adaptation in the educational domain has seen signif-icant progress over the last two decades. Each model discussed here have contributed to the deeper understanding of SRL and its various implications in the learner. All pro-posed models identify self-regulation as a cyclical process that occurs in various phases. In general, these phases can be categorised into three Panadero (2017) which are.
1. preparatory phase: which involves subprocesses such as forethought, identification, interpretation, task representation, goal setting, primary and secondary appraisals, social reinforcement etc.
2. performance phase: when the learner applies the planned strategies towards achiev-ing personal and group goals through the task activity through subprocesses of monitoring and control of performance, regulation of motivation, affect and effort, metacognitive awareness of cognition such as feeling of knowledge.
3. appraisal phase: as the learner engages in self-judgement and performance eval-uation through subprocesses of reaction and reflection generating cognitive judge-ments and attributions as well as affective reactions to the generated learning out-come.

Table of contents :

I Introduction 
1 Context of the Thesis 
1.1 Introduction
1.1.1 Embodied Conversational Agents
1.2 Research Questions
1.3 Summary of contributions
1.4 Thesis Structure
1.5 Publications
II Theoretical Background 
2 Theoretical Background 
2.0.1 Learning and Theories of pedagogy
2.0.2 Self-regulated learning
2.0.3 Models of SRL
2.1 Socially shared regulation model(SSRL)
2.1.1 Evidences of SSRL
2.2 Summary and insights
III Related Work 
3 Related Work 
3.1 Evolution of pedagogical agents
3.1.1 Computer-assisted scaffolding
3.1.2 Computer-Based Learning Environments(CBLEs)
3.2 Theories of pedagogical agents
3.3 Pedagogical agent research characteristics
3.3.1 Levels of agent design
3.3.2 Agent embodiment
3.4 Roles of pedagogical agents
3.5 Impact of pedagogical agents
3.5.1 Social and motivational outcomes
3.5.2 Learning and performance outcomes
3.6 Self-regulation and pedagogical agents
3.6.1 SSRL based systems
3.7 Limitations and Challenges
3.8 Conclusion
IV Shared Learning Interaction Design 
4 Dimensional framework of pedagogical agent roles in SSRL context
4.1 Introduction
4.1.1 Regulation in learning
4.1.2 Roles of agents
4.2 Relational Framework
4.2.1 Social Attitude
4.2.2 Regulation modes
4.3 Multimodal features of pedagogical agent roles
4.3.1 Roles for External regulation
4.3.2 Roles for Co-regulation
4.4 Conclusion
5 CardBot: An affordable humanoid robot peer platform for Human Robot Interaction studies 
5.1 Social robots in education
5.2 Wizard of Oz interactions in HRI
5.3 Design and Architecture
5.4 Conclusion and future perspectives
6 FRACTOS – Learning to be a Better Learner by Building Fractions 
6.1 Introduction
6.2 Learning topic: Fractions
6.3 Task Framework
6.3.1 Agent implementations
6.3.2 Task Levels
6.3.3 Phases
6.4 Task instance
6.5 Wizard of Oz (Woz) Implementation
6.6 Conclusion
7 Multi-agent triadic learning interaction design in socially shared regulation context 
7.1 GRETA VIB
7.2 Generation of communicative gestures
7.3 Triadic learning interaction for SSRL
7.4 Conclusion
V User Studies 
8 User Study 1: Understanding user’s perception of agent roles, behaviours and the learning activity 
8.1 Introduction
8.2 Related work
8.2.1 Multi-agent learning interactions
8.2.2 Self-regulated learning interactions
8.3 Research objectives
8.4 Methodology
8.4.1 System design
8.4.2 Questionnaires
8.4.3 Procedure
8.5 Hypotheses
8.6 Analysis and Results
8.6.1 Participants
8.6.2 Attitudes toward agents
8.6.3 Activity perception
8.6.4 Role perception
8.6.5 Agent perception
8.6.6 Self-regulation behaviour
8.6.7 Learning gain
8.7 Discussion
8.8 Conclusion
9 User Study 2: Understanding the impact of error making peer agent behaviours on user perceptions and self-regulation 
9.1 Introduction
9.2 Related work
9.2.1 Error making peer agents
9.3 Research questions
9.3.1 Hypotheses
9.4 Methodology
9.4.1 System design
9.4.2 Questionnaires
9.4.3 Procedure
9.5 Analysis and Results
9.5.1 Participants
9.5.2 Attitude towards agents
9.5.3 Activity perception
9.5.4 Agent perception
9.5.5 Role perception
9.5.6 Self-regulation behaviour
9.5.7 Learning gain
9.5.8 Comparison with Study 1
9.6 Discussion
9.7 Conclusion
10 User Study 3: Exploring distinct modes of regulation scaffolding in SSRL context 
10.1 Related work
10.1.1 Modes of regulation
10.2 Research questions
10.3 Methodology
10.3.1 Study design
10.3.2 Questionnaires
10.4 Hypotheses
10.5 Analysis and Results
10.5.1 Participants
10.5.2 Agent Perception
10.5.3 Activity Perception
10.5.4 Role perception
10.5.5 Task performance
10.5.6 Self-regulation behaviours
10.6 Discussion
10.7 Conclusion
11 User Study 4: User-driven approach for regulation scaffolding in a shared learning interaction 
11.1 Introduction
11.2 Research questions
11.2.1 Hypotheses
11.3 Methodology
11.3.1 Questionnaires
11.3.2 System design
11.4 Analysis and Results
11.4.1 Participants
11.4.2 Agent Perception
11.4.3 Role perception
11.4.4 Activity Perception
11.4.5 Performance
11.4.6 Self-regulation behaviours
11.4.7 User preferences
11.5 Discussion
11.6 Conclusion
VI Conclusion 
12 Conclusion and Perspectives 
12.1 Summary of Contributions
12.2 Limitations
12.2.1 Agent and task limitations
12.2.2 Experimental limitations
12.3 Future perspectives
12.3.1 Pedagogical role combinations for SSRL
12.3.2 Online measurement of SRL
12.3.3 Automaticity and Adaptive scaffolding
VII Annexes 
A Publications and Dissemination 
B Questionnaires 
Bibliography 

GET THE COMPLETE PROJECT

Related Posts