Elsevier

Computers in Human Behavior

Volume 52, November 2015, Pages 562-572
Computers in Human Behavior

Promoting socially shared regulation of learning in CSCL: Progress of socially shared regulation among high- and low-performing groups

https://doi.org/10.1016/j.chb.2015.03.082Get rights and content

Highlights

  • Socially shared regulation plays a critical role in successful collaboration.

  • Process discovery is used to reveal how groups progress in regulated learning.

  • High performing groups evidenced temporal variety in challenges and strategies.

  • Low performing groups were often incapable of recognizing challenges.

Abstract

Collaborative groups encounter many challenges in their learning. They need to recognize challenges that may hinder collaboration, and to develop appropriate strategies to strengthen collaboration. This study aims to explore how groups progress in their socially shared regulation of learning (SSRL) in the context of computer-supported collaborative learning (CSCL). Teacher education students (N = 103) collaborated in groups of three to four students during a two-month multimedia course. The groups used the Virtual Collaborative Research Institute (VCRI) learning environment along with regulation tools that prompted them to recognize challenges that might hinder their collaboration and to develop SSRL strategies to overcome these challenges.

In the data analysis, the groups reported challenges, and the SSRL strategies they employed were analyzed to specify the focus and function of the SSRL. Process discovery was used to explore how groups progressed in their SSRL. The results indicated that depending on the phase of the course, the SSRL focus and function shifted from regulating external challenges towards regulating the cognitive and motivational aspects of their collaboration. However, the high-performing groups progressed in their SSRL in terms of evidencing temporal variety in challenges and SSRL strategies across time, which was not the case with low performing groups.

Introduction

Collaboration is increasingly required in today’s academic contexts and in working life. Although collaboration has the potential to promote learning, it is not easy. It requires groups to cope with various tasks while coordinating between multiple individuals with unique perspectives as they attempt to achieve a shared understanding in a joint task (Dillenbourg, 1999, Rochelle and Teasley, 1995). Research has shown that groups encounter a wide range of challenges during their collaboration (Kreijns et al., 2003, Phielix et al., 2011). In order to collaborate successfully, group members need to recognize the challenges that might hinder their collaboration and to develop appropriate strategies together to overcome these challenges through interaction with others. Typically, collaborative learning research is framed by intentions to explore cognitive aspects of the collaboration (e.g. Hmelo-Silver and Barrows, 2008, Weinberger et al., 2007), including the social interaction (Kreijns et al., 2003, Rochelle and Teasley, 1995). However, despite the importance of the interactive role of motivation and emotions in collaborative learning, these factors are often ignored (Järvelä and Hadwin, 2013, Järvelä et al., 2014). The research focusing on socially shared regulation of learning (SSRL) goes beyond cognitive aspects of the collaboration by acknowledging the important role of motivation and emotions in learning. To some extent, both of these research lines complement each other when building an understanding of socially shared regulation of learning in social interaction. Socially shared regulation of learning emerges when individuals negotiate shared task perceptions, goals, plans, and strategies (Hadwin, Järvelä, & Miller, 2011), and maintain positive socio-emotional interaction during collaboration by listening and by taking each other’s opinions into consideration (Rogat & Linnenbrink-Garcia, 2011). Hadwin et al. (2011) conceptualized SSRL as unfolding in four loosely sequenced and recursively linked feedback loops. During the first loop, groups negotiate and construct shared task perceptions based on internal and external representations of the current task (Winne & Hadwin, 1998). During the second loop, groups set shared goals for the task and make plans on how to approach the task together. During the third loop, groups coordinate their collaboration strategically and monitor their progress. Based on this monitoring activity, the groups can change their task perceptions, goals, plans, or strategies to elevate their collective activity towards the shared learning goal. In essence, when groups engage in SSRL, they extend their regulatory activity from the “I” to the “we” level in order to regulate their collective activity in agreement (Hadwin & Oshige, 2011). We argue that it is exactly this “transfer in sharing” during SSRL that is essential for successful collaborative learning.

Contemporary models of regulated learning propose that social and contextual features affect SSRL, and therefore how it is shaped, are dependent on the learning situation (Hadwin et al., 2011, Winne and Hadwin, 1998). On that account, the learning situation has a mediating effect on how SSRL is constructed and shaped during collective activity. The key issue in SSRL is that it builds on and merges individual and social processes, and is not reducible to the individual. Rather, it is explained by the activity of the social entity in a learning situation, including situational affordances that provide opportunities for SSRL (Volet, Vauras, & Salonen, 2009). This means that learners need to perceive the relevant information from the environment, and to integrate that information to the previous knowledge of task, self and social context. Finally, learners need to be able to anticipate opportunities that might hinder collaborative learning. However, learners often fail to recognize the challenges that invite the development of strategies for socially shared regulation, which is a critical aspect of collaborative learning. Learners need to be aware of where the group as a whole stands with respect to challenges, and to construct adaptive regulation strategies together when the opportunity arises (Järvelä, Järvenoja, Malmberg, & Hadwin, 2013). Therefore, we argue that challenges experienced by groups provide SSRL opportunities.

The issue of awareness has received much attention in the area of CSCL research (Gutwin and Greenberg, 2002, Kirschner and Erkens, 2013). The concept of awareness in CSCL can be divided into cognitive group awareness (i.e. information about other members’ knowledge) and social group awareness (i.e. information about other members’ participation and contributions to the collaboration process). Especially in the context of CSCL it is more difficult to be aware of what the group members are actually doing and thinking. Bodemer and Dehler (2011) argued that perceiving and processing perceptions of the social awareness of other group members is a prerequisite to progress in collaborative learning (Gutwin & Greenberg, 2002). If the group members do not process this information correctly, they may face the same problem many times and their collaboration may be less satisfying.

Emerging empirical evidence indicates that SSRL increases group performance in collaborative learning (see Panadero & Järvelä, 2015). In order to ensure successful collaboration, the group members need to realize the focus of their SSRL and accordingly regulate their cognition (i.e. strategy use, task perceptions), motivation and emotions (i.e. willingness to work according to task, maintaining the socio-emotional balance) and environment (e.g. taking advantage of the features of the learning environment). Especially in the context of CSCL, regulation of all these aspects in joint agreement is equally important (Järvelä & Hadwin, 2013). Building on the results of research into collaborative learning and theories of SSRL, we posit that in order to progress in socially shared regulation of learning, the collaborating groups must engage in all of these forms of regulation. More specifically, this study aims to understand how collaborative groups progress in their socially shared regulation in computer-supported collaborative learning (CSCL).

Earlier research has shown that collaboration can result in cognitive, motivational, social, or environmental challenges (Järvelä et al., 2013, Van den Bossche et al., 2006). Cognitive challenges may derive from difficulties in understanding the other’s thinking process or in negotiating multiple perspectives (Kirschner et al., 2008, Mäkitalo et al., 2002). Motivational challenges, in turn, can emerge due to differences in group members’ emotional well-being, goals, priorities, and expectations (Rogat et al., 2013, Volet and Mansfield, 2006). By contrast, positive socio-emotional interactions, such as commitment and mutual trust, can support collaborative learning (Kreijns, Kirschner, & Vermeulen, 2013). However, negative interactions may also accumulate and result in a negative interaction loop (Linnenbrink-Garcia, Rogat, & Koskey, 2011). Especially CSCL contexts have been criticized as posing challenges for collaboration (Kreijns et al., 2013), mainly because they offer limited opportunities for social interaction: Group members might not be aware of each other’s task perceptions or goals and the group might have limited possibilities for socio-emotional interactions, which are crucial to facilitate cognitive aspects related to the group work. In other words, collaboration in the context of CSCL can pose environmental and social challenges. Therefore the concept of “sharing” becomes important in the context of CSCL.

To collaborate effectively, group members need to commit themselves to group work, establish a shared common ground, negotiate, and share their task perceptions, strategies, and goals to coordinate their collaboration (Hadwin, Oshige, Gress, & Winne, 2010). In addition, group members can also enhance their commitment to collaborative learning. For example, if the students perceive the situation as challenging because of individual differences or a lack of commitment, they can use strategies that can strengthen their feeling of togetherness and commitment by employing motivational regulation strategies (Järvenoja, Volet, & Järvelä, 2010). However, if these motivational challenges remain unsolved, the students might not be able to work according to the task demands unless the socio-emotional balance is restored between the group members (Näykki, Järvelä, Kirschner, & Järvenoja 2014). Especially in the context of CSCL, the students might confront challenges that become social challenges to regulate these social challenges effectively, the students need to realize the actual source of these challenges. That is, whether the source of the challenges is because of different goals (motivation), incompatible strategies or task perceptions (cognitive), or due to external constraints, such as time (time management), they need to activate appropriate SSRL strategies to overcome these challenges (Järvenoja & Järvelä, 2009). Ultimately, whether or not the SSRL strategy is appropriate depends on how groups can build on their previous experiences and extend their repertoire of regulatory strategies in their future collaboration.

Recently, Järvelä and Hadwin (2013) argued that it is possible to tailor and modify tools to support SSRL. In this case, the support is provided by offering, for example, structured tools that prompt students to negotiate and share their goals, plans, and strategies, along with the possibility of being able to reflect on whether the goals were achieved, whether the plans were adequate, and how effective the strategies were (Järvelä et al., 2014). Research following this line has evidenced that it is possible to tailor tools to promote SSRL (Järvelä et al., 2013). However, the results obtained from these studies have indicated that there are differences in the depth and quality of socially shared goals, plans, and strategies that groups report using. For example, Järvelä et al. (2013) identified “deep-” and “routine-level” strategies used by groups in a CSCL context. The focus of deep-level strategies was to overcome task-related challenges, whereas the focus of routine-level strategies was related to time management and environmental regulation. It can be argued that groups who used routine level strategies did not progress in their SSRL, despite being provided with the appropriate tools.

It has been argued that the way in which students engage in self-regulated learning (and SSRL) is affected by previous learning experiences, and that these experiences influence each other between and within tasks (Molenaar & Chiu, 2014). Thus, when the focus of the analysis is on the events that characterize the SSRL in the context of CSCL, these events are not independent (Cress & Hesse, 2013). Therefore, there is a need to explore how events during SSRL evolve over time.

Methodological and technological advances in the field of CSCL have made it possible to investigate temporal and sequential characteristics of varying learning activities over longer periods of time by utilizing process-oriented methods (Cress and Hesse, 2013, Puntambekar, 2013). Theories of self-regulated learning (e.g. Zimmerman & Schunk, 2011) propose a general time-ordered description of phases (e.g. planning, monitoring and control, and reflection) in which students engage when performing a task. However, there is no strong evidence concerning whether and how the phases are hierarchically or linearly structured such that earlier phases must occur before later phases (Bannert, Reimann, & Sonnenberg, 2014). In order to advance in the field of self- and socially shared regulation of learning there is a need to understand SRL or SSRL as a process rather than as a static disposition. Thus, understanding of the regulated learning process has “the potential to transform contemporary conceptions of SRL” (Azevedo, 2014, p. 217).

The focus of process analysis is on the timing (temporality) and order (sequences) of the activities that learners use during their collaboration. Temporality can reflect, for example, theory-based dimensions of self-regulated learning (planning, goal setting, strategic enactment, and reflecting) that students engage in during their learning (e.g., Johnson, Azevedo, & D’Mello, 2011). Temporality can also be defined statistically by taking advantage of breakpoints that occur during collaboration by using statistical discourse analysis (Molenaar & Chiu, 2014), or by locating typical patterns that occur at different points in time by using educational data mining (Malmberg, Järvenoja, & Järvelä, 2013). In the field of self-regulated learning, the use of these process-oriented methods has evidenced the differences in the degree and quality of self-regulatory processes used by high- and low-achieving students (e.g., Bannert et al., 2014). Similar results have been obtained in the SSRL field, but research evidence is scarce (Järvelä, Malmberg, & Koivuniemi, submitted for publication). We still have only limited understanding of how students’ progress in SSRL during longer timeframes, and how this progress affects their learning outcomes.

Hitherto, only a few studies have used process-mining techniques in the field of the self-regulation of learning and SSRL (e.g. Bannert et al., 2014; Azevedo, 2014, Schoor and Bannert, 2012). Process-mining techniques are typically used in the business context, for example, to reveal product-management processes based on the actual flow of information. In this context, the advantage of investigating process models may be to reveal gaps or “bottlenecks” in the information flow (e.g. Paszkiewicz, 2013). Similarly, process mining can be adopted when investigating the processes of the self- and socially shared regulation of learning (Bannert et al., 2014). As a methodological approach, process mining is similar to educational data parsing and data mining (Hadwin, Nesbit, Jamieson-Noel, Code, & Winne, 2007). Data mining allows for the investigation of detailed information about learner activities, such as which activities learners use, and when exactly these activities are used. Data parsing illustrates the sequences of the mined activities (Nesbit, Zhou, Xu, & Winne, 2007). Similarly, the process-mining technique reveals the most dominant processes that learners (or groups of learners) engage in when learning. Recently, Schoor and Bannert (2012) used a process-mining technique to investigate whether there were differences in SSRL processes between high- and low-achieving dyads. In their analytical approach, they identified 17 categories that were SSRL markers. In particular, they investigated how these 17 categories were present in the process models of their students’ collaboration during 90-min task executions. In their analysis, Schoor and Bannert (2012) used the fuzzy miner, which stresses the most important paths between events (in their analysis, SSRL events), while the less important paths are abstracted from a process model (Günther & Van Der Aals, 2007).

The key aspect in the investigation of process models is to identify markers that characterize successful learning. This means information on how successful groups progress in the task in comparison to less successful groups. Hitherto, there is not much research available considering SSRL as a process. Therefore, an analysis of extreme groups allows better understanding of the key features of the progress in SSRL (c.f. Bannert et al., 2014). Our hypothesis was that challenge episodes invite students to regulate their learning and that the way in which groups engage in SSRL is affected by previous learning experiences. Following these guidelines, we investigated how SSRL evolves over time and across multiple learning situations.

This study aims to elucidate how collaborative groups progress in their socially shared regulation in the context of computer-supported collaborative learning. The research questions are: (1) How do the groups regulate the challenges they have identified?; (2) How do the groups progress in their socially shared regulation of learning in the context of computer-supported collaborative learning?; and (3) Is there any difference in the progress of SSRL between low- and high-performing groups?

Section snippets

Participants and context

The first-year teacher-education students (N = 103, mean age 24.2 years) participated in a multimedia learning course that lasted for two months. The course was compulsory for the students and it was rated as either a pass or fail. The course consisted of nine collaborative face-to-face learning sessions lasting for 90 min each and nine following collaborative online learning sessions lasting for one week in which the students worked collaboratively in 30 groups of 3–4 members. The students were

Qualitative content analysis

The students’ identified challenges and socially shared regulation strategies were coded through qualitative content analysis (Miles & Huberman, 1994) from the students’ responses from OurEvaluator (F = 270). In the analysis, each response with respect to challenges (F = 270) and activated regulation strategies (F = 270) was read through carefully. Despite the question in OurEvaluator asking for the “main challenge” and consecutive regulation, the groups reported one to three challenges in their SSRL

How do the groups regulate the challenges that they have identified?

In order to provide a holistic picture of how the 30 groups engaged in the SSRL, the types of “challenge–SSRL” pairs to emerge during the collaborative tasks in terms of their focus and function were examined. In Table 4, all of the “challenge–SSRL” pairs (n = 6), which are also presented in a process model that characterizes how the collaborating groups progress in their SSRL, are presented. In addition, descriptive information about how many groups used the specific “challenge–SSRL” pair, the

Discussion and conclusion

The results indicate that despite challenges emerging due to the technology use or time-management problems, the groups were able to recognize these challenges and strategically regulate the cognitive and motivational aspects of the collaborative task execution. However, it was apparent that the focus and function of the SSRL were slightly different between the low- and high-performing groups. Mostly, the high-performing groups progressed in their SSRL in terms of evidencing temporal variety in

Acknowledgement

This research was supported by the Council of Cultural and Social Science Research, Academy of Finland. Grant No. 259214.

References (60)

  • R. Morris et al.

    Designing roles, scripts, and prompts to support CSCL in gStudy

    Computers in Human Behavior

    (2010)
  • P. Näykki et al.

    Socio-emotional conflict in collaborative learning: A process-oriented case study in a higher education context

    International Journal of Educational Research

    (2014)
  • C. Phielix et al.

    Group awareness of social and cognitive performance in a CSCL environment: Effects of a peer feedback and reflection tool

    Computers in Human Behavior

    (2011)
  • C. Schoor et al.

    Exploring regulatory processes during a computer-supported collaborative learning task using process mining

    Computers in Human Behavior

    (2012)
  • A. Weinberger et al.

    Knowledge convergence in collaborative learning: Concepts and assessment

    Learning and Instruction

    (2007)
  • R. Azevedo

    Issues in dealing with sequential and temporal characteristics of self- and socially-regulated learning

    Metacognition and Learning

    (2014)
  • R. Azevedo et al.

    New science of learning: Computers, cognition, and collaboration in education

  • M. Bannert et al.

    Process mining techniques for analysing patterns and strategies in students’ self-regulated learning

    Metacognition and Learning

    (2014)
  • M. Boekaerts

    Emotions, emotion regulation, and self-regulation of learning

  • Clow, D. (2013). MOOCs and the funnel of participation. In Proceedings of the third international conference on...
  • U. Cress et al.

    Quantitative methods for studying small groups

  • P. Dillenbourg

    What do you mean by collaborative learning?

  • J. Fleiss

    Statistical examples for rates and proportions

    (1981)
  • C.W Günther et al.

    International conference on business process management

  • C. Gutwin et al.

    A descriptive framework of work of workspace awareness for real-time groupware

    Computer Supported Cooperative Work

    (2002)
  • Hadwin, A. F., Miller, M., & Webster, E. A. (2012). My planner: Personal planning and reflection tool (Version 1.0)....
  • A.F Hadwin et al.

    Self-regulated co-regulated and socially shared regulation of learning

  • A.F. Hadwin et al.

    Examining trace data to explore self-regulated learning

    Metacognition and Learning

    (2007)
  • A.F. Hadwin et al.

    Self-regulation, coregulation, and socially shared regulation: Exploring perspectives of social in self-regulated learning theory

    Teachers College Record

    (2011)
  • C.E. Hmelo-Silver et al.

    Facilitating collaborative knowledge building

    Cognition and instruction

    (2008)
  • Cited by (120)

    View all citing articles on Scopus
    View full text