A decade of empirical research and theoretical development in SRL has been successful in demonstrating significant cognitive and motivational predictions of students’ academic outcomes (Zimmermann, 2008). Although self-regulation research has historically focused on an individual perspective, there is an increasing interest in considering these processes at the social level with reference to concepts such as social regulation, shared regulation, or co-regulation. This symposium brings together teams of researchers who have grappled with social aspects in the regulation of learning using a wide range of research designs and methodologies. The purpose of this session is threefold: (a) to compare empirical methodologies and analytical techniques for examining social aspect in the regulation of learning, (b) to generate discussion about the challenges in researching and social aspects in the regulation of learning and (c) to co-construct a framework for collectively advancing research about the social-aspects in the regulation of learning.

To date very few empirical studies have considered social aspects of self-regulation in the context of complex tasks that unfold over time. Over the past 20 years, Hadwin and her colleagues have made significant progress in developing theoretical platforms, tools, and analysis methods to further research that investigates the intersect of self-regulation, co-regulation and socially-shared regulation in solo and collaborative tasks. Specifically, this paper draws from past and ongoing empirical studies to introduce methods for : (1) developing and testing measures of co-regulation and socially shared regulation; (2) extending and applying analytical tools and methods for examining those two types of regulation; and (3) conducting applied research studies where technologies are used to support the regulation of learning as it unfolds over time. A theme throughout our research is the social construction of task understanding. For example, this paper demonstrates the use of discourse analysis, and systematic task analysis processes as means for examining the appropriation of self-regulatory activity (co-regulation of learning. Evidence of shared-regulation of learning is demonstrated through self-report instruments, patterns of dialogue, and shared task understanding. The paper concludes with a discussion about the ways technologies can be used to advance research agendas regarding the socially shared aspects of self-regulation.

Introduction

Successful learners are often described as self-regulating their learning. These learners are planful, take time to check how they are doing, and deliberately change directions or strategies when learning does not go as planned. These learners take charge of learning and skillfully adapt to new learning task and situations. Self-regulating engages students in four interrelated phases of studying: defining the task, setting goals and forging plans, enacting tactics, and adapting the process (Winne & Hadwin, 1998). However, regulating learning is rarely a solitary task. We posit that it always involves social context, interaction or support at one time or another (Hadwin, 2000). In order to succeed in both solo and collaborative tasks, students need to develop skills and strategies for regulating their own learning (self-regulation), regulating with peers (co-regulation) and regulating together (socially-shared regulation).

Aim

Specifically, this paper draws from past and ongoing empirical studies to introduce methods for : (1) developing and testing measures of co-regulation and socially shared regulation; (2) extending and applying analytical tools and methods for examining those two types of regulation; and (3) conducting applied research studies where technologies are used to support the regulation of learning as it unfolds over time. A theme throughout our research is the social construction of task understanding. For example, this paper demonstrates the use of discourse analysis, and systematic task analysis processes as means for examining the appropriation of self-regulatory activity (co-regulation of learning. Evidence of shared-regulation of learning is demonstrated through self-report instruments, patterns of dialogue, shared task understanding, and shared goals.

Theoretical Perspectives and Methods 

Co-regulated learning (CoRL) refers to a transitional process in a learner’s development of SRL during which experts and learners share a common problem solving plane and SRL is gradually appropriated by the individual learner through interactions (Diaz et al., 1990; Gallimore & Tharpe, 1990; Hadwin, et al., 2005; McCaslin & Hickey, 2001). CoRL is an interactive process whereby ownership of regulatory activity and thinking is temporarily shared amongst participants. The focus of research in this area has been on aspects of interaction, speech and discourse; it often centered on issues of scaffolding and interdependence. As a result, data primarily consists of traces of interaction in the form of discourse. Research about CoRL strives to examine ways in which social practices support individuals in appropriating knowledge and processes for self-regulation (Hadwin, et al., in press). Rather than modeling self-regulated strategy use, social support in the form of scaffolding tends to take on some of the self-regulatory processes or burdens rather than demonstrating how to engage those processes (Hadwin, et al., 2005). CoRL emphasizes the role of emergent interaction where models and students transition from co-regulating to self-regulating (McCaslin & Hickey, 2001).

Socially-shared regulation of learning (SSRL) refers to processes by which group members regulate their collective activity. Much like successful individuals regulate their motivational, cognitive, and metacognitive engagement, successful groups must learn to share in regulating group processes. From this perspective, goals and standards are co-constructed, and the desired product is socially shared cognition. Similar to CoRL, discourse and traces of interaction are primary sources of data in studying SSRL. However, unlike research about CoRL, research about SSRL tends to examine contributions, roles, the evolution of ideas, and ways groups collectively set goals, monitor, evaluate, and regulate their shared social space. Examining SSRL requires a shift toward new forms of instructional tools, data collection and data analysis that acknowledge individuals as social entities within shared tasks. Carroll, Neale, Isenhour, Rosson, and McCrickard (2003) have proposed tools specifically designed to support students in synchronizing task-oriented collaborative activity using notification and awareness tools. Overall, the field lacks empirically validated tools and methods for supporting: (a) co-construction of tasks and collaborative goals, (b) collective monitoring and regulating toward collective task completion. Future research should examine ways collaborators monitor and regulate activities in these shared spaces to avoid them derailing or being supported/dominated by only one group member.

Educational significance

To date very few empirical studies have considered social aspects of self-regulation in the context of complex tasks that unfold over time. Over the past 20 years, Hadwin and her colleagues have made significant progress in developing theoretical platforms, tools, and analysis methods to further research that investigates the intersect of self-regulation, co-regulation and socially-shared regulation in solo and collaborative tasks. The paper concludes with a discussion about the ways technologies can be used to advance research agendas regarding the socially shared aspects of self-regulation.

References

Currently, the conceptual and empirical focus of research on learning has been how learners build a shared understanding of a task or a learning environment in various social learning models. “Shared” is a topical term in cognitive oriented discussion of social learning, but what exactly “sharing” is, in terms of its direct effects is not yet clear. Are students’ able to activate self-regulated learning strategies in these contexts? And how is motivation regulation activated in groups? This paper introduces our methods for studying social aspects of SRL. In a series of empirical studies we have taken a situated approach to consider social aspects of SRL on an individual and group level. We will provide different illustrations to show how regulation processes are studied in as a self and shared phenomenon in collaborative learning settings through adaptive self-report instrument, video-data and interviews.

Introduction

Currently, the conceptual and empirical focus of research on learning has been how learners build a shared understanding of a task or a learning environment in various social learning models (Resnick, Levine, & Teasley, 1991). “Shared” is a topical term in cognitive oriented discussion of social learning (e.g. Schwartz, 1995), but what exactly “sharing” is, in terms of its direct effects is not yet clear. Are students’ able to activate self-regulated learning strategies in these contexts? And how is motivation regulation activated in groups? The problem identified in our current research program is that mental states of motivation, as traditionally conceptualized, represent subjectively perceived snapshots of the process. As such they give the theoretical baseline for the phenomena, but consider motivation and SRL as a static individual phenomenon, not process-oriented and adaptive. However, today, the pressure for active motivation regulation in shared learning situations is increasing, because of complex interactions in changing learning contexts (Järvelä, Volet & Järvenoja, 2008). In our series of empirical studies we have been keen to overcome above mentioned limitations to explain precisely the role of social and contextual influences on all phases of self-regulated learning (SRL), as well as exploring the distinctions between self and socially shared regulation.

Aim

This paper introduces our methods for studying social aspects of SRL. In a series of empirical studies we have taken a situated approach to consider social aspects of SRL on an individual and group level. We will provide different illustrations to show how regulation processes are studied in as a self and social phenomenon in collaborative learning settings through adaptive self-report instrument, video-data and interviews.

Methods

In our study we investigated what kind of socio-emotional challenges first-year teacher education students (n=63) experience during collaborative learning and whether the students regulate the emotions evoked during these situations. The interplay of the emotion regulation processes between the individual and the group was also studied (Järvenoja & Järvelä, 2008). The students studied in groups of 3-5 members and participated in three different collaborative learning tasks. We used   ‘Adaptive Instrument for Regulation of Emotions’ (AIRE), which aims to provide a measure of the adaptive nature of emotion regulation in socially challenging learning situations (Järvenoja, Volet & Järvelä, 2008). It is designed to access students’ experience of individual and socially shared regulation of emotions in a specific group learning activity that they are currently engaged in. The results from AIRE helped us to identify the variety of social challenges the students experienced during collaborative learning. Furthermore, the results revealed that students’ interpretation of regulation processes during collaborative learning include both socially shared and individual forms of regulation. The AIRE gave us more adaptive and situation-specific instrument to capture the significance of adaptive individual and social context dimensions in the regulation process (cf. Nolen & Ward, 2008) relative to conventionally used self-reports and inventories, which focus on individuals’ capacity to use regulation strategies and dispositions towards completing academic tasks.

In another study (Näykki, Järvenoja, & Järvelä, 2008) we aimed to explore what strategic actions the higher education students activate during their collaborative learning situations and how the different collaborative learning situations (defined as successful and less-successful) differ in respect to strategic actions students use. The participants (n = 22) worked in groups of 4 to 5 students for a period of 12 weeks. In addition to exploring the amount, spectrum and quality of strategy use on group level, content analysis of video data offered us information of how various contextual and situational factors can contribute to different learning outcomes. The findings of this study indicate that successful and less-successful collaborative learning situations differ in respect to the strategic actions students use to regulate their collaborative learning.

Another method we used to analyse the video recordings of students’ group work was to analyse the group interactions (“What a group is doing”). When the aim is to study how regulation processes are actually co-constructed together within a group according to our experience video data is often one of the best, if not only, way to detect the shared/joint processes. The illustration that we will present provides process data on the students’ actual social interactions. It derives from 4-person group whose collaboration was videotaped. The focus of the data analysis was on meaningful micro-level episodes that have been identified on a former phase of the analysis. The unit of the analysis was on the interactions between the group members and on the group as a social entity. This type of video data analysis had the potential to illuminate how motivation and emotion regulation processes are generated together in group members’ interaction in order to restore socio-emotional balance and maintain motivation within the group.

Conclusions and educational significance

We will conclude with the discussion how combining data at the individual and group level can provide a more comprehensive and unique insight into the dynamics of the group. We argue that it makes it possible to examine the extent of congruence in perceptions of challenges in relation to the same event, and to interpret each individual’s data in relation to that of the others. Finally, it enables the study of the group’s collaboration from self- and socially shared regulation perspectives.

References

Järvenoja, H., & Järvelä, S. (2005). How students describe the sources of their emotional and motivational experiences during the learning process: A qualitative approach. Learning and Instruction, 15(5), 465-480.

Järvenoja, H., Volet, S., & Järvelä, S. (2008). Investigating self-, other- and shared-regulation in socially challenging learning situations: An instrument to assess dynamics of students’ emotion and motivation regulation processes. Submitted.

In this study, a process-oriented approach was used to examine how groups of pre-service teachers regulate their mathematical problem solving supported by a text-based and asynchronous learning environment. To analyse the discussion forum data from individual perspective, a qualitative content analysis was performed from the mathematical problem solving and the regulation of cognition points of view. First, it was characterized whether a computer note signifies mathematical problem solving or is it social in nature. Secondly, cognitive and metacognitive, and especially metacognitive regulation computer notes were distinguished. Thirdly, to move beyond the individual level and focus on group level characteristics, an operational definition of socially shared metacognition was used to establish how the groups jointly regulate their joint problem solving process. Finally, to examine the interplay of cognitive, the metacognitive and social dimensions of group’s problem solving process, a process-oriented graph as a function of time was drawn. The results of this study showed that the groups regulated their problem solving problem differently. There were groups in which metacognition was made visible but not shared, or there was not any sign of regulation of cognitive processes in group’s problem solving. There were also groups where the participants jointly regulated their problem solving making socially shared metacognition visible. Although, there was not much regulation of cognitive process made visible, the results of this study suggest that socially shared metacognition is a differentiator making successful problem solving possible at group level.

Introduction

In the field of educational psychology, there is a growing interest to understand how learners regulate their own and their joint problem solving in social learning situations (Järvenoja & Järvelä, 2008; Hadwin, Oshige, Gress & Winne, 2008; Volet, Summers & Thurman, 2008). The current research aims to understand shared processes of learning which cannot be reduced to the attribute of an individual (Järvelä, Järvenoja & Veermans, 2008; Thompson & Fine, 1999) in a particular learning context. Drawing on social constructivist and situative perspectives of learning (Brown, Collins & Duguid, 1989; Anderson, Reder & Simon, 1996; Vygotsky, 1978; Palinscar, 1998; Alexander, 2007), we move over from an individual mathematical problem solving to social learning situation, especially from metacognition point of view. Generally, the concept of metacognition refers to thinking about thinking or knowledge of cognition and regulation of one’s cognitive activities (Flavell, 1979; Brown, 1987), and research on metacognition explains how an individual problem solver regulates and controls his or her cognitive processes (Schraw & Dennison, 1994; Schoenfeld, 1985; Garofalo & Lester, 1985). To explore metacognition in social learning situation, a conditional definition of socially shared metacognition was developed based on the current understanding of shared processes of learning (e.g. Hadwin, Oshige, Gress & Winne, 2008; Järvenoja & Järvelä, 2008). According to our understanding socially shared metacognitionrefers to metacognitive regulation in a group’s cognitive activities where shared and jointly constructed knowledge are used to interrupt, change or promote the collaborative problem solving process.

 Aim

 Method

In our study, we examined how six triads of pre-service teachers regulate their collaborative mathematical problem solving supported by a text-based and asynchronous learning environment. For this explorative study, we chose a total of 16 various mathematical problems of different difficulty levels but within the limits of the distribution of skills. In this presentation, we discuss three of these problems where pre-service teachers have conceptual difficulties in mathematics when understanding of fractions and operations with rational numbers is required (Merenluoto & Lehtinen, 2004). The data of this study consist of discussion forum data and the written computer notes are considered as individual thinking made visible. In the analysis, no generalizations were made into thinking behind of messages.

In the qualitative content analysis (Chi, 1997) of discussion forum data, it was first characterized whether the computer note was cognitive, metacognitive or social in nature. Metacognitive computer notes included an explicit explanation to control (Nelson, 1996) and regulate (Schraw & Moshman 1995; Kluwe 1987; Brown 1987) as to why, for example, the current way of thinking was or was not working. Secondly, to focus on group-level characteristics, an operational definition for socially shared metacognition was proposed on the basis of the current understanding of shared processes of learning (e.g. Hadwin, Oshige, Gress & Winne, 2008; Järvenoja & Järvelä, 2008). For socially shared metacognition to emerge in the discussion forum, it was required that: (a) an individual’s metacognitive regulation message was contributed to the discussion forum in order to interrupt, change or promote the group’s problem solving; (b) the peers acknowledged the other group members’ thinking; and (c) they were building on the other group members’ thinking to develop the idea further. Finally, based on the previous analysis, a process-oriented graph of group’s problem solving as a function of time was drawn. The group’s problem solving graph made it possible to visualize how interactive the group was, how much time they used time for the task, what the cognitive level of the group’s performance was, and in what kind of situations metacognition became visible and/or shared in joint problem solving.

The results of this study show that in groups’ problem solving 1) metacognition becomes visible and shared; or 2) metacognition becomes visible but not shared; or 3) only individual regulation becomes visible. Although regulation was not much joint in groups’ problem solving, socially shared metacognition can be considered as a differentiator making problem solving successful at group level.

Conclusions and educational significance

To conclude, we propose that a tight operational definition is needed to examine shared metacognition and regulation of learning. We argue that focusing on the group level characteristics and utilizing the process-oriented graph of groups’ problem solving provides a strategy for examining the interplay between social, cognitive and metacognitive dimensions in joint mathematical problem solving and how joint processes are regulated. This knowledge is needed to design and scaffold such collaborative learning situations in technology-based learning environments where participants could engage in successful problem solving as groups forming a community for learners. The pre-service teachers participated this study, could be encouraged to use collaborative learning situations with and without technology in their future mathematics classroom with children.

 References

Hadwin, A. F., Oshige, M., Gress, C.L.Z. & Winne, P.H. (2008). Innocvative ways for using gStudy to orchestrate and research social aspects of self-regulated learning. Computers in Human Behavior, in press.

Järvelä, S. Järvenoja, H. & Veermans, M. (2008). Understanding the dynamics of motivation in socially shared learning. International Journal of Educational Research, 47 (1), 122-135.

Järvelä, S., Veermans, M & Leinonen, P. (2008). Investigating student engagement in computer-supported inquiry: a process-oriented analysis. Social Psychology of Education 11, 299–322.

Järvenoja, H., & Järvelä, S. (2008). Emotion control in collaborative learning situations - Do students regulate emotions evoked from social challenges? Submitted for publication.

Merenluoto, K. & Lehtinen, E. (2004). Number concept and conceptual change: Outlines for new teaching strategies. Learning and Instruction, 14(5), 519-534.

Numerous studies have shown how self-efficacy for specific tasks and longer-term self-regulated learning (SRL) can be improved. Most studies involve modeling, the key variable in social cognitive theory, and employ the four-phase social cognitive development model: observation, emulation, self-control, and self-regulation. This presentation will explore the idea that self-efficacy and SRL might also be improved by educational practices that essentially ignore them. Emerging situative theories of cognition suggest that practices that focus directly on communal participation in discourse can indirectly increase self-efficacy towards those practices and enhance SRL in the longer term. Compared to social cognitive interventions, it appears possible that “participatory” interventions might (a) have a greater impact on self-efficacy and SRL, (b) have greater impact on behavioral outcomes such as time on task, and (c) overcome the validity problem of using the same constructs in the intervention and the outcome measures. These implications are being explored using a comprehensive framework consisting of multiple levels of increasingly formal discourse guidelines. Event guidelines shape informal observation of the enactment of the activity and feedback (“communal regulation”); activity guidelines shape semi-formal observation of reflection and feedback on those enactments (“collaborative regulation”).   Subsequently, assessment guidelines shape the assessment of individual student’s actual artifacts and self-efficacy for completing such tasks (“self-regulation”).   Finally, measurement guidelines shape the eventual measurement of a curriculum’s impact on aggregated achievement and SRL (self-regulated learning). Two ongoing examples of the application of this framework are presented. The first involves educational videogames and the enduring debate over extrinsic incentives. The second involves using new media literacy practice in an alternative high school that serves students who have not exhibited self-efficacy or developed SRL in conventional settings.

Researchers have long been concerned with improving self-efficacy (the perceived capabilities for learning and performing in specific contexts at designated levels, e.g., Schunk & Rice, 1989) and self-regulated learning (SRL, self-generated thoughts, feelings, or actions that affect learning, e.g., Zimmerman & Bandura, 1994).   Reflecting social cognitive theory (Bandura, 1989), most efforts assume that SRL begins with social sources (via models of self-efficacy) and gradually shift to self-sources. Most efforts follow the four hypothesized phases of social cognitive development: observation, emulation, self-control, and self-regulation. Experimental manipulations of models and feedback consistent with social cognitive theory have consistently resulted in greater self-efficacy and enhanced self-regulation; these have then been shown to improve academic outcomes such as reading (Schunk & Rice, 1989) and writing (Schunk & Swartz, 1993; Zimmerman & Kitsantas, 1999). This research has been extremely influential. Most future educators are taught to teach self-regulation strategies, use adult and peer modeling, provide feedback, etc. 

This presentation will explore different ways of thinking about self-efficacy and self-regulated learning, and different strategies for improving them. This approach draws from situative (Greeno, et al., 1998; Gee, 2004) and participatory (e.g., Lave & Wenger, 1991) views of cognition and learning. Consistent with the EARLI 2009 Community of Learners theme, these theories assume that knowledge is distributed across tools, technologies, and social rituals that human cultures construct to work together. In this view, classroom learning is communal engagement in the knowledge represented in the enactment of social rituals and collective classroom practices. Applied to motivation, this view implies that the goals and values that motivate engagement primarily reside in the social context, alongside the relevant socially defined knowledge (Hickey & Granade, 2004). This means, for example, that if a community of learners does not value participation in those rituals and practices, it is practically impossible for a given individual to participate (Hickey, 2003). 

Turner & Patrick (2008, p. 121) interpret this stance as asserting incompatibility with social cognitive tenets; Walker (2007) interprets this stance to be “based upon a misunderstanding of the transformative nature of internalization.” Neither interpretation appears warranted. The most important attribute of situative views is that they treat individual activity as “secondary” phenomena. This distinguishes this situative approach from Vygotskian sociocultural approaches (e.g., McCaslin & Good, 1996; Nolen, 2007). While acknowledging the social origins and internalization of self-efficacy and self-regulation, they remain “special cases” of socially situated activity (along with all other individual activity). Indeed, the framework below asserts that individual engagement and learning should be assessed and used for feedback and improvement. But this must be done carefully and in isolation, to avoid undermining the more formatively useful communal activity. This is accomplished using design-based methods (e.g., Barab, 2007) and a naturalistic stance; this distinguishes this approach from ostensibly similar studies using mixed methods and a more naturalistic stance (e.g., Järvenoja,, & Järvelä, 2005; Nolen, 2007; Turner & Patrick, 2008). 

Thus, the relevant debate is whether it is useful to treat individual activity (self-efficacy, SRL, behavior, understanding, etc.) as special cases of socially situated activity, but not vice versa. Assuming so, several important implications for practice follow. One is that directly targeting communal participation can consistently impact, self-efficacy, SRL, and individual behavior more than directly targeting self-efficacy & SRL, or individual behavior (Hickey, 2003). Another implication is that directly targeting communal participation can improve self-efficacy & SRL, but that directly targeting self-efficacy or SRL will have little impact on communal participation (Hickey & Schafer, 2005). 

These implications are embedded in a participatory multi-level framework. The basic idea is that (1) communal participation and co-regulated learning should be interpreted and enhanced before (2) individual self-efficacy and understanding are assessed and enhanced, and before (3) aggregated self-regulated learning and academic achievement are measured. This framework consists of discourse guidelines that shape participation around these increasingly formal characterizations of engagement and learning: 

 Event guidelines frame the enactment of specific curricular activities and outline the domain formalisms that students should enlist in the discourse defining those enactments. Features of the teacher and student materials guide informal observation and guidance of this discourse. This feedback is provided on an immediate timescale, as the discourse is occurring. This activity is communally-regulated.

 Activity guidelines frame reflection on the enactment of the activities around the students “working” artifacts (but not the artifacts themselves). A simple rubric guides the teacher’s semi-formal observation and guidance of that discourse, ensuring that all students have participated in discourse around the intended formalisms. To avoid considering individual activity, the discourse is directed towards the intended curricular rituals. This is intended to be consistent with McCaslin & Good’s (1996) co-regulated learning.

 Assessment guidelines frame the teacher’s formal assessment (e.g., grading) of the artifacts (essays, assignments, expressions) that students produce as evidence of individual understanding, using conventional scoring rubrics. This is the point at which it becomes appropriate to consider individual students’ self-efficacy for completing such tasks (self-regulation). This includes informal consideration by teachers and more formal assessment of self-efficacy or other task-specific or state-specific motivational variables by researchers.

 Measurement guidelines frame the eventual measurement of longer term consequences of the curriculum, across broader populations of students, potentially in cross-curricular comparisons. This includes aggregated achievement using external measures, and aggregated self-regulated learning measured as a stable trait of individuals.

 The framework is intended to be used in increasingly formal design cycles, starting with the first two levels, leading to formal evaluation using the measurement guidelines (Hickey, et al., 2006). It is being refined in two ongoing programs of research funded by the MacArthur Foundation’s Digital Media and Learning initiative. The first is refining assessment, feedback, and incentives in the Quest Atlantis educational videogame (Barab et al., 2007). The second is refining a “21st Century” English curriculum developed by Project New Media Literacy (Jenkins, et al., 2006) in an alternative school serving students who have not exhibited self-efficacy or developed SRL in traditional settings. Evidence of communal, collaborative, and individual engagement and learning and design principles from from two design-based refinement cycles in each effort will be presented.