=Paper=
{{Paper
|id=Vol-3029/paper07
|storemode=property
|title=Towards socially shared regulation within CSCL scripts: mirroring group participation in PyramidApp
|pdfUrl=https://ceur-ws.org/Vol-3029/paper07.pdf
|volume=Vol-3029
|authors=Emily Theophilou,Ishari Amarasinghe,Davinia Hernández-Leo,René Alejandro Lobo,Francisco Crespi
|dblpUrl=https://dblp.org/rec/conf/lasi-spain/TheophilouAHQC21
}}
==Towards socially shared regulation within CSCL scripts: mirroring group participation in PyramidApp==
https://ceur-ws.org/Vol-3029/paper07.pdf
Towards socially shared regulation within CSCL scripts:
mirroring group participation in PyramidApp
Emily Theophilou [0000-0001-8290-9944], Ishari Amarasinghe [0000-0003-2960-4804], Davinia Her-
nández-Leo [0000-0003-0548-7455], René Lobo [0000-0003-2989-5357], Francisco Crespi [0000-0002-
6791-3299]
TIDE, ICT Department, Universitat Pompeu Fabra, Spain
[emily.theophilou, ishari.amarasinghe, davinia.hernandez-
leo, rene.lobo]@upf.edu,francisco.crespi01@estudiant.upf.edu
Abstract. Computer-Supported Collaborative Learning (CSCL) has been shown
to enhance learning by promoting peer interactions. With the support of collabo-
rative learning flow patterns (CLFP), the collaboration process within the so-
called CSCL macro-scripts can be further structured and constrained. However,
social dynamics within tasks, even within scripts, have the potential to hinder the
collaborative element as students might not participate sufficiently in conversa-
tions. To determine if participation in discussions between students was balanced
within a CSCL script we analysed data from an implementation of the Pyramid
CLFP. The data showed that student participation varied across different groups,
with some groups and students not initiating conversation between them. So-
cially shared regulation is a social form of regulation where students regulate
their learning as a group rather than as individuals through the means of discus-
sion, negotiation, and perspective taking. As research has shown that for collab-
orative learning to be successful students need to engage in fruitful discussions
and support each other to regulate their learning, we propose the implementation
of a social awareness feature mirroring group participation through learning an-
alytics in a CLFP tool (PyramidApp) to promote socially shared regulated learn-
ing.
Keywords: Socially shared regulated learning, CSCL, Pyramid CLFP, Social
awareness, Learning analytics
1 Introduction
This study focuses on motivating the need and presenting a design of an application of
learning analytics to enhance social interactions in Computer Supported Collaborative
Learning (CSCL). In particular, it proposes the design of a social awareness feature
mirroring group participation to support student regulation in the context of scripted
Pyramid-based collaborative learning flows.
1.1 CSCL and Collaborative Learning Flow Patterns (CLFPs)
Collaborative learning is an effective educational approach that has been shown to
enhance students' learning by allowing them to externalize their knowledge, monitor
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
74 Learning Analytics in times of COVID-19: Opportunity from crisis
their peers' learning and negotiate their opinions and ideas. However, when learners are
asked to collaborate without any instructions the results are not as effective as with
scaffolded collaboration [1]. In CSCL, there is significant research supporting the ef-
fectiveness of collaboration scripts that structure the flow of collaborative learning ac-
tions with the help of technology. CSCL scripts scaffold learners towards collaboration
by promoting specific activities in a sequence and implicitly or explicitly distributing
roles amongst the learners which effectively guide and improve collaboration processes
and subsequent individual learning [2]. Collaborative learning flow patterns (CLFP)
provide specific structures to CSCL scripts [3]. One such CLFP is the one of Pyramid
where students are faced with a problem and through different levels discuss and im-
prove their initial answer [4].
Alternative or complementary approaches seeking effective collaborative learning
are related to supporting student’s regulation. Research has shown that collaborative
learning can be hindered by social dynamics as this type of learning situation poses
socio-emotional challenges and requires regulation of learning [5, 6]. During face-to-
face collaborative learning students are more likely to be aware of each other's socio-
emotional interactions as they can be conscious of each other's task perceptions or
goals. Collaborative learning supported by socially shared regulated learning (SSRL)
has shown to further assist collaboration by encouraging group members to support
each other to regulate their learning and by promoting self-regulation [7].
1.2 Co-regulation and socially shared regulation in collaborative learning
During individual learning, students internally regulate their thoughts and behav-
iours to help them reach a goal. The theory of regulated learning addresses this process
and concerns how learners develop and effectively use learning skills [8]. An important
contribution of this theory is the method of self-regulated learning. Self-regulated learn-
ing (SRL) refers to an iterative process where learners through self-generated thoughts,
feelings and behaviours take control of their learning by planning, monitoring, and
evaluating their learning to attain a goal [9]. Regulated learning can also be applied in
collaborative learning settings through the social modes of co-regulation and socially
shared regulated learning (SSRL).
Co-regulated learning describes a relationship between individuals where one person
is more knowledgeable or more skilled. Co-regulated learning is a dynamic metacog-
nitive process where individuals self-regulate their learning and share regulation of
their cognitions, emotions, behaviours, and motivations towards the aim of an academic
goal [10]. SSRL occurs when groups perform interdependently on a set task and nego-
tiate perceptions, goals, and strategies as a group rather than as individuals [7,11]. Such
a process assumes that a positive socio-emotional interaction will be maintained by
promoting listening and perspective taking in individuals [12]. Group awareness and
group mirroring features in CSCL solutions can support regulation in collaborative
learning.
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
Learning Analytics in times of COVID-19: Opportunity from crisis 75
1.3 Group awareness and group mirrors in CSCL
Group awareness was defined by Dourish and Bellotti (1992) [13] as the “understand-
ing of the activities of others, which provides a context for your own activity”. Tools
that take into account the group awareness, can provide the participants valuable infor-
mation. Particularly, CSCL-related group awareness tools offer such kind of additional
value, as they often focus on variables that are not directly available in face-to-face
settings, such as assessing and providing information about how a learning group eval-
uates the group members’ contributions [14]. However in group awareness tools the
information has to be salient enough to capture the learners‘attention. But at the same
time, awareness information needs to be easy to understand and interpret.
There are three types of group awareness: Behavioural awareness refers to the in-
formation about other members’ activities in the group, such as what they are doing
and what they will do later. Cognitive awareness refers to cognitive information, in-
cluding knowledge, beliefs or goals of the members in the group. Social awareness
refers to being aware of what other members of the group are doing or gathering con-
tinuous information about them and acting accordingly [15].
One example of social awareness in CSCL is the Reflect table proposed by Bachour
et al (2008) [16]. The table provides a display on their surface showing participants
their interaction level (contributions to a conversation) with the others. However, the
visualization on the table did not suggest changes in participant behaviour. It simply
showed the participants the state of their conversation as a group mirror, and it was up
to the participants to decide if a change was needed.
Section 2 describes the need of supporting student regulation in the case of CSCL
scripts structured according to the Pyramid CLFP. Then, section 3 presents a design of
a social awareness feature aiming at facilitating socially shared regulation in Pyramid
App, an application enabling the design and enactment of Pyramid scripts [4].
2 Need for socially shared regulation in Pyramid collaborative
learning flows
PyramidApp is a web-based tool that facilitates the deployment of Pyramid pattern
based CSCL activities where students can collaborate during several stages. Initially
students are required to provide an individual answer to a given task. Then they are
allocated into small groups where they can see the answers that were provided by their
group members and rate them. In the next stage, the “answer improving” stage, students
can discuss between them the given task and provide an improved answer. At the end
of the small group level, students enter a second individual rating stage. Finally, the
smaller groups are merged to formulate larger groups where they can further engage in
discussions and vote for the best answers to reach a consensus at the end of the activity.
Log data from the PyramidApp tool was analysed to determine the discussion par-
ticipation of students during Pyramid activities. In the following we first present stu-
dents’ overall participation in the discussion in three Pyramid activities. Then in order
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
76 Learning Analytics in times of COVID-19: Opportunity from crisis
to better understand the discussion participation differences of each student in each
Pyramid activity we provide a detailed analysis.
2.1 Overall Discussion Participation
The first column in Table 1 shows the activity ID. We considered three Pyramid
activities conducted by a teacher online in a postgraduate course named “Learning
Technologies”. During activity A, B, C teacher proposed open-ended knowledge shar-
ing tasks for students. The third column represents the Pyramid group level, the fourth
column provides an identifier for different groups and the fifth column shows the total
number of students in each group. In the final column students’ chat participation is
given which was calculated based on the percentage of students who wrote chat mes-
sages out of total students in a given group.
Table 1. Chat Participation of Students.
Activity Topic / type of task Group Group Group Chat Participation
ID Level ID size %
A Brainstorming ideas (for a Small GA0 5 60%
particular type of com- GA1 5 80%
puter system design) based GA2 4 80%
on previous experiences Large GA3 14 57.14%
B Compare two types of ed- Small GB0 5 60%
ucational technologies GB1 6 0%
GB2 4 100%
Large GB3 15 66.67%
C Reflection with critical Small GC0 3 100%
analysis based on a read- GC1 5 66.67%
ing GC2 6 80%
Large GC3 14 14.29%
As presented in Table 1, it can be seen that students’ discussion participation varied
across different groups. Only GB2 and GC0 groups had 100% participation in the dis-
cussion meaning every student in the group participated in the discussion at least once.
In the other groups however, the discussion participation varied across participants. It
is noticeable that in GB1 and GC3 students had the least discussion participation.
2.2 Detailed Discussion Participation
The graphs under Fig. 1 present along the X-axis the number of chat messages posted
by each student during the first and second rating levels of the Pyramid activities and
the Y-axis represents students. As it can be seen in Fig.1(a) four students (S1-1 , S1-3,
S1-6 and S1-7) only contributed to the first rating level and student S1-13 only partici-
pated in the second rating level. Noticeably S1-12 and S1-14 did not participate in dis-
cussion at any Pyramid level. In Activity B (see Fig. 1(b)) there is unequal participation
in the discussion. For instance, it can be seen that seven students (S2-1, S2-2, S2-3, S2-
4, S2-5, S2-6 and S2-7) mainly contributed and dominated the discussion while others
have less participation or no participation at all in the discussion. In Activity C (see Fig.
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
Learning Analytics in times of COVID-19: Opportunity from crisis 77
1(c)) many students contributed to the discussion in the first rating level, whereas only
two of them contributed to the discussion in the second rating level.
Fig. 1. Chat participation distribution of students in (a) activity A, (b) activity B and (c)
activity C.
3 Design of a social awareness feature for socially shared
regulation in PyramidApp
The analysis of the PyramidApp log data has shown that student’s participation varied
across different groups, reaching both extremes of 100% and 0%. As empirical evi-
dence indicates that SSRL has the potential of increasing group performance in collab-
orative learning [17], we hypothesise that the implementation of a feature to support
SSRL within PyramidApp has the potential of enhancing discussion dynamics in the
PyramidApp. We therefore propose a social awareness feature to be implemented to
allow participants to become aware of their social contribution and support SSRL.
Similarly, to the approach of Reflect [16], which achieves social regulation by display-
ing the levels of participation on the surface of an interactive table, we propose a design
mirroring participation in the PyramidApp level of “answer improvement”. In particu-
lar, we use the metaphor of a participation bar that is visible to all the participants (Fig
1). The design has taken into consideration the space limitations and amount of infor-
mation within the GUI. The expectation is that this feature supports socially shared
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
78 Learning Analytics in times of COVID-19: Opportunity from crisis
regulation by improving participants' awareness of their own contributions in relation
to other group members’ contributions.
To calculate user participation, an algorithm counts the characters sent, edited, or de-
leted from the chat and the answer improvement text-editor. Then to compute the user’s
participation the following algorithm has been proposed:
(User participation / (All participants average participation * 2)) * 100
This algorithm implies the following cases: at the beginning, all user’s participation
will be 50% as the average will be 0, a user’s progress will be under 50% if their par-
ticipation is under the average, and the user's progress will be at 100% if their partici-
pation is equal or greater than twice the average. See the notes of figure 2 for a detailed
use case using the algorithm.
Fig. 2. The proposed social awareness feature can be seen under each user. Use case: User1
introduced 29 characters, User2 introduced 21 characters, User3 introduced 46 characters and
User4 introduced 0 characters. The average of the characters is 24. The participation bar for each
user has been calculated based on the above algorithm. User1 participation → (29 / (24 x 2)) x
100 = 60% , User2 participation → (21 / (24 x 2)) x 100 = 43%, User3 participation → (46 / (24
x 2)) x 100 = 95% and User4 participation → (0 / (24 x 2)) x 100 = 0%.
4 Conclusions and future work
This study aimed to investigate students' participation in Pyramid pattern based scripted
CSCL situations and explore regulated learning theories to help enhance participation
in collaborative activities. Log data from PyramidApp were analysed to determine the
distribution of students’ discussion participation within three classroom based CSCL
situations. Results of the analysis showed that participation levels varied across differ-
ent groups with a number of students not participating at all. The sample size that was
used to analyze the data was small, however, as CSCL can be hindered by social dy-
namics and socio-emotional factors, we found it important to further investigate mech-
anisms that can reduce this effect. Thus, this paper further proposed the implementation
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
Learning Analytics in times of COVID-19: Opportunity from crisis 79
of a social awareness feature within a pyramid CLFP to promote socially shared regu-
lation by allowing group members to monitor each other’s contributions. However, its
effectiveness has not yet been evaluated in practice and its future implementation could
have some limitations. For instance, the design of the social awareness feature could
potentially hinder its effectiveness as students might not become aware of its purpose.
Additionally, the amount of information that appears on the page could possibly lead
to an overload of their cognitive load, not allowing them to concentrate on the given
task. Follow up research needs to be conducted to evaluate the design of the feature and
its effectiveness in collaborative tasks. In contrast, students becoming aware of the
tool’s functionality may cause the conversation to become a competition of who will
send the most messages and derail the discussion from being productive. The current
design of the platform prevents such behaviors from happening through a teacher mon-
itoring system where teachers can intervene in the conversation. However, the feature’s
limitation could be addressed in future work by implementing an algorithm to analyze
the quality of contributions rather than word volume. Follow-up research will be con-
ducted to determine if the implementation of the proposed feature enhances discussion
participation within a pyramid CLFP. Further future work may also address how the
feature affects dominant speakers' participation level and propose amendments to the
design to help prevent dominant speakers from controlling and overpowering the con-
versation. The proposed design in this paper does not explicitly address this. Such a
design needs to be carefully thought to not adversely affect the participation levels of
dominant speakers.
5 Acknowledgements
We would like to acknowledge the help of Alexandre Argenté Pérez-Milá and Pascual
Pablo Abenia Polo for assisting on the technical development of this study. This work
has been partially funded by the Volkswagen Foundation in the framework of the pro-
ject COURAGE (no. 95567) and by FEDER, the National Research Agency of the
Spanish Ministry, TIN2017-85179-C3-3-R and MDM-2015-0502. D. Hernández-Leo
also acknowledges the support of ICREA under the ICREA Academia programme.
References
1. Fischer, F., Kollar, I., Haake, J. M., & Mandl, H.: Perspectives on Collaboration Scripts.
Scripting Computer-Supported Collaborative Learning Cognitive, Computational and Edu-
cational Perspectives, Springer, pp.1-9, (2006)
2. Radkowitsch, A., Vogel, F., & Fischer, F. (2020). Good for learning, bad for motivation? A
meta-analysis on the effects of computer-supported collaboration scripts. International
Journal of Computer-Supported Collaborative Learning.
3. Hernández Leo, D., Asensio-Pérez, J. I., & Dimitriadis, Y. (2005). Computational represen-
tation of collaborative learning flow patterns using IMS learning design. Journal of Educa-
tional Technology & Society 2005; 8 (4): 75-89.
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)
80 Learning Analytics in times of COVID-19: Opportunity from crisis
4. Manathunga, K., & Hernández‐Leo, D. (2018). Authoring and enactment of mobile pyra-
mid‐based collaborative learning activities. British Journal of Educational Technology,
49(2), 262-275.
5. Järvelä, Sanna, Simone Volet, and Hanna Järvenoja. 2010. “Research on Motivation in Col-
laborative Learning: Moving Beyond the Cognitive-Situative Divide and Combining Indi-
vidual and Social Processes.” EDUCATIONAL PSYCHOLOGIST 45(1): 15–27.
6. Volet, Simone, and Caroline Mansfield. Group Work at University: Significance of Personal
Goals in the Regulation Strategies of Students with Positive and Negative Appraisals (2007).
7. Järvelä, S., Järvenoja, H., Malmberg, J., & Hadwin, A. F. (2013). Exploring socially shared
regulation in the context of collaboration. Journal of Cognitive Education and Psychology,
12(3), 267-286.
8. Zimmerman, B. J., & Schunk, D. H. (2001). Reflections on theories of self-regulated learn-
ing and academic achievement. In B. J. Zimmerman & D. H. Schunk (Eds.), Self-regulated
learning and academic achieve- ment: Theoretical perspectives (2nd ed., pp. 289–307). Mah-
wah, NJ: Lawrence Erlbaum Associates.
9. Zimmerman, Barry J. 2002. “Becoming a Self-Regulated Learner : An Overview Becoming
a Self-Regulated Learner : An Overview.” Theory Into Practice 5841(2002): 64–70.
10. Hadwin, Allyson, Sanna Järvelä, and Mariel Miller. 2017. “Self-Regulation, Co-Regulation,
and Shared Regulation in Collaborative Learning Environments.” Handbook of Self-Regu-
lation of Learning and Performance: 83–106.
11. Hadwin, A., et al. (2011). Self-regulated, co-regulated, and socially shared regulation of
learning. Handbook of self-regulation of learning and performance, 30, 65-84.
12. Rogat, Toni Kempler, and Lisa Linnenbrink-Garcia. 2011. “Socially Shared Regulation in
Collaborative Groups: An Analysis of the Interplay between Quality of Social Regulation
and Group Processes.” Cognition and Instruction 29(4): 375–415.
13. Dourish, P., & Bellotti, V.: Awareness and coordination in shared workspaces. ACM con-
ference on computer-supported cooperative work (pp. 107–114) (1992).
14. Buder, J., & Bodemer, D. (2008). Supporting controversial CSCL discussions with aug-
mented group awareness tools. International Journal of Computer Supported Collaborative
Learning, 3, 123–139.
15. Ma, X., Liu, J., Liang, J., & Fan, C. (2020). An empirical study on the effect of group aware-
ness in CSCL environments. Interactive Learning Environments, 0(0), 1–16.
16. Bachour K, Kaplan F, Dillenbourg P (2008) Reflect: An interactive table for regulating face-
to-face collaborative learning. In: Lecture Notes in Computer Science (including subseries
Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). pp 39–48
17. Malmberg, J., Järvelä, S., Järvenoja, H., & Panadero, E. (2015). Promoting socially shared
regulation of learning in CSCL: Progress of socially shared regulation among high- and low-
performing groups. Computers and Human Behavior.
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0)