Guiding the Integration of Multimodal Learning Analytics in the

Glocal Classroom: A Case Study Applying MAMDA

Hamza Ouhaichi

, Daniel Spikol

and Bahtijar Vogel

Departement of Computer Science and Media Technology, Malmö University, Malmö, Sweden

Department of Science Education, Copenhagen University, Copenhagen, Denmark

Keywords: Glocal Classroom, Multimodal Learning Analytics, Smart Learning Environment.

Abstract: This study explores the integration of Multimodal Learning Analytics (MMLA) within the dynamic learning

ecosystem of the Glocal Classroom (GC). By employing the MMLA Model for Design and Analysis

(MAMDA), our research proposes a conceptual model leveraging the GC's existing infrastructure into an

MMLA system to enrich learning experiences and inform course design. Our methodology involves a case

study approach guided by the six phases of MAMDA. Building on previous studies, including a systematic

mapping of MMLA research and an investigation into MMLA system design. We seek to employ MMLA

insights to comprehensively understand the learning experience, identify issues, and guide improvement

strategies. Furthermore, we discuss potential challenges, mainly focusing on privacy and ethical

considerations. The result of this work aims to facilitate a responsible and effective implementation of MMLA

systems in educational settings.

1 INTRODUCTION

Recently, several innovative learning environments

have emerged at the intersection of technology and

education, such as the Glocal Classroom (GC). The

GC is a dynamic learning space that transcends

geographical boundaries, fostering collaboration

among students from diverse cultural backgrounds

(Messina et al., 2014). In line with the development

of educational technology, there is a growing

recognition of the potential benefits that Multimodal

Learning Analytics (MMLA) systems can bring to the

design of next-generation teaching and learning

environments. MMLA leverages diverse data

modalities, including text, audio, and visual inputs, to

gain deeper insights into the learning process

(Blikstein et al., 2013).

This study presents a conceptual integration of

MMLA in the learning environment GC. The

motivation behind implementing MMLA in the GC is

to enhance the educational experience and optimize

course design. By leveraging the multimodal data

generated within the GC, MMLA promises to provide

https://orcid.org/0000-0002-9278-8063

https://orcid.org/0000-0001-9454-0793

https://orcid.org/0000-0001-6708-5983

a holistic understanding of student engagement,

learning patterns, and the effectiveness of

pedagogical approaches (Cukurova et al., 2020). This

integration aligns with the broader goal of learning

analytics for advancing educational practices in line

with technological advancements (Ahad et al., 2018).

Applying a case study approach, we could examine

the GC in depth and develop a conceptual integration

of MMLA. This methodology enables a better

understanding of the interactions between technology

and Glocality principles in education (Patel & Lynch,

2013). We employ the MMLA Model for Design and

Analysis (MAMDA) to guide the study and

integration of MMLA in the GC. We aim to test and

reflect upon this newly developed model in a real-life

scenario. MAMDA serves as a framework,

addressing the key considerations for responsible data

handling, and the overall development of MMLA

systems. It offers a structured approach to ensure that

MMLA integration aligns with ethical standards and

educational objectives .

The remainder of this paper is structured as

follows: Section 2 sets the context and introduces the

478

Ouhaichi, H., Spikol, D. and Vogel, B.

Guiding the Integration of Multimodal Learning Analytics in the Glocal Classroom: A Case Study Applying MAMDA.

DOI: 10.5220/0012690900003693

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 1, pages 478-485

ISBN: 978-989-758-697-2; ISSN: 2184-5026

GC and MMLA. Section 3 describes the chosen

methodology. In section 4, the MAMDA is explained

in detail. Following that, Section 5 provides an

illustration of the integration of MMLA within the

GC setting. Section 6 engages in a discussion, delving

into the implications and insights derived from the

study. Section 7 reflects on the lessons learned,

followed by Section 8 outlines the identified

limitations. The paper concludes in Section 9,

summarizing key findings and proposing directions

for future research.

2 BACKGROUND

2.1

Glocal Classroom and MMLA

The GC is an innovative and dynamic educational

environment designed to bridge geographical gaps

and foster collaboration among students from diverse

cultural backgrounds. Equipped with communication

technology, the GC serves as a hybrid learning space

(Patel & Lynch, 2013). It features microphones,

cameras, projectors, and screens that facilitate

participation and interaction both on campus and

remotely. In our study, the GC is adopted by multiple

universities across four countries, exemplifying its

global reach and impact (Christensen et al., 2022).

The hybrid learning approach of the GC allows

students to engage in courses conducted at partner

universities, providing an educational experience that

combines the benefits of in-person and remote

learning. Our study explores the intersection of

MMLA technology and education to enhance the

learning experience within the GC. MMLA

represents an analytical data-driven approach that

harnesses diverse data modalities, including text,

audio, and visual inputs, to gain insights into the

learning process (Worsley, 2018). As a concept,

MMLA has gained prominence for its potential to

understand, monitor, and improve educational

practices. MMLA provides a holistic understanding

of student engagement, learning patterns, and the

effectiveness of pedagogical approaches (Blikstein et

al., 2013). The integration of MMLA into educational

settings has gained traction, driven by its potential to

understand, monitor, and improve educational

practices. As educational institutions seek to harness

the benefits of MMLA, there is a growing recognition

of the need for structured frameworks to guide its

effective implementation. In terms of understanding,

modeling, and supporting learning, MMLA has

shown promising results. Multiple research works

envision MMLA as a tool for supporting the

educational experience and refining learning designs

(Cukurova et al., 2020), which motivates us to

incorporate MMLA into the GC. With this

integration, comprehensive analytics become

possible, allowing a variety of perspectives on

educational experiences across various scales of time

and space. Considering the GC's existing

communication technology infrastructure, the

foundations are already in place for integrating an

MMLA system seamlessly.

2.2 MMLA Model for Design and

Analysis

The MAMDA model provides a framework and set

of key considerations to guide the development of

MMLA systems. Our primary objective is to assess

and reflect upon the viability and efficacy of the

newly developed model in a real-world case scenario.

The provided set of considerations for MMLA design

are derived from a previous study (Ouhaichi et al.,

2023). As a practical tool, the model will facilitate the

incorporation of MMLA into the learning

environment at GC. The primary goal of this study is

to test and reflect upon the MAMDA model within

the dynamic and innovative learning ecosystem of the

GC. Understanding how MMLA can be effectively

implemented in diverse settings becomes paramount

as educational landscapes evolve. With its global

reach and collaborative learning model, the GC

serves as an ideal case for this exploration. By

adopting a case study approach, we aim to assess the

applicability of the MAMDA model in guiding the

integration of MMLA within the GC. This study

seeks to contribute valuable insights into the potential

enhancements that MMLA can bring to collaborative

learning scenarios and the MAMDA model's

adaptability in shaping educational technology's

future.

3 METHODOLOGY

The methodology employed in this research adopts a

case study approach to explore the GC and investigate

the integration of MMLA. This section justifies using

a case study approach, outlines its characteristics, and

highlights its suitability for investigating the

relationship between analytics technology and

Glocality principles in education. We chose the case

study approach for its ability to provide an in-depth

exploration of the GC, offering a better understanding

of its dynamics and functionalities. Following

established guidelines for conducting and reporting

Guiding the Integration of Multimodal Learning Analytics in the Glocal Classroom: A Case Study Applying MAMDA

479

case study research in software engineering, proposed

by (Runeson et al., 2012) and in alignment with

practical recommendations from (Hancock et al.,

2021) in "Doing case study research: A practical

guide for beginning researchers" the case study

design is considered appropriate for our research

objectives. This approach enables a flexible

examination of the GC, considering various

dimensions such as its technological infrastructure,

user interactions, and the overarching Glocality

principles that shape its educational goals.

The data collection methods employed in this

study encompass a mixed approach to gain insights

into the GC. Observations constitute an initial

element involving a first-hand examination of the

learning scenarios facilitated within the GC. This

includes attending lectures held in the GC, providing

an immersive experience to understand the dynamics

of student engagement, interaction, and the overall

educational environment. In addition, a

documentation review is conducted, focusing on

digital platforms showcasing the usage of the GC.

This digital platform also functions as a booking

system, offering insights into the scheduling and

utilization patterns of the GC. We used the MAMDA

model to guide the case study and inform the data

collection process. The motivation behind employing

the model is to examine and reflect upon the practical

applicability and effectiveness of the newly

developed model in an authentic educational context.

This approach serves as a valuable means to refine the

proposed model, particularly within the complex and

dynamic setting of the GC. MAMDA, designed

through a qualitative study involving interviews with

researchers and experts in the field of MMLA, is

applied as a structured framework. The model

addresses key considerations across various phases,

providing an initial approach to MMLA system

design. Each phase of MAMDA corresponds to

specific considerations, including learning scenarios,

human factors, research orientation, data collection,

data management, privacy, and ethics. The

application of MAMDA serves as a guidepost,

aligning the data collection efforts with the

overarching objectives of designing an MMLA

system tailored to the context of the GC.

While the GC is present in four different countries

across four continents (Sweden, Canada, South

Africa, and Australia), the observations and data

collection activities are specifically centered on the

Swedish instance of the GC. This focused

examination allows for a specific exploration of the

GC's characteristics, technological capabilities, and

the embodiment of Glocality principles in education.

Figure 1: MMLA Model for Design and Analysis.

4 MMLA INTEGRATION IN GC

Aligned with the foundational principles of the

Community of Practice (CoP), derived from John

Dewey's notions and defined by Lave and Wenger, it

is applied to the GC setting (Christensen et al., 2020).

the framework leverages CoP's emphasis on shared

interests, open dialogue, and collaborative learning

within a common domain. In the GC setting, the CoP

perspective becomes a catalyst for transformative

educational experiences, evolving the GC into a

community where global interactions, shared

learning, and collaboration form the essence of the

learning process. These elements gain heightened

significance in an era where interconnectedness and

cross-cultural understanding are imperative for

education. The synergy between MMLA and the GC's

goals is a cornerstone of our framework. MMLA's

focus on leveraging multimodal data aligns

seamlessly with the GC's objectives, showcasing its

adaptability of MMLA and potential to enhance or

assess education practices in the GC. This alignment

underscores the reciprocal relationship between

MMLA and the GC's educational philosophy.

MMLA benefits from the GC's infrastructure, while

the GC gains from enhanced analytical capabilities

CSEDU 2024 - 16th International Conference on Computer Supported Education

480

offered by MMLA. This reciprocal enrichment forms

the envisioned glocal learning environment, where

global insights contribute to local educational

excellence and vice versa.

4.1 Preparations and Needs Analysis

We addressed the first three considerations in the

initial phase: Learning Scenarios, Human Factors,

and Research Orientation. Our aim involved the

identification of educational objectives aligned with

curriculum requirements and the technical

infrastructure. CoP emerged as a fundamental

educational framework within the GC. Consisting of

students and educators, the CoP shaped the

collaborative and community-driven educational

practices observed in the GC. The examination also

revealed the elements of the basic technological

infrastructure essential for communication and

collaboration. This included audio and video

components and digital platforms that facilitate

interactions. The main actors in the GC, namely

students and lecturers, are identified as collectively

forming the CoP. The implementation details, such as

the placement and distribution of communication

tools (data streams), become critical considerations.

Designing the classroom layout, determining optimal

tool placement, and understanding what aspects of

interactions to capture are essential for effective

MMLA integration. While not delving into an

exhaustive technical inventory, the emphasis remains

on the minimal yet pivotal technology necessary for

CoP and MMLA, ensuring a focused and purposeful

approach to the integration framework.

4.2 Data Collection

In the second phase, we address two considerations:

Data Collection and Sensors and modalities leading

to selecting sensors capable of capturing multimodal

data and identifying the appropriate modalities,

closely aligning outcomes from the preceding phase.

By leveraging the capabilities of the GC's

communication and collaboration technology,

encompassing microphones, cameras, and digital

platforms, we can establish the foundational

groundwork for data collection. Importantly, we

provide in Table 1 an explicit mapping of CoP

learning indicators with digital modalities. This

deliberate alignment emphasized modalities

associated with communication and collaboration

patterns, echoing the collaborative nature inherent in

the CoP framework. MMLA leverages the data

streams generated by the GC’s technological

infrastructure. Interaction patterns, both verbal and

non-verbal, are captured by cameras, providing

insights into the dynamics of collaboration and

engagement. Audio data offers a qualitative

understanding of the tone, intensity, and frequency of

interactions, enriching the contextual characteristics

of communication. Screens contribute valuable

information on collaborative efforts, displaying

shared content and illustrating the evolution of ideas.

The data collection types encompass a spectrum of

modalities, including visual, auditory, and textual

elements. Interaction patterns unveil the social

dynamics within the Community of Practice, while

engagement levels provide a quantitative measure of

participation and involvement. These modalities

collectively offer an overview of the collaborative

learning process, enabling the analysis of the

multifaceted interactions within the Global

Classroom.

4.3 Privacy and Ethics

In the Privacy and Ethics phase, our primary objective

was to ensure that the selected modalities, aligned

with the CoP learning indicators, complied with

established privacy and ethical standards. Building

upon the pre-established requirements from the

previous phases, we validated the alignment of

selected data streams concerning GDPR regulations,

data anonymization, and transparent data processing

policies. Furthermore, we developed a policy

outlining the transparent data processing practices

adopted in our MMLA system. This policy provided

a clear explanation of how data collection and

processing contribute to educational objectives. By

emphasizing transparency, we aimed to build trust

among users, fostering an environment where

individuals are aware of and understand the purpose

of data processing in the context of enhancing

educational experiences.

4.4 Interpretation and Feedback

Based on the outcomes of preceding phases, namely

the established CoP, the identified technological

infrastructure, the selected modalities for data

collection, and the robust privacy and ethics

considerations. Building upon these foundations, we

determined the nature of feedback and interpretations

that the MMLA system should provide in the context

of the GC. We envisioned a simple visualization

framework to capture individual and classroom

performance modalities. These modalities included

indicators such as attendance patterns, levels of

Guiding the Integration of Multimodal Learning Analytics in the Glocal Classroom: A Case Study Applying MAMDA

481

activity, and the frequency and quantity of speech

within the learning environment. To refine our

approach further, we aim to conduct surveys among

both students and lecturers, seeking their insights on

the initial modes of feedback that the MMLA system

should offer. This iterative process allows us to

identify and prioritize feedback elements that can be

subsequently evaluated and tested for their

effectiveness, ensuring that the MMLA system's

interpretations align closely with the needs of the

stakeholders.

4.5 Development

In the Development phase, we address three

considerations namely, Design and Development,

Data Management and Technology integration. All

activities in this phase culminate in insights and

considerations from the preceding phases,

encompassing the established CoP learning

indicators, the identified technological infrastructure,

the modalities chosen for data collection, the privacy

and ethics framework, and the envisioned feedback

mechanisms. Based on the understanding of these

elements, the design and development process,

consist often incorporation of data interpretation

techniques, including algorithms and machine

learning, to provide insights. A key focus is

establishing real-time feedback loops, enabling

continuous responses to learners and educators based

on collected data. This integration not only capitalizes

on the available technological infrastructure but also

ensured a cohesive learning experience for the diverse

CoP. Data analytics is the key understanding

collected data about collaborative learning dynamics

within the CoP. Natural Language Processing (NLP)

techniques decode textual interactions, uncovering

themes, sentiments, and emerging patterns in the

discourse. Computer vision algorithms analyze visual

data, identifying non-verbal cues, group dynamics,

and engagement levels. Quantitative analytics,

complemented by qualitative assessments, provide a

comprehensive understanding of the collaborative

learning experience.

Practically implementing the MMLA integration

involves a phased approach. Initially, educators and

students undergo training to familiarize themselves

with the technological tools and the MMLA system.

Throughout the course, data collection occurs

seamlessly in the background, preserving the natural

flow of the learning process. Periodic reviews and

reflections on the analytics generated by the system

inform iterative adjustments, ensuring continuous

improvement and optimization of the learning

experience.

4.6 Refinement and Validation

The final phase, Refinement and Validation,

represents a continuous improvement cycle,

acknowledging the dynamic nature of educational

environments. As future work, this phase involves a

reflective process that considers the challenges

encountered throughout the previous phases. A

critical aspect is the evaluation and testing of the

MMLA system, which serves as a feedback loop to

inform further refinements. The outcomes of this

phase are twofold. First, it entails an iterative return

to the initial phase, Preparation and Needs Analysis,

to address any identified issues, adapt to emerging

Table 1: Mapping of CoP learning indicators and digital modalities.

Learning Indicator Modality Data Format Sensor/Data

Source

Feedback/Output

Learning Interactions Speech patterns, collaborative

activities, frequency of

contributions

Audio, Video,

Interaction logs

Microphones,

Cameras, Digital

Platforms

Real-time feedback on

group communication

Knowledge Sharing and

Construction

Shared documents, contributions to

discussions, collaborative problem-

solving sessions

Text, Interaction

logs, Screen sharing

Documents, Digital

Platforms

Visualizations of

collaborative activities

Social Presence and

Community Engagement

Attendance patterns, frequency of

interactions, participation in

collaborative activities

Interaction logs,

Video

Digital Platforms,

Cameras

Notifications in case

of lack of presence

Problem-Solving

Strategies

Interaction patterns during

discussions, key contributors,

effectiveness of solutions propose

Interaction logs,

Screen sharing

Digital Platforms,

Cameras

Individual profile

analytics

Role Identification and

Collaboration

Contributions, expertise-based

interactions, distribution of

responsibilities

Interaction logs,

Screen sharing

Microphones,

Cameras

Visualization of roles

dynamics

Reflection and Feedback Individual contributions, responses

to feedback, evolution of ideas

Interaction logs,

Video, Text

Microphones,

Cameras

Mapping reflections

and feedback with

earlier activities.

CSEDU 2024 - 16th International Conference on Computer Supported Education

482

constraints, and ensure ongoing flexibility in the

MMLA system. Second, the validation of the

conceptual model is achieved through continuous

refinement, incorporating expert input and addressing

potential limitations. This process ensures the

alignment of the MMLA system with real-world

applications and scholarly discourse, contributing to

the effectiveness of analytical systems in diverse

educational settings such as the GC.

5 MMLA MODEL FOR THE

GLOCAL CLASSROOM

Building upon the foundational principles of MBOX

(Ouhaichi et al., 2021), an IoT-based system allowing

the collection and processing of multimodal data from

collaborative learning tasks. We present a tailored

MMLA model specifically designed for the GC. This

model envisions a dynamic and adaptive system that

integrates seamlessly with the unique characteristics

of the GC environment. Drawing inspiration from the

three-layered architecture of MBOX, our model

revolves around the Users Layer, Communication

Interface Layer, and Processing Layer.

5.1 Model Components

The model is made up of three layers. The users layer

constitutes the first tier engaging students, teachers,

and various stakeholders such as researchers and

education managers. In the GC, this layer represents

participants interacting within the collaborative

learning space across multiple remote classrooms.

The users layer serves as the focal point for

capturing diverse perspectives and interactions within

the GC. Second, comes the sensing interface layer,

which represents communication technology

embedded in the GC, functioning both as a facilitator

of interactions and a data capture mechanism.

Communication tools, including microphones,

cameras, and projectors, form the technological

infrastructure of this layer. This interface serves a

dual purpose, enabling seamless communication

among participants while also capturing multimodal

data generated during the collaborative learning

process. The arrow connecting the users’ layer to the

sensing interface symbolizes capturing interactions

and communication within the GC, creating a bridge

between human engagement and data collection. At

the top of our model lies the processing layer, where

data storage, management, and MMLA activities

unfold. This layer is the hub for processing

multimodal data collected from the sensing interface.

Leveraging the insights gained from MBOX, this tier

could adopt a multi-level architecture, comprising

edge, fog, and cloud layers. The edge layer handles

real-time data processing at the group level,

facilitating immediate feedback through local

mechanisms. The fog layer extends the perspective to

the classroom or school level, offering additional

insights and feedback mechanisms. Finally, the cloud

layer processes data at a global level, contributing to

a comprehensive understanding of collaborative

learning experiences within the broader educational

system.

Figure 2: MMLA model for Glocal education spaces.

The arrows in our model illustrate the cyclical

nature of MMLA within the GC. Data flows from the

users’ layer to the sensing interface, where

interactions are captured. This data is then directed to

the processing layer, where MMLA activities,

including analytics and interpretation, take place.

Feedback and outputs generated are then looped back

to the users, completing the cycle.

5.2 Usage of Integrated MMLA

The integration of MMLA into the Glocal Classroom

transforms the user experience for students,

educators, and researchers. For students, the MMLA

integration introduces a personalized learning

journey. Real-time feedback mechanisms from the

edge layer provide instant insights into collaborative

efforts, allowing students to adapt and optimize their

learning strategies. The Sensing Interface Layer

captures nuanced interactions, providing a detailed

view of their participation and engagement.

Educators benefit by gaining a more holistic

perspective on collaborative learning dynamics. By

Guiding the Integration of Multimodal Learning Analytics in the Glocal Classroom: A Case Study Applying MAMDA

483

capturing both verbal and non-verbal cues, educators

can better understand group dynamics and tailor their

teaching strategies. This enhanced view facilitates

informed decision-making, allowing educators to

refine their pedagogical approaches. Researchers now

have access to a wealth of multimodal data, fostering

in-depth analyses and scholarly contributions. The

multi-level architecture of the Processing Layer

provides a scalable and adaptable framework for

researchers to explore human learning across various

scales. The cyclical flow of data enables continuous

refinement, aligning the MMLA system with real-

world applications and scholarly discourse. The

integrated MMLA model serves as a valuable tool for

advancing research in the field of educational

technology.

6 DISCUSSION

In this study, MAMDA serves as a foundational guide

for integrating MMLA into the GC. The adaptation of

the MAMDA model reveals crucial insights into the

dynamics of the learning environment, prompting

necessary adjustments for a more effective

integration. Notably, the repositioning of Data

Management considerations to the Development

phase and the relocation of Sensors and Modalities to

the second phase of Data Collection exemplify a more

coherent workflow that aligns with the dependencies

of educational technology integration (Ouhaichi et

al., 2019). The phased consideration of Interpretation

and Feedback acknowledges the distinct production

workflow of educational systems, enhancing the

responsiveness of the learning environment (Messina

Dahlberg & Bagga-Gupta, 2014). These modifica-

tions not only refine the MAMDA model but also

address challenges associated with privacy and ethics,

providing a more robust framework for responsible

MMLA integration. The study's insights contribute to

the ongoing evolution of MMLA systems, opening

avenues for future research and advancements in

smart learning environments.

Researchers can leverage our model as a

comprehensive guide and template for integrating

MMLA into diverse learning environments,

especially those characterized by global

communication and collaborative learning. The

structured six-phase approach outlined in our model,

derived from MAMDA, serves as a roadmap for

researchers seeking to understand, implement, and

refine MMLA systems. By employing our model,

researchers and other practitioners can navigate the

complexities of data collection, ethical considerations

(Alwahaby et al., 2021), and technological integration

within learning environments. The adaptable nature

of our framework allows researchers to tailor it to

different educational settings, ensuring relevance

across various cultural and geographical contexts.

Ultimately, our model provides a methodological

foundation for advancing the exploration and

implementation of MMLA, fostering a deeper

understanding of collaborative learning experiences

and paving the way for further innovations in

educational research and technology.

7 LIMITATIONS

While integrating MMLA into the Glocal Classroom

presents a promising avenue for enhancing

collaborative learning, certain limitations inherent in

this study must be acknowledged.

The success of the MMLA model relies heavily

on the existing technological infrastructure within the

GC. Limitations in the hardware, network

capabilities, or compatibility issues may impact the

integration and functioning of the MMLA system

(Blined). Variability in technology across different

instances of the GC may introduce challenges in

achieving uniform MMLA implementation. The case

study approach employed in this research focuses

specifically on the GC setting, representing a unique

implementation of Glocality in education. While the

findings contribute valuable insights within this

context, the generalizability of the MMLA model to

diverse educational environments remains a

consideration. Variations in classroom structures,

cultural nuances, and technological setups may affect

the applicability of the model in other settings. The

mapping of CoP learning indicators with MMLA

modalities is a complex task. The accuracy and

relevance of this mapping may be subject to

interpretation and could impact the fidelity of

learning insights derived from the MMLA system.

Further refinement and validation of this mapping

process are essential for ensuring the robustness of

the model. Acknowledging these limitations is crucial

for a comprehensive understanding of the study's

scope and implications. Future iterations of MMLA

integration in diverse educational contexts can build

upon these insights to address and overcome these

limitations, contributing to the continued evolution of

collaborative learning analytics.

CSEDU 2024 - 16th International Conference on Computer Supported Education

484

8 CONCLUSION

In this paper, we present a case study exploring the

integration of MMLA into the GC. Guided by the

MAMDA model, our study unfolds six phases:

Preparation and Needs Analysis, Data Collection,

Privacy and Ethics, Interpretation and Feedback,

Development, and Refinement and Validation. The

main contribution lies in demonstrating the practical

application of MAMDA in the GC, offering insights

into the integration of MMLA to enhance

collaborative learning experiences and inform

educational practices. Through this study, we tested

the application of MAMDA within the dynamic

educational ecosystem of the GC. We also critically

examined and refined the model to better align with

the dependencies of educational technology

integration. The strategic repositioning of key

considerations, such as Data Management, Sensors

and Modalities, and Interpretation and Feedback,

attests to the adaptability and responsiveness of the

model in real-world scenarios. As a result of this

systematic approach, we have demonstrated how

MMLA can be embedded into the GC context,

leveraging various data sources. The incorporation of

MMLA aligns with the CoP principles within the GC,

fostering a collaborative and community-driven

educational environment. To illustrate the resulting

conceptual integration, we present a three-layered

architecture that facilitates the capture, processing,

and analysis of multimodal data.

REFERENCES

Ahad, M. A., Tripathi, G., & Agarwal, P. (2018). Learning

analytics for IoE based educational model using deep

learning techniques: architecture, challenges and

applications. Smart Learning Environments, 5(1).

https://doi.org/10.1186/s40561-018-0057-y

Alwahaby, H., Cukurova, M., Papamitsiou, Z., &

Giannakos, M. (2021). The evidence of impact and

ethical considerations of Multimodal Learning

Analytics : A Systematic Literature Review. In The

Multimodal Learning Analytics Handbooks (pp. 1–34).

EdArXiv. https://doi.org/10.35542/OSF.IO/SD23Y

Blikstein, P., Ochoa, X., & Blikstein, P. (2013). Multimodal

learning analytics. ACM International Conference

Proceeding Series, 102–106. https://doi.org/10.1145/

2460296.2460316

Christensen, J., Altenreiter, M., & Meixner, K. (2022).

Remote Is Not So Far Away: A Self-Reflective Case Of

Internationalisation Using Collaborative Online

International Learning In Social Work Education.

Tiltai, 1–17. https://doi.org/10.15181/TBB.V88I1.2495

Christensen, J., Thönnessen, J., & Weber, B. (2020).

Knowledge Creation in Reflective Teaching and Shared

Values in Social Education: A Design for an

International Classroom. Educatia 21, 19, 11–23.

https://doi.org/10.24193/ed21.2020.19.02

Cukurova, M., Giannakos, M., & Martinez-Maldonado, R.

(2020). The promise and challenges of multimodal

learning analytics. In British Journal of Educational

Technology (Vol. 51, Issue 5). Wiley Online Library.

https://doi.org/10.1111/bjet.13015

Hancock, D., Algozzine, B., & Lim, J. (2021). Doing case

study research: A practical guide for beginning

researchers. https://books.google.com/books?hl=en&l

r=&id=e7lLEAAAQBAJ&oi=fnd&pg=PP1&dq=Doin

g+case+study+research:+A+practical+guide+for+begi

nning+researchers&ots=5-q8SHSS8j&sig=mMvG9h

OKduudYk0Se17a3w0fpgc

Messina Dahlberg, G., & Bagga-Gupta, S. (2014).

Understanding glocal learning spaces. An empirical

study of languaging and transmigrant positions in the

virtual classroom. Learning, Media and Technology,

39(4), 468–487. https://doi.org/10.1080/17439884.20

14.931868

Ouhaichi, H., Olsson, H. H. H. H., & Bosch, J. (2019).

Dynamic data management for machine learning in

embedded systems: A case study. Lecture Notes in

Business Information Processing, 370 LNBIP, 145–

154. https://doi.org/10.1007/978-3-030-33742-1_12

Ouhaichi, H., Spikol, D., & Vogel, B. (2021). MBOX:

Designing a flexible IoT multimodal learning analytics

system. Proceedings - IEEE 21st International

Conference on Advanced Learning Technologies,

ICALT 2021, 122–126. https://doi.org/10.1109/

ICALT52272.2021.00044

Ouhaichi, H., Spikol, D., & Vogel, B. (2023). Rethinking

MMLA: Design Considerations for Multimodal

Learning Analytics Systems. L@S 2023 - Proceedings

of the 10th ACM Conference on Learning @ Scale.

https://doi.org/10.1145/3573051.3596186

Patel, F., & Lynch, H. (2013). Glocalization as an

Alternative to Internationalization in Higher Education:

Embedding Positive Glocal Learning Perspectives.

International Journal of Teaching and Learning in

Higher Education, 25(2), 223–230. http://www.ise

tl.org/ijtlhe/

Runeson, P., Höst, M., Rainer, A., & Regnell, B. (2012).

Case Study Research in Software Engineering:

Guidelines and Examples. In Case Study Research in

Software Engineering: Guidelines and Examples.

https://doi.org/10.1002/9781118181034

Worsley, M. (2018). Multimodal learning analytics’ past,

present, and, potential futures. CEUR Workshop

Proceedings, 2163.

Guiding the Integration of Multimodal Learning Analytics in the Glocal Classroom: A Case Study Applying MAMDA

485