Navigation – Plan du site

AccueilNuméros33-1/2A platform approach to experienti...

A platform approach to experiential learning

Data literacy and technical skills development
Une approche de l’apprentissage expérientiel par les plateformes. Cultures des données et développement des compétences techniques
Sandra Robinson
p. 101-124

Résumés

Cet article s’attache à rendre compte de la conception et de la restitution d’un programme de formation en apprentissage expérientiel fondé sur la pratique, qui initie des étudiants de premier cycle en étude de la communication et des médias, au sein d’une université canadienne, aux défis des données sociales, à travers l’utilisation d’une plateforme d’analytique. Le développement de ce programme de formation est fondé sur une approche pédagogique de l’apprentissage expérientiel, des technologies de l’éducation et de l’enseignement, tout en examinant les théories récentes de l’étude des plateformes au sein du domaine de la communication et des médias. Cet article développe l’idée selon laquelle, à travers l’affirmation d’un lien clair entre théorie, éthique de l’information, et pratique, les étudiants sont en mesure d’acquérir des compétences en analyse critique nécessaires à la compréhension du lien complexe entre données sociales et communication, mais aussi de développer des compétences en culture des données qui leur permettent d’étendre leur formation en étude critique de la communication.

Haut de page

Texte intégral

Introduction

  • 1 The curriculum design project is currently in the third and final year of a three year pilot projec (...)

1In this contemporary moment, social media is a popular and, in many instances, a problematic social networking phenomenon. Social media platforms, from Twitter to Instagram and more, provide individuals, celebrities, politicians and political parties, social activists, sports teams, as well as governments and corporations, a platform to communicate with diverse audiences. The upside is in capturing the attention of a particular audience at a particular moment to influence attitudes or amplify a message. The downside is that in the ensuing cacophony, conversations on social media are increasingly difficult to track and makes sense of and determine what the “noisy chatter” means for an issue, event, campaign, product, or service. This article discusses preliminary observations of a three-year curriculum design and delivery project for a practice-based, experiential learning course in data analysis and visualisation1. The course, called Visualising social media : hashtags, keywords, and conversations (abbrev. VSM), is designed to respond to the growing demand for communication and media professionals to acquire specific competencies in data analytics and social media analysis to be able to work with large data sets that social networking platforms generate (Botibol, 2016 ; Whorton, Casillas, Oswald, and Shaw, 2017). It introduces students in an undergraduate communication and media studies program, at a Canadian university, to the challenges of working with social data using an analytics platform. The course enables students to probe more deeply into the “noisy chatter” of social media data through basic data mining, by making data visualisations, and by exploring what their dataset ‘tells’ them—about what it means—in relation to a contemporary problem or issue they are researching. Students completing the VSM course will have developed professional research skills suitable for advanced social media monitoring and data analysis.

2Early student feedback suggests the course meets its main goal, which is to give students the opportunity to dig deeper into data produced through social media and to understand what insights may be gained through analysis and visualisation, while being sensitive to the ethical and critical challenges, and potential limits, of data-driven analysis. The following discussion outlines core theories in experiential learning and the role of the active classroom, and the shift in communication and media technologies toward platforms that enable rich social communication and interaction, which can be collected for meaningful social analysis. Many of the foundational concepts in experiential learning are appropriate to the curricular design and delivery of a course in social media analysis and visualisation designed to build practical skills and expand data literacy necessary for postgraduate professional practice in the communication and media fields.

3The first section of the paper reviews some of the theories and concepts that ground the experiential learning approach before turning to issues relevant in course design for experiential learning in a technologically enabled active classroom environment. The discussion diverges from some of the conventional literature on experiential learning by exploring research in platform studies, which explore “how platforms preconfigure specific practices through designed features and functions” as well as specific uses of the data they generate (Weltevrede and Borra, 2016 : 1). Following that discussion, we outline the curricular design, course structure, and learning objectives and explore the challenges and affordances of the platform approach. We argue that a platform approach is required for a robust and interoperable social media analysis workflow as a necessary condition for professional research practice and skills building.

A Platform Approach to Experiential Learning and Technical Skills-building

Experiential Learning

4Creating a successful and productive learning experience for students, in addition to the opportunity for experiential learning, requires a teaching and learning approach that is student-centred and combines theory and practice (Kolb, 1984). The instructor in this scenario is a facilitator and guide through the learning process. In the context of teaching computer-based skills in the VSM course, the instructor must be sensitive to the particular challenge of teaching non-technical students, who themselves must overcome anxieties and personal challenges in learning computationally intensive software (McGrath et al., 2015). Teaching data analytics therefore requires a certain “sensitization” early in the course to introduce students to data analytics and build exposure and understanding around the concept of data and the potential for data-driven research in professional, postgraduate practice (Asamoah et al., 2017 : 167 ; see also Bruns, 2016). This adds a unique context to the role of social media as part of a listening apparatus through which communication researchers can conduct conversation monitoring and response tracking to understand audiences and markets through analysis that can yield particular kinds of insights. This sensitization process is accompanied by the application of theory and principles of data mining and analysis, network analysis, and data visualisation as well as ethical principles that guide informed research practice. Taken together, students can then apply these foundational concepts as they work on building their data analysis competencies and professional analytics skills (Sommerville, 2013 ; Asamoah et al., 2017).

5Experiential learning has a rich tradition in pedagogical scholarship and has been successful in computer-based learning environments (Winne, 2006). John Dewey inspired contemporary experiential learning models by encouraging a pedagogical approach that enhances understanding by connecting real-world experience and practice to learning objectives to encourage collaboration, participation, and engagement (1966). This learning modality unfolds within a broadly co-constructed learning environment (Vygotsky, 1978) ; that is, students learn through social interaction with their peers in a continuous process grounded in experience. In this view, learning unfolds as an emergent dialectic between experience and practice wherein learning is both transformative and dynamic (Kolb, 1984 : 38). The emphasis, therefore, is on the social process itself within the context of self-regulated student learning that encourages a reflexive encounter between experience and practice and the student’s goals (and see Table 2 for examples).

6Scholarship on teaching and learning over the last several decades has drawn attention to the experiential learning model as well as the use of digital technology in the classroom for student learning and engagement (Kolb, 1984 ; Winne, 2006). As technology in the classroom becomes widespread, there is also attention on instructor knowledge and specialized skills development, or « professional learning » focused on an « educator’s practice that is linked to students’ learning outcome » within a contextual environment that connects theory with practice (Ng, 2015 : 25, 28 ; White, 2013). Therefore, the requirement for courses that adopt the experiential learning model, such as the one under discussion in this paper, must also ensure the instructor has the requisite skills, experience, and theoretical background to teach the course.

7In experiential learning courses, students are self-directed ; they regulate their activities as part of a more flexible in-class experience. While there has to be some kind of course structure and assessment to meet overall program degree requirements (and see Figure 2 and Table 2), the students have more autonomy in organizing their activities to meet the course deliverables. An experiential learning environment is not simply one in which students do a few tasks that simulate a professional experience ; rather, students create and work within a milieu of experience that links content, substantive ideas and theory, with a process that facilitates learning by doing (Chapman et al., 1995 ; Sommerville, 2013). In this environment of experience and practice, students are considered as active agents engaged in the process of learning and much of the recent scholarship on experiential learning in computer-based courses stress the importance of creating “instructional conditions” that support students (Gasevic et al., 2016 : 208 ; Gebre et al., 2015). This “instructional scaffolding,” discussed further on, is designed to provide support to students on an as-needed basis for aspects such as core learning concepts, procedures, and other domain knowledge, self-regulated learning, and learning about procedures and functionality of the analysis platform (Azevedo and Hadwin, 2005 : 370). While there is ample flexibility built into the experiential learning course, it needs a clear structure as a foundation for the learning objectives and activities to create optimal instructional conditions. An experiential learning course requires a concrete vision with an assessment strategy for students working in small teams with clearly “defined educational parameters [and] group working agreements, activity learning goals, and a big-picture design plan” (Chapman, McPhee and Proudman, 1995 : 243). These parameters and workflow are illustrated in Figure 2 and Table 2 and detailed below.

8Teaching social media analytics aligns neatly with experiential learning theory conceptualized as “networked learning” that leverages digital technologies for use in teaching and learning activities (Gruzd et al., 2016 : 5). Studies of social media use, to support teaching and learning, fit within the core experiential model by emphasizing practice-based skills building, “extending the range of the learning environment” through collaborative, “networked learning” and promoting social interaction among the students (ibid. ; Jones, 2015).

The platform approach as methodology

9The methodology behind the data analysis and visualisation course discussed here borrows conceptually from platform studies, as a subfield within communication and media studies, in which a platform is understood as a technical architecture that enables interoperability. Platforms are larger-scale software architecture that enables programmability and customization, interaction, participation, and connectivity, between individual users, collectives, and markets. In formal terms, a platform’s “technical architecture emphasizes the provision of connection, programmability, and data exchange with applications” (Plantin, 2016 et al. : 6). Platform studies provide students a theoretical grounding through which to explore substantive scholarly work that addresses technological, political, economic, and social dimensions linked to contemporary social media platforms. Research in platform studies extends from the politics of platforms (Gillespie, 2010 ; Burgess and Bruns, 2015) to their political economy (Srnicek, 2016) and surveillant tendencies (Zuboff, 2019) and to critical concerns about the way the technical architecture shapes and organizes sociality (Bivens, 2017 ; Bucher, 2012 ; Van Dijck, 2013) through a technocultural logic (Langlois, Elmer, McElvey, and Devereaux, 2009 ; Gerlitz and Helmond, 2013). Students are encouraged to think from this research orientation about how platforms—both large scale social platforms and small-scale analytics platforms—do discursive work in shaping how we make sense of social conversation.

10Platforms highlight the turn in the last two decades to technical architecture that extends « programmability, affordances and constraints, connection of heterogeneous actors, and accessibility of data and logic through APIs » which are an important dimension for the collection and analysis of social media data (Plantin et al., 2016 : 2 ; Janetzco, 2017). In media and communication studies the definition of a platform thus turns on a technical architecture that is programmable, as discussed above, through an Application Programming Interface (API), which other software can connect to, adding new features, functions, and interoperability. Early research into social media platforms, such as Facebook and YouTube, were strictly defined in these computational terms drawing clear boundaries around what constituted a platform (Andreesen, 2007 ; Montfort and Bogost, 2009). However, this narrow computational definition of a platform forecloses on other conceptions including technical solutions and analytics workflows that are designed to be interoperable and seamlessly integrated. By linking research in platform studies and adapting some of its key terms that service pedagogical goals, we are re-situating the term platform by suggesting that techniques and practices required for social media analysis demand a unified approach that operates as a scaled-down integrated workflow platform. In the VSM course, multiple software serves one integrated workspace for students to develop competencies in data analysis and visualisation. We refer to this workspace as a platform.

The platform

11In the context of the VSM course, the platform is conceived as an assemblage of software, technique, and practice required to build a repertoire of knowledge and competencies in data analysis for students. This is standard practice in data science whereby several applications and programming interfaces are assembled into a platform for analysis, which includes data collection and archiving and data management software connected to analysis and visualisation applications to ensure a robust workflow (Asamoah et al., 2017). The platform approach ensures that the technical features in the suite of applications required to conduct the data analysis course are well integrated and function reliably.

12The platform for the VSM course includes data collection using a Google Sheets add-on called the Twitter Archiver that connects to the Twitter API and streams data into an archive sheet on Google Drive, which functions as basic data management (see Table 1 : Analytics Platform). The dataset can be connected to Tableau Software for analysis and visualisation and then used to create a structured dataset that can be imported into Gephi—The Open Graph Viz Platform, to generate a network analysis and visualisation2. This software assemblage was selected and tested to provide an integrated workflow that enables students to effectively move from data collection to visual analysis. The notion of platform also incorporates the common usability metaphor in analytics software known as the work canvas conjoined to a specific workflow designed for Twitter analysis and works well to give students the experience of interoperability between different applications.

Table 1 : Analytics Platform

Data collection & analysis

Application (software program)

Data Source

Twitter (primary data source)

Optional data sources :

  • Facebook

  • Instagram

  • Other social media

Data Archiver &
data management

Connected to the Twitter API

Twitter Archiver Google Sheet (TAGS)

Or, Twitter Archiver by Digital Inspiration

Optional tools :

  • Microsoft Excel

  • Libra Office Calc

  • Google Sheets

  • Numbers

Data analysis & visualisation

Tableau Software (primary, required)

Optional tools :

  • Microsoft Excel

  • Libra Office Calc

  • Google Sheets

  • Numbers

Network analysis & visualisation

Gephi—The Open Graph Viz Platform (primary, required)

13Using the platform concept as the model to define and shape the approach in the VSM course is, we argue, crucial for an introductory course designed for non-technical learners outside of specialized data science and computer science fields. It helps to create a sense of continuity and consistency in the user experience. Students enrolled in the VSM course have already completed training in communication and media theory and in qualitative and quantitative research methods in the year preceding the course discussed here, including an introduction to IBM’s statistical package SPSS, but have not been trained in social data visualisation and network analysis using specialized tools for those purposes. Therefore, from a teaching perspective, a platform approach, coupled with important components of experiential learning within a digital workspace, facilitates « active learning » among students (Felder and Brent, 2009). The approach enables an emergent and lively learning environment that more easily accommodates student diversity, social dynamics, and learning styles because they have a robust research platform to work with (Kolb, 1984 ; Roehl, Reddy, and Shannon, 2013).

Curricular design goals, learning objectives, and course structure

Curricular design goals

14The data mining and analysis course discussed here enables students to explore what data are, as noted in the foregoing section, through basic data mining, by making data visualisations, and by exploring what their dataset tells them—the meaning in the data—in relation to the contemporary issue they choose to research. This cycle of mining, making, and meaning is part of the thought style and habit of practice that emerges as a condition of experience for students working on the analysis platform (cf. Fleck, 1979). The process emphasizes a clear link between theory, information ethics (Floridi, 2011), and practice to build critical skills necessary for analyses of the complex relationship between data and communication. Through this process, students can develop data literacy skills that extend their training in critical communication studies (boyd and Crawford, 2012 ; Gruzd et al., 2016). The course follows the pedagogical method established in Asamoah et al. (2017) for the design and delivery of a data analytics course in data science, but adapted for communication and media studies students who are not technical specialists. While there are resources available for social media analysis methods (Bruns, 2016 ; Sloan and Quan-Haase, 2017 ; Borra and Rieder, 2014) there is limited research on course design in this area outside of data science and computer science pedagogy such as found in Asamoah et al. (2017), or in business analytics (Atwong, 2015). In this way, the curriculum design and pilot project reported on in this paper addresses a gap in communication and media studies around how to teach data analysis and visualisation at the undergraduate level.

15The curricular design goals are distinct from the learning objectives of the course. The key curricular design goal is to develop and deliver the fourth-year workshop course—Visualising social media : hashtags, keywords, and conversations—as part of a three-year pilot project (2017–2020) to assess the demand for, and efficacy of, an applied course in data analysis and visualisation for communication and media studies students. The core learning objective is to give students the opportunity to explore data generated on Twitter (or, in future, other social media platforms) and to understand what insights may be gained through analysis and visualisation using the industry standard analytics package Tableau, and Gephi. In addition to the core learning objectives, which are discussed in more detail below, the course design and development project has three goals :

  1. To research and test an analytics platform, which incorporates a minimum of three software applications for social media analysis that enables a reasonable degree of interoperability ;

  2. To design a scalable course that after the three-year pilot project can scale up to enable more students to take the course through a blended course model that combines on-campus intensive classroom sessions with online delivery ;

  3. To provide adaptable workshop-style modules applicable to different learner styles and levels from introductory research methods for second year undergraduate students to more advanced techniques for graduate students.

16The discussion in this paper focuses on the design and implementation of these goals, which are aspects of the project that are nearing completion after 24 months of the three-year pilot project. In the spirit of interdisciplinary and constructivist learning as part of an experiential methodology, the workshop-style course discussed here is a hands-on class that functions as a software lab. The students are engaged in a semester long project of research design, data collection, data analysis and visualisation, and a final project report. The mode of learning closely adheres to the theoretical tenets of constructivist learning whereby « outcomes of the learning process are varied and often unpredictable » (Wurdinger, 2005 : 69). This unpredictability is part of experiential learning and students are self-regulated although supported by adaptive « instructional scaffolding » (Gasevic et al., 2016 : 208) in the form of instructor lead tutorials to guide them through how to use the social media analysis platform in the VSM class and link its value to a learning objective. The term scaffolding invokes the notion of a more supportive and flexible instructional apparatus that is directed toward providing « assistance to students on an as-needed basis, fading the assistance as learner competence increases » (Azevedo and Hadwin, 2005 : 368).

17To create a sound pedagogical framework for the course that can leverage “instructional scaffolding” and support students without recourse to a traditional focus on “micromanaging” them or their activities, the course requires careful planning. To ensure the course runs tethered to its learning objectives, there are several recommendations that can help to integrate experiential learning concepts into the course design to keep it on track. At the very outset of the curricular design process, the course was designed around one final or main project that can be approached incrementally throughout the term to motivate students toward a clear goal (Wurdinger, 2005 : 63). As the course proceeds, we constructed a combination of lab activities that contribute to the main project to give students a sense of building not only their knowledge, but to enable students to see their project taking shape over the duration of the course (see Figure 2 and Table 2). In experiential learning courses, lectures and conventional course readings may be scaled back, but when we do incorporate substantive theory, we ensure it links back to experiential activities so that the readings are resources for the final project. In-class tutorials and software lab activities need to be challenging but manageable with clear expectations. It is important the students have enough time to identify core learning objectives, clarify their experience, practice the skills, and keep a focus on the end goal of their final project. Finally, students sometimes need to change direction whether it is because their data was corrupt or difficult to collect, or they were working on an idea that was simply not meaningful and relevant to their interests, so flexibility is important for student success (adapted from Wurdinger, 2005 : 63).

Learning Objectives

18There are three main learning objectives. The core learning objective for the VSM course, from which the others flow, is to build professional data analysis skills following the mining, making, and meaning workflow that the platform approach encourages. Through a range of activities students learn how to explore a dataset collected on Twitter and conduct analysis and visualization to make sense of what the data “tells” them. Students are introduced to the common work canvas usability metaphor in analytics software and a structured workflow from analysis to visualization. This usability metaphor translates well to other analytics platforms students may encounter in professional settings after they graduate such as Facebook Insights or Google Analytics.

19The second learning objective is to develop research methodology with a focus on ethical research practice. Students engage in a research-intensive process that requires a well-developed methodology, which is sensitive to the ethical and critical challenges, and potential limits, of data-driven analysis. Finally, data literacy as a learning objective ensures that as communication and media professionals, graduates of our program can understand the implications of social media and networked publics in our global mediasphere by locating data-intensive research within wider social, political, and cultural concerns. Tools and techniques exist to enable access to data, but well-informed researchers also need to understand that just because we can access the data does not mean we should—and sound research methods and good research design ensure reflexive practice even in commercial settings (Sloan and Quan-Haase, 2017 : 670).

Course structure and activities

20The VSM class is twelve weeks long—one semester in the Canadian post-secondary education system—in either the Fall term from September to December or the Winter term from January to April. Student enrolment varies between 12 to 25 students in each term and through 2017–2019, the first two years of the pilot project, 84 students have taken the workshop with an overall, class-wide, final grade average of A- (80–84 percent). The workshop is intentionally capped at a maximum of 25 students and class size varies from semester to semester although a smaller size enhances the learning environment and instructor to student ratio. Students work in small teams of two to three students and are encouraged to collaborate and problem solve within and between groups in the class and share techniques, feedback, and work processes. The course runs for 180 minutes once a week in a computer-based learning environment, or active classroom, fitted with state-of-the-art HDMI visual projection systems, effective lighting controls, modular tables and ergonomic seating (see Figure 1). Students work in self-selected groups of two or three for the term at a worktable ‘pod’ with streaming connectivity to a large screen monitor. Students can connect their own laptops to the pod monitor and the instructor can also project to all pods during software tutorials.

Figure 1 : An active classroom at Carleton University

Figure 1 : An active classroom at Carleton University

Image used by permission of Instructional Media Services at Carleton University.

21The semester proceeds through an introduction to social data collection and visualisation, research ethics and research design, exploring open-source browser-based automated ‘dashboards’ for visualising Twitter data, and then into progressively more challenging software tutorials in Tableau Software and Gephi. The dashboard concept is crucial for students to understand how data can be represented visually and how particular kinds of narratives are shaped through distinct visual techniques that use social data to tell a story. The open access social media dashboard Social Bearing is used to introduce the idea of a data dashboard, but also to explain some of the limitations of automated dashboards in which the underlying data and visualisation technique are not always transparent (Langlois, Redden, and Elmer, 2016)3. Social Bearing also enables students to explore different hashtags and keywords in circulation on Twitter and decide what topic they will research for the term. After working within Social Bearing and documenting their findings, students then move into the data collection phase based on the hashtag or keyword they choose to track for three to five weeks.

22During the first half of the semester students are introduced to theoretical concepts about social networks, data analysis and visualisation, and research methods and ethics, and key terms are incorporated into the software tutorials and assessments. Students in our undergraduate degree program are well versed in qualitative and quantitative research methods by the time they register in the VSM course. However, in the context of social media analysis, students are required to consider the particular ethical challenges that arise in research using publicly available, and socially rich, data. These ethical challenges can be convened around three central concerns. Firstly, in an era of “networked publics,” the audience is not immediately visible in a tangible sense : it is easy to lose the original context of a tweeted message because our networked environment is “persistent, replicable, searchable, and scalable,” meaning we have no control in most cases of who sees our tweets and other social media posts and whether they go “viral” (boyd, 2010 : 12). The “context collapse” in which distinctions between public and private are blurred make data handling practices challenging for researchers (ibid.). Secondly, Twitter users consent to information being seen and shared on the Twitter platform, but its data is widely available to be collected on a limited basis through the public API, and in even larger amounts through privatized data “pipes” accessible for a fee through Twitter’s Gnip service4. A central tenet of ethical research practice is to minimise the “risk of actual or potential harm during data collection, analysis and publication” (Williams, Burnap, and Sloan, 2017 : 1160). Professional researchers need to consider how they will handle individual tweets that they want to use as examples or as part of a larger case study. Researchers should seek informed consent if they want to feature or discuss a particular tweet or Twitter account holder in their research (ibid.). Finally, professional corporate and academic research is based on trust around data handling and ethical practice must inform research on a case-by-case basis : there is simply no defined standard and no substitute for “good research design and constant reflexive practice” (Sloan and Quan-Haase, 2017 : 670). This focus on research ethics is tied to the course learning objectives.

23For most of the term students spend the first 60 minutes of each class participating in a tutorial and then the remainder of time (120 minutes) on their group research project with support from the instructor if required to resolve problems with the software, datasets, research scope, and related project issues. Students gradually gather data and begin to work through an analysis of their Twitter dataset in Tableau to understand the role of different measures, such as the number of tweets, and different dimensions in their dataset, including message content, retweets, @mentions, and followers, which are all important factors to understand in terms of analyzing what the data can tell them about the issue they are tracking on Twitter. In the second half of the semester students spend more of their time in self-directed activities working on more advanced functions in Tableau to create visualisation worksheets, dashboards, and stories from the Twitter dataset.

Figure 2 : Twelve-week course structure

Figure 2 : Twelve-week course structure

24Students work across the analytics platform moving from the Twitter Archiver into Tableau to trace out some of the key dimensions in their dataset. Tableau allows students to create a time series visualisation to show the volume of tweets over time and to locate highly active Twitter accounts, popular tweets, and to analyze and explain thresholds of activity in the context of the wider world or local events related to their research topic. The time series is the first visualisation students create and from that point on, students can analyze secondary hashtags, keywords, and retweets, and explore dimensions that can help to explain key influencers in their Twitter network. Students create data visualisations that enable them to explore who is tweeting and at whom ; what are the most popular retweets ; and, who is most active versus who is most visible in their network. Mining the social dataset “allows for the extraction of all kinds of social knowledge” and professional communicators are faced with problematic challenges in private industry around the incessant monetization of these data (Langlois, Redden, and Elmer, 2016 : 4). A reminder of the ongoing challenge for ethical considerations and reflexive practice.

25While the semester has a clear structure and core learning objectives, the pace of activity can be adjusted according to each groups’ capabilities and all software tutorial guides are posted to the university’s Moodle-based learning management system so that students can work through different stages or repeat steps as necessary at their own pace. There are four distinct graded assignments to provide instructional feedback and encouragement and, with the exception of the student presentation, have flexible due dates : 1. social media and ethical research ; 2. project outline detailing the issue, event, or topic being tracked on Twitter and research design ; 3. presentation of findings ; and, 4. main (final) project consisting of the written report, Tableau Workbook, and Gephi network visualisation (see Figure 2 and Table 2). Working with Gephi is an important capstone in visualising their Twitter network.

Table 2 : Course modules and lab activity

Modules
(3-week segments)

Detailed Learning Objectives

Description

Tools

Lab Activity

Evaluations*

1. Research Methods and Ethics

I. Research methods in communication studies : social data, and data mining and analytics

II. Social media and ethical research practice

III. Ethical frameworks and Twitter data

- Effective communication

- Critical thinking

- Ethically inflected research

- Research methodology & platform studies

-Understanding social network analysis

I. Explore the foundations of social network analysis

II. Examine substantive methodologies for social media research

III. Create an ethical framework for term project

- Twitter

- Google Trends

- Social Bearing

Consult assigned readings


Explore hashtag and keyword activity by tracing conversations, events, people, and issues on Twitter and other social media


Draft research design and ethical framework for term project

Activity report : Self-reflection and group feedback


Research design and research question


Ethical review of social media company

2. Visua-
lisation techniques

I. Introduction to Tableau Software

II. Data visualisation techniques

III. Advanced data visualisation techniques

- Analytics platform and work canvas skills

- Under-
standing data management

- Tableau skills

- Data mining

- Visualisation techniques and best practices

I. Explore analytics platform and Tableau work canvas

- Manage Twitter dataset

II. Generate Time Series and retweet visualisation

III. Advanced visualisations : Generate retweet pattern analysis ; secondary hashtags ; and network influencers

- Tableau Software

- TAGS

- Google Sheets

- Other spreadsheet applications

- Tableau tutorials

- Data collection

- Data management

- Project research (social & political dimensions)

- Analyzing Twitter dataset and visualisations


Experimentation and reflection using the analytics platform and work canvas

Activity report : Self-reflection and group feedback


Tableau Workbook


Project outline and research design

3. Network Analysis and Visualisation

Introduction to Gephi - The Open Graph Viz Platform

- Network analysis skills

- Structured datasets

- Network properties

I. Create structured dataset from archived Twitter data

II. Create network visualisation in Gephi

III. Analyze and explain network visualisation

- Gephi


- Tableau Software

- TAGS

- Google Sheets

- Gephi tutorials

- Data analysis and visualisation in Tableau workbook

- Network analysis and visualisation in Gephi


Experimentation and reflection using the analytics platform and work canvas

Activity report : Self-reflection and group feedback


Tableau Workbook


Gephi project file


Project presentation : work in-progress

4. Final project

- Effective communication skills

- Critical analysis

- Advanced data visualisation skills

- Professional report writing

Produce a final report detailing the project and analyzing the Twitter dataset

Analytics platform

Self-directed group work on the final project.


Refining Tableau workbook including worksheets, dashboards, and story feature


Refining Gephi network visualisation


Writing final report

Final project report

(data analysis and visualisations and professional report)

- 25-page report

- Tableau Workbook with worksheets, dashboards, and story feature

- Gephi network analysis visualisations

*All activities are posted by research teams to the Moodle-based learning management system for review by the instructor.

Italics indicate components formally graded by the instructor.

26There are three instructional modules show in Table 2 : research methods and ethics, visualisation techniques, and network analysis and visualisation. Attached to each instructional module are detailed learning objectives, description of the modules, software tools, lab activity, and the form of evaluation. Across these modules, students produce an activity report each week with a self-reflective assessment (self-assessment) and exchange group feedback on what worked or did not work or new techniques and workflow processes that can be helpful for everyone in the course. These activity reports, which are posted to the Moodle-based learning management platform for the course, allow the instructor to contribute more informally to activities and support students’ learning process and outcomes. There are three assignments due, one in every module as listed in Table 2, which are graded assessments and formal opportunities for instructor feedback and guidance, and also an important means for assessing how engaged students are with the course material. The three modules enable experimentation and learning reflection for the students as the visualisation techniques and analysis grow more complex through the tutorials.

27Each of the three instructional modules offer different ways for students to learn about visualisation techniques and learn to use Tableau and Gephi. Each week, for the first nine weeks of the course, a tutorial is conducted by the instructor at the start of the lab and streamed to the large screens in the room. Course materials that support the tutorial are uploaded to the learning management platform and include written step-by-step instructions, annotated visuals created from Tableau and Gephi screenshots, and how-to videos recorded through screen capture. The technical instructions are supported by course materials that include theory and principles of social network analysis (Borgatti and Halgin, 2011), data ethics (Tufecki, 2014), platform studies approaches (Gillespie, 2010 ; Helmond, 2015), and research methods in communication and social media analysis (Williams, Burnap and Sloan, 2017 ; Beninger, 2017).

  • 5 Students do not have time to do in-depth content analysis by developing a coding scheme through whi (...)

28In the fourth module, students are required to document and explain research constraints or limitations, and the social, cultural, political, economic, and technological dimensions of their study, to construct an analysis about what the Twitter dataset tells them in terms of network structure, properties, diffusion of conversations, debates, and ideas, in relation to the context of their study5. Working with Tableau, students build worksheets with single visualisations, dashboards on which they can combine and annotate multiple visualisations, and a story feature through which students can combine multiple visualisations and dashboards with annotations within a “story” sequence structured through separate tabs to move through the visualisations. Importantly, students must do more than create data visualisations in Tableau and generate the network analysis in Gephi ; they must be able to explain and analyze the network’s features in the context of historical and social processes that give rise to the phenomena under study, and the cultural, political, and economic conditions that situate their object of study within the contemporary moment.

The platform approach : affordances and challenges

Affordances

29Why a platform approach? The research platform used in the VSM class is not the same as the social media platform on which Twitter account holders are active. Students in the course are not tweeting on the Twitter platform, they are using it as a data source to analyze particular conversations in the Twittersphere organized around a hashtag or keyword. The platform approach, however, is intended to show how different software, processes, practices, and people interact across a workflow that introduces non-technical students to a powerful set of professional data analysis skills. In a sense, the idea of a platform approach to social media analysis might be best described as a “platform-sensitive affordance” because it gives students a level of interoperability between distinctly different components (Burgess et.al., 2016 : 4). If we consider affordances “manifest in relations between platforms and their different types of users” (ibid.), then the platform approach leaves ample room for collaboration, participation, and engagement within experiential learning environments.

30A platform approach to teaching a social media analysis course is intended to emphasize a suite of software tools that include publicly accessible software such as Social Bearing that work well alongside, and coupled to, proprietary software that together can form a teaching and learning platform as an extended work canvas across which students can conduct in-depth social media analysis. The proprietary analytics software, Tableau, affords students the control to create their own dashboards and extend their understanding of how social data is visualised and interpreted using an industry-standard application. Students in experiential learning settings such as the VSM class must come to understand the value of the tool(s) they are learning, how the tool can be applied to specific tasks, and then be able to build a sufficient set of skills to meaningfully use the tool in an applied approach relevant to professional work (Winne, 2006). Experience in the VSM course so far suggests the platform approach supports the value/task relationship and using Twitter data to track an event, issue, or campaign through a keyword or hashtag makes the project meaningful and relevant.

31At the level of classroom study, taking a platform approach to social media analytics is practical and manageable. It conveys a sense of control for students over the process of data collection and analysis particularly as their proficiency with data analysis increases and they begin to feel adept at working on advanced visualisations. Arguably, the way we mobilize the platform approach diverges from the way researchers in platform studies define and explain computational platforms. We suggest a generous idea of platforms as adaptable technical ensembles, whereas Ian Bogost and Nick Montfort would argue that platforms must also be programmable (2009). However, the notion of a platform comprised of several components is what enables the workflow to take shape within the VSM course and provides a conceptually useful way of thinking about how multiple applications can be assembled to work together (Borra and Rieder, 2014).

Challenges

32The research platform, as noted in a preceding section, consists of the Twitter Archiver with a programmed connection to the public (free to access) Twitter API, which limits the data archiver to one per cent of all tweets on Twitter’s platform linked to the hashtag, issue, or keyword being tracked. Even with this limitation students collect a substantial amount of Twitter data for their project across four to five weeks of data collection. However, while this workflow is manageable in the VSM course, it remains one of the biggest challenges and students must explain this research limitation in their methodology because their Twitter dataset will never include all tweets on a topic organized around an issue, hashtag, or keyword.

33Working with or integrating datasets from other social media is difficult because of the proprietary nature of social media data on most platforms including from Facebook, Instagram, tumblr, and other social networking sites. Access to large proprietary commercial social datasets is out of reach for most academic researchers and students. Limited data archives for public social media platforms are sometimes shared between researchers, but the value in the VSM course is in learning how to collect and analyze real-time social data flows and Twitter remains the most accessible platform even with its API limitations.

34Integrating other analysis tools within the platform was proposed at the outset of the pilot project as an option to grow the analytics platform, but experience in the workshop class suggests this could lead to software overload (and see Asamoah et al., 2017). Developing professional skills in social media analysis using Tableau Software and Gephi meets the core goal of the course to build professional data analysis skills to foster critically inflected understanding through the mining, making, and meaning workflow enabled within the existing analytics platform. Expanding the analytics platform will be reassessed in the future.

35One of the challenges evident in the pilot project is the fascination students have for the software applications and analytical process. The analytics platform can easily become the sole focus in the course as students are enthusiastic about data visualisation, which brings a certain ‘cool factor’ to social media analysis. The final project works well to mitigate this effect and keep students focused on the requirement to make sense of the data and to be able to situate their dataset and findings within a conceptually grounded analysis that links theory to practice.

Learning Assessment

36There are challenges and affordances to the analytics platform designed and applied in the VSM course. However, recent end of semester course evaluations completed by students indicate that the course imparted learning materials and course content clearly and met course objectives and rated it at 4.83 out of 5. The instructor evaluation was 5.0 out of 5. The students provided substantive feedback on the quality of the workshop course that pointed to the segmented, modular structure, critical and analytical skills building, and job market ready professional capabilities :

It was a lot to take in and learn in a few months, but the way it was broken down and segmented made it manageable and much easier to enjoy learning the programs.

This workshop-style course engaged my critical/analytical skills in a way that is applicable to the real-world work environment.

Data mining and analytics have given me an edge in entering the job market with a new set of skill acquired in this class.

37For students, the VSM course expands their data literacy and gives them “an edge” as they graduate and enter the workforce to begin their professional life.

Conclusion

38The shift toward programmability, distributed across a range of applications to ensure cross-connectivity, enables complex software to interoperate with a more unified user experience as a platform, which is crucial for introductory courses designed for non-technical learners outside of specialized data science and computer science fields. Across the semester, students work toward this capability and they produce ethically informed, professional research reports with a rich set of data visualisations and network analysis on topics as diverse as climate change, federal political parties, student government, food waste, and Indigenous politics.

39The VSM course is a work in progress and so far, there have been five sections in the last 24 months. Future VSM courses may include an introductory lesson in R for advanced social media visualisation, augmented by an optional module on NVIVO’s proprietary software for organizing and managing research data, but as discussed above, expansion of the analytics platform must be considered with some caution6. Early feedback through course evaluations suggests students are very satisfied with the course. A supplemental survey may be conducted after the pilot project is completed in 2020 to ascertain whether or not students are employed, and if so, whether they use data visualisation and analysis in a professional setting. Wherever students go in their postgraduate life, they leave the VSM course with a set of skills and an understanding of what is entailed in social media analysis and how to make critically inflected decisions informed by ethically sound research practices.

Haut de page

Bibliographie

ANDREESSEN M. (2007), The three kinds of platforms you meet on the Internet [Blog post], <URL : https://pmarchive.com/three_kinds_of_platforms_you_meet_on_the_internet.html. >

ASAMOAH A., SHARDA R., ZADEH A. H., KALGOTRA P. (2017), “Preparing a Data Scientist : A Pedagogic Experience in Designing a Big Data Analytics Course”, Decision Sciences Journal of Innovative Education, vol. 15 (2), pp. 161–190.

ATWONG C. T. (2015), “A Social Media Practicum : An Action-Learning Approach to Social Media Marketing and Analytics,” Marketing Education Review, vol. 25 (1), pp. 27–31.

AZEVEDO R., HADWIN A. (2005), “Scaffolding self-regulated learning and metacognition : Implications for the design of computer-based scaffolds”, Instructional Science, vol. 33, pp. 367-379 <DOI 10.1007/s11251-005-1272-9>.

BENINGER K. (2017), “Social Media Users’ Views on the Ethics of Social Media Research”, In SLOAN L., QUAN-HAASE A. (ed.), The SAGE Handbook of Social Media Research Methods, London: Sage, pp. 57–73.

BIVENS R. (2017), “The gender binary will not be deprogrammed : Ten years of coding gender on Facebook,” New Media & Society, vol. 19 (6), pp. 880–898.

BOGOST I., MONTFORT N. (2009), “Platform Studies : Frequently Questioned Answers”, UC Irvine - Plenaries : After Media - Embodiment and Context, <URL : https://escholarship.org/uc/item/01r0k9br >

BORGATTI S., HALGIN D. (2011), “On Network Theory”, Organization Science, pp. 1–14.

BORRA E., RIEDER B. (2014), “Programmed method : developing a toolset for capturing and analyzing tweets,” Aslib Journal of Information Management, vol. 66 (3), pp. 262–278, <http://dx.doi.org/10.1108/AJIM-09-2013-0094>.

BOTIBOL A. (2016), “How data analysis became the new social media marketing”, Marketing Tech., October. <URL : http://www.marketingtechnews.net/news/2016/oct/11/how-data-analysis-became-new-social-media-marketing/> (accessed 01.11.2017).

BOYD D. (2010), “Social Network Sites as Networked Publics : Affordances, Dynamics, and Implications”, In PAPACHARISSI Z. (ed.), Networked Self : Identity, Community, and Culture on Social Network Sites, London : Routledge, pp. 39–58.

BOYD D., CRAWFORD K. (2012), “Critical questions for big data”, Information, Communication & Society, vol. 15, pp. 662–679 < DOI : 10.1080/1369118X.2012.678878>.

BRUNS A. (2016), Social Media Analytics Workshop, AoIR Berlin 5–8 October 2016, < https://aoir.org/aoir2016/preconference-workshops/>.

BUCHER T. (2012), “A technicity of attention : How software “makes sense”, Culture Machine, vol. 13, pp. 1–13. <URL : http://www.culturemachine.net/index.php/cm/article/viewArticle/470 >

BURGESS J., BRUNS A. (2015), “Easy Data, Hard Data : The Politics and Pragmatics of Twitter Research After the Computational Turn,” In BRUNS A., ELMER G., LANGLOIS G., REDDEN J. (ed.), Compromised Data : From Social Media to Big Data, London : Bloomsbury Publishing, pp. 93–111.

BURGESS J., BAYM N., BUCHER T., HELMOND A., JOHN N., NISSENBAUM A., CUNNINGHAM S., CRAIG D. (2016), Platform studies : the rules of engagement. Panel presented at AoIR 2016 : The 17th Annual Conference of the Association of Internet Researchers, Berlin, <URL : http://spir.aoir.org.>

CHAPMAN S., MCPHEE P., PROUDMAN B. (1995), “What is Experiential Education ?”, In WARREN K. (ed.), The Theory of Experiential Education, Dubuque : Kendall-Hunt Publishing Company, pp. 235–248.

DEWEY J. (1966), Democracy and Education : An introduction to the philosophy of education, New York : Free Press, 378 p.

FELDER R., BRENT B. (2009), “Active learning : an introduction”, ASQ Higher Education Brief, vol. 2 (4), <URL : http://www4.ncsu.edu/unity/lockers/users/f/felder/public/Papers/ALpaper (ASQ).pdf > (accessed 02.24.2018).

FLECK L. (1979), Genesis and development of a scientific fact, BRADLEY F., TRENN T. (trans.), Chicago : University of Chicago Press, 203 p.

FLORIDI L. (2011), The Ethics of Information, Oxford, UK : Oxford University Press, 384 p.

GASEVIC D., MIRRIAHI N., DAWSON, JOKSIMOVIC S. (2017), “Effects of instructional conditions and experience on the adoption of a learning tool,” Computers in Human Behavior, vol. 67, pp. 207–220.

GEBRE E., SAROYAN A., AULLS M. (2015), “Conceptions of Effective Teaching and Perceived Use of Computer Technologies in Active Learning Classrooms”, International Journal of Teaching and Learning in Higher Education, vol. 27 (2), pp. 204–220.

GERLITZ C., HELMOND A. (2013), “The Like economy : Social buttons and the data-intensive web,” New Media & Society, vol. 15 (8), pp. 1348–1365.

GILLESPIE T. (2010), “The politics of platforms”, New Media & Society, vol. 12 (3), pp. 347–364. <URL : https://doi.org/10.1177/1461444809342738 >

GRUZD A., HAYTHORNTHWAITE C., PAULIN D., GILBERT S., DEL VALLE M. E. (2016), “Uses and gratifications factors for social media use in teaching : instructors’ perspectives”, New Media & Society, vol. 20 (2), pp. 1–20, <DOI : 10.1177/1461444816662933>.

HELMOND A. (2015), “The Platformization of the Web : Making Web Data Platform Ready”, Social Media & Society, July-December, pp. 1–11.

JANETZCO J. (2017), “The role of APIs in data sampling from social media”, In SLOAN L., QUAN-HAASE A. (eds.), Social Media Research Methods, London : SAGE Publications, pp. 146–160.

JONES C. (2015), Networked Learning : An Educational Paradigm for the Age of Digital Networks, London : Springer, 249 p.

KOLB D. (1984), Experiential Learning : Experience as the Source of Learning and Development, Englewood Cliffs, N. J. : Prentice-Hall, 256 p.

LANGLOIS G., ELMER G., MCKELVEY F., DEVEREAUX Z. (2009), “Networked publics : The double articulation of code and politics on Facebook”, Canadian Journal of Communication, vol. 34 (3), <URL : https://www.cjc-online.ca/index.php/journal/article/viewArticle/2114 >

LANGLOIS G., REDDEN J., ELMER G. (2015), “Introduction : Compromised Data – form social media to big data”, In BRUNS A., ELMER G., LANGLOIS G., REDDEN J. (ed.), Compromised Data : From Social Media to Big Data, London : Bloomsbury Publishing, pp. 1–14.

MCGRATH A. L., FERNS A., GREINER L., WANAMAKER K., BROWN S. (2015), “Reducing Anxiety and Increasing Self-efficacy within an Advanced Graduate Psychology Statistics Course”, The Canadian Journal for the Scholarship of Teaching and Learning, vol. 6 (1), pp. 1–17, <DOI :10.5206/cjsotl-rcacea.2015.1.5>.

MONTFORT N., BOGOST I. (2009), Racing the beam : The Atari video computer system, Cambridge, MA : MIT Press.

NG W. (2015), “Adopting New Digital Technologies in Education : Professional Learning”, In WANG W. (ed), New Digital Technology in Education : Conceptualizing Professional Learning for Educators, Heidelberg & London : Springer, pp. 25–48.

PLANTIN J., LAGOZE C., EDWARDS P. N., SANDVIG C. (2018), “Infrastructure studies meet platform studies in the age of Google and Facebook”, New Media & Society, vol. 20 (1), pp. 293–310, <DOI : 10.1177/1461444816661553>.

ROEHL A., REDDY S. L., SHANNON G. J. (2013), “The flipped classroom : An opportunity to engage millennial students through active learning strategies”, Journal of Family & Consumer Sciences, vol. 105 (2), pp. 44–49, <DOI 10.1007/978-3-319-05822-1_7>.

SLOAN L., QUAN-HAASE A. (2017), Social Media Research Methods, London : SAGE Publications.

SOMMERVILLE I. (2013), “Teaching cloud computing : A software engineering perspective”, The Journal of Systems and Software, vol. 86, pp. 2330–2332.

SRNICEK N. (2016), Platform Capitalism, Cambridge and Boston : Polity.

TUFEKCI Z. (2014), "Big Questions for Social Media Big Data : Representativeness, Validity and Other Methodological Pitfalls", In ICWSM ’14 : Proceedings of the 8th International AAAI Conference on Weblogs and Social Media, 2014, <URL : https://arxiv.org/abs/1403.7400 >.

VAN DIJCK J. (2013), The culture of connectivity : A critical history of social media, New York, NY : Oxford University Press.

VYGOTSKY L. (1978), Mind in Society : The Development of Higher Psychological Processes, Cambridge, MA : Harvard University Press.

WELTEVREDE E., BORRA E. (2016), “Platform affordances and data practices : The value of dispute on Wikipedia,” Big Data & Society, vol. 3 (1). <DOI : 10.1177/2053951716653418 >

WHITE G. K. (2013), Digital fluency : skills necessary for the digital age”, Professional Educator, vol. 6, <URL : http://works.bepress.com/gerry_white/36> (Accessed 25.02.2018).

WHORTON R., CASILLAS A., OSWALD F. L., SHAW A. (2017), “Critical Skills for the 21st Century Workforce”, In BURRUS J., MATTERN K., NAEMI B. D., ROBERTS R. D. (ed.), Building Better Students : Preparation for the Workforce, Oxford, UK : Oxford University Press, pp. 48–72.

WILLIAMS M., BURNAP P., SLOAN L. (2017), “Towards an Ethical Framework for Publishing Twitter Data in Social Research : Taking into Account Users’ Views, Online Context and Algorithmic Estimation,” Sociology, vol. 51 (6), pp. 1149–1168.

WINNE P. H. (2006), “How Software Technologies Can Improve Research on Learning and Bolster School Reform”, Educational Psychologist, vol. 41 (1), pp. 5–17, <DOI : 10.1207/ s15326985ep4101_3 >.

WURDINGER K. (2005), Using Experiential Learning in the Classroom, Lanham : Scarecrow Education, 112 p.

ZUBOFF S. (2019), The Age of Surveillance Capitalism, New York : Public Affairs.

Haut de page

Notes

1 The curriculum design project is currently in the third and final year of a three year pilot project generously funded by the Faculty of Public Affairs at Carleton University, Ottawa, Canada.

2 See Tableau Software, <URL : https://www.tableau.com/> and Gephi – The Open Graph Viz Platform, > URL : https://gephi.org/ >, and, TAGS, < https://tags.hawksey.info/> or the Twitter Archiver from Amit Agarwal’s Google App development company, Digital Inspiration, <URL : https://digitalinspiration.com/product/twitter-archiver >.

3 Social Bearing < https://socialbearing.com/ >.

4 See Twitter.com Gnip service description : <URL : http://support.gnip.com/apis/>.

5 Students do not have time to do in-depth content analysis by developing a coding scheme through which to examine tweet texts and are limited to creating and studying data visualisations they create in Tableau such as “retweets” and measures of influence such as “followers”, and then creating network visualisations in Gephi.

6 The R Project for statistical computing and visualisation <URL : https://www.r-project.org/>. NVIVO is a commercial, fee-based research management application <URL : https://www.qsrinternational.com/nvivo/home>.

Haut de page

Table des illustrations

Titre Figure 1 : An active classroom at Carleton University
Crédits Image used by permission of Instructional Media Services at Carleton University.
URL http://journals.openedition.org/netcom/docannexe/image/4004/img-1.jpg
Fichier image/jpeg, 336k
Titre Figure 2 : Twelve-week course structure
URL http://journals.openedition.org/netcom/docannexe/image/4004/img-2.jpg
Fichier image/jpeg, 52k
Haut de page

Pour citer cet article

Référence papier

Sandra Robinson, « A platform approach to experiential learning »Netcom, 33-1/2 | -1, 101-124.

Référence électronique

Sandra Robinson, « A platform approach to experiential learning »Netcom [En ligne], 33-1/2 | 2019, mis en ligne le 09 septembre 2019, consulté le 16 avril 2024. URL : http://journals.openedition.org/netcom/4004 ; DOI : https://doi.org/10.4000/netcom.4004

Haut de page

Auteur

Sandra Robinson

The School of Journalism and Communication, Carleton University, Ottawa, Canada, sandra.robinson@carleton.ca

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search