Conference Agenda

Dashboards, progression curves, benchmarks, and traffic lights. All are examples of data visualizations used to mobilize data about educational institutions and their students. Data visualizations are graphic representations of digital data which summazise large amounts of data to patterns and trends within data sets. Data visualizations signpost the emergence of educational institutions and students as data objects, which can be observed and compared on a computer screen. They are thus shaping educational administrators’ and teachers’ socio-technical ways of ‘seeing’ educational quality and learning (Selwyn et al., 2022), and it is crucial to investigate their world making capacities and their ‘social life’ in educational worlds. As the main ‘interface’ through which educational administrators and educators access data, they are an underlooked but central aspect of the datafication of education.

This symposium investigates the role of data visualizations as a distinct way of making ‘education’ or ‘learning’ tangible and knowable. Although praised for making data accessible and interpretable, data visualizations also imply a distancing from data. Issues relating to how data is categorized in a database and how statistical techniques are performed on data are not included in visualizations (Ratner & Ruppert, 2019). The software developers of data visualizations make numerous design choices rendering some things absent and others present (Greller & Drachsler, 2012). While visualizations may appear factual and transparent, data visualizations provide neither direct nor neutral access to the object they are deemed to represent. Rather, they may be seen as persuasive and value-laden devices that privilege certain viewpoints (Latour, 1990).

This symposium examines data visualizations as entry point for discussing issues related to power, governance and automation. Dashboards visualizing the performance of educational institutions are today mundane artifacts in educational governance and require actors at different levels of governance hierarchies to turn performance gaps into improved outcomes (Decuypere & Landri, 2021; Ratner & Gad, 2018). Here, visualizations may have an affective dimension with e.g. rankings encouraging a dynamics of faming and shaming (Brøgger, 2016; Sellar, 2015), which, in turn, may situate education in a wider political context of competition and accountability. We may also examine questions of automation through data visualizations. With data visualizations increasingly presenting pre-fabricated interpretations of data, they now conduct some of the professional judgment formerly done by teachers (e.g. identifying low performing students). This may naturalize new forms of knowledge such as ‘at risk students’. It also maps out new responsibilities for teachers, such as ‘acting on’ visualizations to improve student learning. It is thus likely that visualizations both shape what counts as educational quality and signal to administrators and educators what they should prioritize. This raises important questions about how data visualizations reconfigure human judgment and decision-making in a digital and datafied age. While powerful, however, data visualizations can never fully determine the social contexts they are part of. Users may take them up in unanticipated ways. Thus, it is equally important to examine how educators and administrators make sense of data visualizations, ignore them, resist them or put them to other uses than those anticipated by the designers.

This conference symposium will explore the role of data visualizations in education across Europe and beyond. It does so by comparing different European and international cases of how data visualizations are used in education, including historical and contemporary examples. The symposium includes contributions examining both the production and consumption of data visualizations. Across the different contributions, it will also discuss conceptual and methodological questions arising from the study of educational data visualizations.

Recognizing that EdTech is increasingly shaping the teaching profession through the datafication and visualization of student learning, this paper advances an analytical framework for eliciting the “politics” of data visualizations. With inspiration from Science and Technology Studies (STS), well-suited for analyzing the co-constitution of technology and society, the paper suggests a framework for analyzing data visualizations’ ‘aesthetics’, ‘agencies’ and ‘alternatives’ as important if we are to understand their implications for teachers’ expertise. Rather than assuming the ‘technical’ and ‘social’ to be separate domains, STS invites an ‘infrastructural inversion’ (cf. Bowker & Star, 2000) where questions of politics, ethics, and knowledge are examined through infrastructural entities. Specifically, the notions of ‘aesthetics’, ‘agencies’ and ‘alternatives’ allow eliciting how the aesthetics of data visualizations also entail an interpretation of data, how visualizations configure expertise across human and machine agencies, and how visualizations themselves are contingent results of ongoing negotiations of their makers (Coopman, 2014; Ratner & Ruppert, 2019; Schaffer, 2017; Suchman, 2007). This approach thus casts light on the power of data visualizations as a device for shaping expertise but also appreciate them as cultural and social artifacts that could be otherwise. The paper demonstrates these analytics through a qualitative case study of a widely used digital mathematics platform for the primary and lower school (“folkeskole”) in Denmark, ‘MathTraining’. Launched as an adaptive and self-correcting platform in 2010, MathTraining today has become one of the most popular Danish platforms in mathematics. The analytical sections show, respectively, 1) how the aesthetics of data visualizations shape expertise by calculating and also interpreting student learning on behalf of the teacher; 2) how data visualizations configure expertise across human and machine agencies, automating student assessments and attracting teachers’ attention towards student engagement and progression, and 3) map out the alternative visualizations that never became part of the platform, demonstrating the contingent aspect of data visualizations, in terms of how different actors in the EdTech company have different ideas about how and which data should be visualized. Examining both the intentions inscribed into the visualizations as well as ongoing mundane negotiations about which data to visualize and how, allow us to better appreciate the normative dimensions of unsettled and ethical questions about the role of automated digital systems in education, including how they reconfigure teachers’ socio-technical way of ‘seeing’ and attending to learning.

Bowker, G., & Star, S. L. (2000). Sorting Things Out: Classification and Its Consequences. MIT Press. Coopmans, C. (2014). Visual analytics as artful revelation. In C. Coopmans, J. Vertesi, M. Lynch, & S. Woolgar (Eds.), Representation in Scientific Practice Revisited (pp. 37–59). MIT Press. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-151377 Ratner, H., & Ruppert, E. (2019). Producing and projecting data: Aesthetic practices of government data portals. Big Data & Society, 6(2), 2053951719853316. Schaffer, S. (2017). Introduction. In S. Schaffer, J. Tresch, & P. Gagliardi (Eds.), Aesthetics of Universal Knowledge. Palgrave Macmillan UK. Suchman, L. (2007). Human-Machine Reconfigurations: Plans and Situated Actions (2nd ed.). Cambridge University Press.

Education is, for many educational theorists, inherently uncertain, open-ended and risky (Biesta 2013, Allert et al. 2017). Yet the algorithmic systems of today‘s datafied world increasingly prioritise ‘certainty and the promise of guaranteed outcomes’ (Zuboff 2019: 497). It is widely known that AI-powered predictive systems have high margins of error. However, the data visualisations of algorithmic systems across health, social work, policing, government and other social fields tend to visualise certainty, thus invisibilising the underlying approximations and uncertainties of both the algorithmic systems and the social settings in which these systems operate. This can have harmful consequences for people, in particular minoritised populations. For example, while software providers and policy makers assure the public that algorithmic systems merely provide suggestions, and that users make the final decisions, research has shown that civil servants and other practitioners find it difficult to override algorithmic recommendations, even in their area of expertise (Eubanks 2019, Allhutter et al. 2020). Critical algorithm studies have hence raised questions about how (valid) knowledge is produced and circulates through algorithmic systems, and how truth claims are made (Jarke et al. forthcoming). In this paper, we argue that data visualisations and their production of certainty play a crucial role in devaluing situated knowledges that embrace the inherent uncertainty and open-endedness of our world. After introducing this field of research, we analyse a corpus of data visualisations from English- and German-language predictive analytics platforms for education. We explore the extent to which they show uncertainty. Findings indicate that data visualisations of uncertainty in education are exceedingly rare. The paper then discusses the implications when educators make decisions based on these visualisations. It reflects on the dashboard construction of ‘at-risk’ students (Jarke & Macgilchrist 2021), the distribution of benefits and harms to students, and the constitution of possible futures. It discusses three moves to contest the encoding of certainty into spaces of educational uncertainty: First, increased algorithmic literacy, which, however, individualises responsibility for action and transformation with the user. Second, artistic data visualisations which highlight uncertainty, which, however, tend to remain within the same frame in which data are collected about individuals. Third, then, the paper draws on a feminist/critical perspective to propose data visualisations of uncertainty that move beyond individualised data to show, for instance, structural inequalities, and that are embedded in collective (sociotechnical) practices. The paper concludes by identifying methodological challenges and open questions for future research.

Allert, H., Asmussen, M., & Richter, C. (2017). Formen von Subjektivierung und Unbestimmtheit im Umgang mit datengetriebenen Lerntechnologien – eine praxistheoretische Position. Zeitschrift für Erziehungswissenschaft, 21(1), 142-158. Allhutter, D., Cech, F., Fischer, F., Grill, G., & Mager, A. (2020). Algorithmic Profiling of Job Seekers in Austria: How Austerity Politics Are Made Effective. Frontiers in Big Data, 3. Biesta, G. (2013). The Beautiful Risk of Education. London: Paradigm Publishers. Eubanks, V. (2018). Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. New York: St. Martin's Press. Jarke, J., Prietl, B., Egbert, S., Boeva, Y., Heuer, H. (forthcoming). Algorithmic Regimes: Methods, Interactions, Politics. Amsterdam University Press Jarke, J., & Macgilchrist, F. (2021). Dashboard stories: How the narratives told by predictive analytics reconfigure roles, risk and sociality in education. Big Data & Society, 8(1), 1-15. https://doi.org/10.1177/20539517211025561 Zuboff, S. (2019). The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. London: Profile Books.

Writing from a new materialist perspective, Kosciejew (2017) proposes the concept of material-documentary literacy, reminding us that one of the main functions of documentation is to materialise information. He points out that ‘information’ is commonly regarded as being an abstract, dematerialised entity, distanced from its materiality which is regarded as secondary. In contrast, he foregrounds the materiality of documentation, to ‘help (re)configure our understanding of information, as something not immaterial and intangible, but something material and tangible’ (Kosciejew 2017: 97). Drawing on Suzanne Briet’s (1951) groundbreaking work on the nature of documentation, and his concept of informative material objects, this paper will examine datafication and data visualisation in higher education, avoiding the limitations of mainstream analyses in educational research so far. The concept of the informative material object allows us to analyse information and data visualiation as material phenomena which are embedded in specific sociomaterial instantiations and enmeshed with human agency, in contrast with the dominant paradigm of data and information being abstract, disembodied entities. This, I propose, is a subtle but important distinction which moves the focus onto the entanglement of human, material, digital and analogue agency which constitutes the ‘datafied’ university. I will examine a specific case of data visualisation via the production of representations of student engagement via ‘learning analytics dashboards’, a pedagogical practice which has been described in terms of tracing students’ ‘digital footprints’ (Sclater et al. 2016: 4). I will focus on the visual digital tracing of students, discussed via critiques of neoliberal uses of algorithms in society at large (e.g. Aneesh 2009, Beer 2019), of digital higher education (e.g. Prinsloo 2017, Jarke & Breiter 2019, Joksimović, Kovanović & Dawson 2019, Selwyn & Gasevi 2020), and of surveillance studies (e.g. Lyon 2018). I will argue that these critiques and theoretical resources, although invaluable, do not go far enough in their conception of data visualizations’ world-making capacities, in particular, in terms of their constitutive force, focusing particularly on student subjectivities in this case. My argument will be that the act of tracing undertaken via visual representation in learning analytics is an act of documentation in Briet’s terms, which fundamentally shifts how we might understand this educational practice, moving from a notion of surveillance towards a conception of ontological change – even violence – in which the student is rendered into a document. The implications for theory and practice will be discussed.

Aneesh, A. (2009). Global labour: algocratic modes of organization. Sociological Theory, 27(4), 347-370. https://doi.org/10.1111/j.1467-9558.2009.01352.x. Beer, D. 2019. The Data Gaze. London: SAGE. Beer, T. (Ed.) (2022). The Social Power of Algorithms. London: Taylor and Francis. Briet, S. (1951) 2006. What is Documentation? English Translation of the Classic French Text. Day, R., Martinet, L. and Anghelescu, H. (Eds.) Toronto: Scarecrow. Jarke, J. and Breiter, A. 2019. The datafication of education. Learning, Media and Technology 44(1), 106. Joksimović, S., Kovanović, V. and Dawson, S. 2019. The journey of learning analytics. HERDSA Review of Higher Education 6, 37-63. Kosciejew, M. (2017). A material-documentary literacy: documents, practices and the materialization of information. Minnesota Review 88, 96-111. Lyon, D. (2018). The Culture of Surveillance: Watching as a Way of Life. Cambridge: Polity. Prinsloo, P. (2017). Fleeing from Frankenstein’s monster and meeting Kafka on the way: algorithmic decision-making in higher education. Elearning and Digital Media 14(3), 138-163. Sclater, N., Peasgood, A. and Mullan, J. (2016). Learning Analytics in Higher Education: A Review of UK and International Practice. Bristol, UK: JISC. Selwyn, N. and Gasevi, D. 2020. The datafication of higher education: discussing the promises and problems. Teaching in Higher Education 25(4), 527-540.

Educational dashboards are increasingly prevalent, commended, and diverse visualising technologies that display outcomes of datafied educational processes to help students and pedagogical actors ‘keep track’ of learning pathways, alert for deviations, and make interventions, so students remain ‘on track’ (Gašević et al., 2022). Dashboards have long been connected to learning management systems and gained momentum as promising tools in learning analytics, an interdisciplinary field seeking to produce and deploy data-driven technologies and methods for improving education (Guzmán-Valenzuela et al., 2021). Descriptive and pre-emptive dashboards, for instance, are believed to enhance students’ self-monitoring and reduce dropouts as they allow students to reflect on their visualised learning (Safsouf et al., 2021). In response to the heightened attention for educational dashboards and the datafication of education generally, critical scholarship has investigated assumptions and consequences of data-driven technologies in education and called for research that details how such technologies engender (un)foreseen effects in situ (e.g., Jarke & Macgilchrist, 2021). With this contribution, we aim to scrutinise how educational actors relate in the production and deployment of higher distance education dashboards. Distance learning has a history of being organised through (digital) technologies, and existing issues intensified during the Covid-19 pandemic. As educational dashboards were used at universities to alleviate problems, research has predominantly focussed on (dis)advantages of dashboards for distance learning rather than their usage in distance learning (Celik et al., 2022). We examine dashboards from a Dutch university because they are telling cases about data visualisation for and in distance learning. Therefore, this article takes a science and technology studies (STS) approach to critically investigate modes of ordering and their effects (Law, 1994): specific ways of relating to visualising technologies situated in wider educational settings. We focussed on different relations with visuals, i.e., employees doing techno-scientific work to produce visuals and learners learning with the deployed visuals (Burri & Dumit, 2008). We followed insights from visual/digital ethnography during our fieldwork because the pandemic required us to pay close attention to the daily, technology-intensive practices of participants and ourselves (Pink, 2021). The results show modes of producing and learning with higher distance education dashboards in the Netherlands. The cases exemplify a ‘dashboarding of learning’ as well as a ‘learning to dashboard’, meaning that data visualisations enter educational practices and encourage – though not always with success – learners to understand and realise their education in close proximity to the underpinning techno-pedagogical ideas of production teams.

Burri, V., & Dumit, J. (2008). Social studies of scientific imaging and visualization. In J. Hackett, O. Amserdamska, M. Lynch, & J. Wajcman (Eds.), The handbook of science and technology studies (pp. 297–318). The MIT Press. Celik, I., Gedrimiene, E., Silvola, A., & Muukkonen, H. (2022). Response of learning analytics to online education challenges during pandemic: Opportunities and key examples in higher education. Policy Futures in Education, 1–18. Gašević, D., Tsai, S., & Drachsler, H. (2022). Learning analytics in higher education: Stakeholders, strategy and scale. The Internet and Higher Education, 52(1), 1–5. Guzmán-Valenzuela, C., Gómez-González, C., Rojas-Murphy Tagle, A., & Lorca-Vyhmeister, A. (2021). Learning analytics in higher education: A preponderance of analytics but very little learning? International Journal of Educational Technology in Higher Education, 18(1), 1–19. Jarke, J., & Macgilchrist, F. (2021). Dashboard stories: How narratives told by predictive analytics reconfigure roles, risk and sociality in education. Big Data and Society, 8(1). Law, J. (1994). Organizing modernity. Blackwell. Pink, S. (2021). Doing visual ethnography. SAGE Publications. Safsouf, Y., Mansouri, K., & Poirier, F. (2021). TaBAT: Design and experimentation of a learning analysis dashboard for teachers and learners. Journal of Information Technology Education: Research, 20, 331–350.