Final reflections

Looking back on my introductory post, my perspective has shifted hugely. I started out with a view of the power of data and the insights and details it could uncover. I struggled quite a lot during the course when understanding the many ways data is not as definitive as it appears: the fact that the act of measuring is a social process and is not neutral, and how data processes are reductive and can lead to decisions being based on partial views. This work threatened to derail my ambitions to work on personalised learning solutions, but I think it has given me a much more realistic viewpoint as a base.

Through this process I have understood the complexity and highly relational nature of data, and how it cannot ‘exist independently of ideas, techniques, technologies, systems, people and contexts’ Kitchen 2014a:24’ (Williamson, 2017, P30). I’ve appreciated that whilst data can uncover details, it should be used as a starting point to ask more questions in context rather than as a definitive answer.

Creating the visualisations

The process of collecting data and creating a visualisation showed me how limited the factors that data represent are. When creating my visualisations, I always wanted to add more detail, more context, more subcategories but the ease of understanding was the issue that stood in the way.

Creating the data visualisation takes planning in terms of the structure, the collection, the categorisation and how it will look visually to aid understanding. I believe my method of producing my visualisations improved over the course. In the first visualisation my focus was on representing data as simply as possible. I did not pre-determine categories before starting my visualisation, which meant I expanded those through the process. The result is a visualisation where the design hinders the understanding, highlighting the importance of the format of data in conveying the meaning.

Whilst there is still lots of room for improvement in terms of my design choices of my visualisations, I think I have shown progress across my three visualisations in presenting richer data as clearly as possible.

I gained understanding that the way data is represented is hugely important, despite the assumption data is inherently unambiguous. The connotations of certain structures or colours have fundamentally shaped the meaning of the visualisation through design choices. For example, in teaching with data, my visualisation made certain books appear much more successful as the most obvious factor was star rating, with categorisation of number of ratings less obvious and needing to look at the key to decipher. My later visualisation for governing with data was created with more design thought and intent; but this visualisation production showed me how even though more information was included and in a clearer way, the data is still imprecise as it is infused the with the ideology of what I wanted to portray.

Learning with data – key themes

My learning with data visualisation allowed me to interrogate digital behavioural traces and understand that the data is a proxy for behaviour but is not an actual representation and does not capture intent, reasoning and neglects human factors.

The way that data is categorised appears to work against autonomy and independence ‘seeking to intervene in educational conduct and shaping learner behaviour towards predefined aims’ (Knox et al, 2019). It assumes that individuals are uncomplicated and straightforward and can be understood as simplified archetypes. Data has the reputation of precision and accuracy of detail but in fact ignoring the contexts and elements that are incongruous with data structures reduces richness of meaning. Much like the emojis in my first visualisation, the simplification distorts intent from authentic expression to a model of best-fit.

Teaching with data – key themes

Teaching with data demands digital literacy and the ability to mediate practical knowledge with data (Anagnostopoulos, 2013). This exploration demonstrated to me the labour involved with teaching with data; the demands on other processes, the expectation to use technology and the pressures enacted on practical teaching expertise to include data as part of the process.

My visualisation process showed me that extent that the structure of platforms (and visualisations) hides or illuminates certain elements of the practice. This has huge influence on the information that is privileged in teaching discussions. The way in which certain information is privileged is routed in the neoliberalist focus on marketisation and measurement. Using data to optimise performance and evidence progress are political choices which fundamentally affect the act of teaching. The emphasis on showing where value is added is an example of ‘pedagogical reductionism’ (Williamson et al, 2020, P358) where only learning that can be measured is considered valuable. In this way the data elements which show success link with what wider society values in education.

Governing with data – key themes

The main learning from the governing with data block is the act of producing data makes it knowable and therefore able to intervene and manage. The choice to ‘datafy’ something is not a neutral or a straightforward translation of information, it is a social process which has consequences to what is being measured, changing it through this process. Creating data is an act of simplification, Selwyn warns against the ‘modelling’ of education through digital data and that fact it can create ‘algorithmically driven ‘systems thinking’ through which complex (and unsolvable) social problems associated with education can be seen as complex (but solvable) statistical problems’ (Ozga, 2016), when in fact the simplification changed the perception of the supposed problem initially, ignoring contexts, biases and other real-world issues.

Conclusion

Data require many compromises: the social change brought about by its measurement, the surveillance required and the labour-intensive processes from teachers and policy makers to make it useful. The output produced is simplified and privileges performance and outcome over human factors.

Whilst data must be perceived as a proxy representation of a richer contextual social event, it does uncover some information that was less available without the act of datafication and enables a different perspective. Data serve their purpose in looking at things differently, but it should not be mistaken for indisputable insight. Data require our participation in the process but space must be preserved to consider factors not easily quantifiable or represented by data.

A view of data will always be reductive in some sense, and it is irresponsible to expose people to datafied information with the promise that it is concrete fact, and this should be made explicit when working with data.

References

Anagnostopoulos, D., Rutledge, S.A. & Jacobsen, R. 2013. Introduction: Mapping the Information Infrastructure of Accountability. In, Anagnostopoulos, D., Rutledge, S.A. & Jacobsen, R. (Eds.) The Infrastructure of Accountability: Data use and the transformation of American education.

Lupi, G. 2016. Dear Data, ed. by Stephanie Posavec (London] UK: Particular Books,)

Williamson, B. 2017. Conceptualising Digital Data, in Big Data in Education: The Digital Future of Learning, Policy and Practice, (London: SAGE Publications)

Williamson, B. Bayne, S. Shay, S. 2020. The datafication of teaching in Higher Education: critical issues and perspectives. Teaching in Higher Education. 25(4), pp. 351-365.