Data infrastructuring in schools: New forms of professional edu-data expertise and agency

Abstract

The digitalisation and datafication of education has raised profound questions about the changing role of teachers’ educational expertise and agency, as automated processes, data-driven analytics and accountability regimes produce new forms of knowledge and governance. Increasingly, research is paying greater attention to the significant role of digital intermediaries, ‘in-between’ edtech or State authorities and the classroom itself, in educational transformations. School data infrastructures, understood as comprising diverse sociomaterial elements including teachers, data, software, standards and pedagogical practices, is one such intermediary through which teacher expertise and agency is reconfigured. In this paper, I focus on teachers’ involvement in processes of data infrastructuring in which people, platforms, systems and tools come together to create, enable and maintain data flows. Drawing on a sociomaterial ethnography of a secondary school in England, I analyse the work of a school data office in the behind-the-scenes work of data infrastructuring. The findings detail the significant labour and expertise involved in data infrastructuring, the dynamic, expanding and bespoke nature of the school data infrastructures that emerged, and processes of decontextualising and recontextualising numbers. The paper argues that the work of data infrastructuring undertaken by and through the school data office was an intermediary process which worked to both de-professionalise and re-professionalise teachers in new ways. In the process, this created new kinds of educational data experts and expertise, who gained significant influence and power within and beyond the school, both challenging and reinforcing existing organisational and governing power flows.

Keywords

Data infrastructure teacher professionalism teacher agency teacher expertise data labour digital intermediaries

Introduction

As has been well documented, schools have become sites of intensive datafication in recent years, with increasing volumes of performance and administrative data collected, processed, circulated and applied within and beyond individual schools, making up an extensive “data infrastructure” through which educational knowledge and practice are being profoundly reconfigured (Sellar, 2015). These trends are reshaping how schools, teachers and pupils have their behaviour and subjectivity increasingly governed and disciplined “by numbers”, including through State-driven performance accountability targets and commercial edtech platforms (for example, see Jarke and Breiter, 2019; Ozga, 2009; Souto-Otero and Beneito-Montagut, 2016; Williamson, 2015).

Of particular concern has been the effects on teachers’ professionalism, and in particular, the way that datafication has enabled a culture of performativity in which ‘good teaching’ becomes captured and governed by quantified measurement systems (Ball, 2003; Bradbury and Roberts-Holmes, 2017; Hardy and Lewis, 2016; Holloway and Brass, 2018). Performative accountability regimes and digital datafication have been seen as a threat to teachers’ professionalism, with educator’s professional judgement, expertise and agency potentially delegated to automated and datafied systems (Fenwick and Edwards, 2016).

So far, there has been less exploration of how teachers themselves actively participate in facilitating and mediating processes of educational datafication. Indeed, the work of education professionals’ data labour has often been overlooked, as attention has tended to concentrate on policy and commercial actors in this space. Teachers are often positioned as responding to rather than participating in datafication, and variously exhorted to either enthusiastically adopt or critically resist technologies of datafication, which appear to enter schools and take effect ‘from elsewhere’.

Increasingly, research is attending to the role of intermediaries, actors positioned ‘in-between’ policy-makers and schools, or big tech players and classroom practice (Hartong, 2016). Teachers themselves may act as digital intermediaries, facilitating and mediating digitalisation and datafication within their schools, creating bespoke data systems, brokering connections between platform providers and educational institutions, or translating policy into practice. In this paper, I focus on teachers’ involvement in practices of data infrastructuring in which people, platforms, systems and tools come together to create, enable and maintain data flows. I analyse the role of teachers in the behind-the-scenes work of data infrastructuring to address how teachers mediate and shape the flow, meanings and effects of data within their school. In particular, the paper addresses what the implications of teachers as data intermediaries might be for processes of de- and re-professionalisation; and how they might reconfigure, reinforce or reframe existing data governance power flows. More specifically, the paper addresses the following research questions: what new forms of teacher professionality are emerging around teachers’ roles as digital data intermediaries? And how are teachers’ professional expertise and agency being reshaped through digital data infrastructures?

Teacher professionalism and digital data

Exactly what constitutes ‘professionalism’ is contested, though generally refers to knowledge-based occupations requiring both higher education and vocational training (Evetts, 2013). Expertise and agency are key parts of professionalism: the knowledge and ability to act according to professional judgement. There are many definitions of what a professional teacher ought to be, defined through standards specified by national governing bodies, or as an individual’s personal virtues, behaviours and characteristics. These approaches, however, miss the materiality of professional practice, including how digital and data infrastructures might be reshaping professional agency and expertise. A sociomaterial approach acknowledges the materiality of professional practice and helps to analyse how teacher professionalism is not simply a question of an individual’s skills, but is a collective sociomaterial enactment resulting from the coming-together of many heterogenous human and non-human elements, including software and technologies, policies, actions, narratives and spaces (Fenwick, 2016). Similarly, agency and expertise, as core parts of professionalism, are not characteristics of individuals, but emerge through the relations between individuals, local practices and policies and other human and non-human elements (Biesta et al., 2015).

The digitalisation and datafication of education have raised profound questions about the changing role of teachers’ professional expertise and agency, as automated processes and data-driven analytics combine with intensified accountability regimes to produce new forms of knowledge and governance. Digital technologies and data analytics are powerful new actors in sociomaterial assemblages, transforming and giving rise to new kinds of digital education professionalities (Fenwick and Edwards, 2016; Hartong and Decuypere, 2023). Increasingly, teachers are expected to become data-literate professionals who apply data insights to their own teaching practice, while schools are encouraged to establish ‘data cultures’ in which data is applied to strategic decision-making (Lasater et al., 2020; Spina, 2020).

Datafication has also been integral to the rise of high-stakes accountability regimes in education, supporting the emergence of a ‘performative’ professional culture, in which producing the ‘right’ data becomes an aim in and of itself. (Ball, 2015; Grek, 2009; Moss, 2017; Ozga, 2016). These datafied and performative accountability regimes can be seen to be displacing possibilities for more ‘educative' practices and concerns and redefining teachers’ professional expertise and agency, even leading to de-professionalisation (Ball, 2003; Holloway and Brass, 2018; Lewis and Holloway, 2019; Manolev et al., 2019; Roberts-Holmes, 2014). Evetts (2011: 407) identifies a shift in notions of teacher professionalism associated with intensified accountability and governance, away from “partnership, collegiality, discretion and trust” towards “managerialism, bureaucracy, standardization, assessment and performance review”, which Anderson and Cohen (2015: 3) term a form of “de-professionalised professionalism”. Some teachers in recent studies have been described as internalising data targets, becoming disciplined as “marketized, managed and performative teachers” (Holloway and Brass, 2018: 18), or seeing data as a more professional source of knowledge than other professional sources of knowledge (Grant 2022). Becoming proficient with and engaging with data has come to be associated with being a ‘good’ teacher, “transforming teaching professionals into professors of data” (Lewis and Holloway, 2019: 49). Datafication and digitalisation can be seen as supporting performative accountability regimes that undermine and displace teachers’ professional expertise and agency.

Taking a sociomaterial perspective, however, draws attention to how assemblages of educational professionalism are changing with the entrance of new actors, including digital technologies and data, and giving rise to new education professionalities with multiple and varying effects (Fenwick, 2016; Hartong and Decuypere, 2023). Rather than straightforward accounts of professional decisions being delegated and automated to digital technologies, recent analyses show how some education professionals are able to exercise agency and expertise within education data assemblages. For example, school supervisors demonstrated a “reflexive engagement” in which they (re)contextualised data according to their prior conceptualisation of their professional roles (Dabisch, 2023: 59), ‘caring’ for data systems (Clutterbuck, 2023), or exercising agency as teacher ‘influencers’ in a new form of “postdigital teacher identity” (Arantes and Buchanan, 2022: 13). Rather than a straightforward account of digital and data governance de-professionalising teachers, then, it is possible to see how education professionalities are becoming at once fragmented, intensified and reconstituted, as a complex and ambivalent “matter of differential enactment” (Hartong and Decuypere, 2023: 6).

Teachers’ new roles: Data workers and digital intermediaries

While significant attention has been given to how teachers respond and react to processes of datafication, there has been less attention given to tracing the involvement of teachers and other education professionals in creating and shaping processes of datafication. In other words, how teachers themselves participate in ‘behind-the-scenes' education data work necessary to making data flow and take effect (Selwyn, 2021). Indeed, while edtech companies often promise digital solutions that bring automation and greater efficiency to routine administrative work, the introduction of new digital and data systems often results in the intensification of teachers’ labour, as well as that of students, parents, technicians and educational support professionals involved in the production and flow of data in and through schools (Perrotta et al., 2021; Selwyn, 2021).

Teachers in schools have diverse orientations to and engagements with data; they are not uniformly impacted by data in the same ways, and some teachers have much more agency in their engagements with data than others. For most teachers, producing and reporting data, and responding to data-generated reports is a routine part of their professional roles (Selwyn, 2021; Selwyn et al., 2022). Other, more ‘data-savvy’ teachers – those who have formal responsibility for, or a particular enthusiasm for education data – may develop a new form of outcomes-based professional identity (Anderson and Cohen, 2015: 3), or able to use their engagement with data as a way to extend their influence within and beyond their own schools (Arantes and Buchanan, 2022). Indeed, having a close professional association with digital data can be understood as a new form of power and influence within schooling (Hardy, 2015).

Tracing these different ways that teachers engage with data work shows how not all teachers are simply complying with, internalising or reacting to the imposition of datafication from external edtech companies or accountability regimes. As Arantes and Buchanan (2022: 2) note in their study of teacher influencers, they are not simply “puppets to commercial platforms”, but exercise their own agency within or through data logics. In this way, the shifts in educational expertise and agency associated with educational data cannot be understood as falling neatly into either performative or educative dispositions, or cleanly between commercial or pedagogical interests.

Teachers’ data work extends to the role of some teachers as digital data intermediaries, ‘in-between’ edtech companies, State authorities and schools and classrooms. Digital and datafied education governance does not arrive fully-formed from external policy instruments and software companies, but is facilitated and mediated by diverse intermediary actors, who consequently also shape the forms, practices, meanings and ‘effects’ of datafication that transpire in a particular context. Intermediaries may be organisations, infrastructures or individuals which work to facilitate and mediate linkages between the global and the local, between the State and schools, between edtech platforms and educational practice (Hartong, 2016). Understood in this way, software giants such as Google or accountability policies are part of a distributed network of organisations and actors governing schools through data, that may also include local education governance structures, professional development networks, technology brokers, private consultants, platforms, etc. (Fenwick et al., 2014; Gulson et al., 2022; Kerssens et al., 2022).

While much research exploring digital intermediaries has focused on actors and organisations outside schools, teachers themselves can also act as data intermediaries. From teacher influencers (Arantes and Buchanan, 2022) to private consultants (Williamson, 2016) to designated data managers (Selwyn et al., 2022), teachers play a crucial role in mediating and enabling data to enter and take effect within their schools. In many cases, schools do not simply adopt and apply off-the-shelf data products and solutions, but work with providers to customise the development of products and services (Gulson et al., 2022; Williamson, 2015). In this way, teachers can be considered “prosumers” of data – users of data who participate in the creation of personalised data products. More attention is still needed to understand teachers’ role as data intermediaries, and how they work to forge connections between diverse elements of digital networked governance (Williamson, 2015) and shape the forms of datafication emerging within their schools.

Data infrastructuring

Analysing teachers’ work as data intermediaries is therefore essential to understand how connections are made and mediated between different elements of a distributed network of educational governance that makes up an education data infrastructure. Such data infrastructures can be understood as helping to create the complex and potentially fragile assemblages that give rise to localised meanings and effects of data in particular contexts (Fenwick and Edwards, 2011; Hartong, 2021; Piattoeva and Boden, 2020). Data infrastructures are not only technical organisations of software but also include the social subjects and practices that enable data practices (Decuypere, 2021; Star, 1999). Infrastructures are dynamic, relational and emergent; that is, they are constituted of shifting relations between diverse elements (Gulson and Sellar, 2019; Hartong, 2021; Piattoeva and Saari, 2018). It is important, therefore, to see data infrastructuring as an ongoing process rather than as a one-time achievement, requiring care and labour to bring and hold together heterogeneous elements. The ongoing process of data infrastructuring works to create new spaces of educational governance, bringing in new actors through processes that enable standardization and interoperability between systems and scales (Gulson et al., 2022; Lewis and Hartong, 2022). In exploring the creation and maintenance of education data infrastructures, then, we need to unpick how different elements are brought together and what mechanisms allow them to cohere or become destabilized (Lewis and Hartong, 2022), including the role of teachers as data intermediaries.

Understanding teachers as active intermediaries and participants in the work of creating and sustaining data infrastructures further complicates the notion that teachers are either internalizing or resisting performative data regimes imposed from elsewhere. Rather, they may be taking on much more active and agentive roles in mediating between and co-creating infrastructures of education data governance. Data – whether required by State accountability policies or generated via edtech software platforms – does not just act upon schools and teachers; teachers must work to integrate school practice into these wider data infrastructures. In the process, teachers themselves become part of data infrastructures, entangled in the relations that they are actively engaged in performing. From this perspective, we need to ask how teachers’ professional agency and expertise may be performed in and through their role as intermediaries in emerging education data infrastructures.

Research context and methods

The analysis in this paper is drawn from a sociomaterial ethnography of a secondary school in England in which I spent three periods in a large, comprehensive, suburban secondary (age 11-16) school, over a single school year. The school, while not particularly atypical, is presented here as an entry point into network of wider digital and data infrastructures and a point at which multiple relations intersect (Nespor, 1997; Sellar, 2015). I was based in the school’s data office, following pupil performance and other forms of education data as it was generated, processed, circulated and acted upon. Consistent with the sociomaterial framing of this study, the data generation and analysis approach pays attention to the material as well as the social and discursive elements of the research setting, drawing on ethnographic observations of the spatial, material and discursive elements of data practices, in-depth interviews with the data office team, interviews and observations with classroom teachers, collected documents and artefacts including reports, digital and physical displays and software outputs.

I was based in the school’s ‘data office’, led by Sarah, the Head of Improving Achievement, Chris, the Data Manager, and Jenny, the data administrator.¹ Sarah, Chris and Jenny were responsible for establishing, maintaining and facilitating the operation of data systems within the school. Sarah and Chris were also mathematics teachers. Being based in the data office allowed me to be present for casual conversations and formal meetings with other school colleagues; to understand the rhythms, routines and preoccupations of those teachers tasked with managing school data systems, which then informed the more in-depth interviews I conducted with them. I also conducted interviews with two classroom teachers, Joe and Sophie, as they talked me through their data reporting processes and explained their classroom resources and data records. These conversations allowed a glimpse of data practices of teachers less closely associated with data in the school, which provides some insight into the variations in changing teacher professionalism in relation to digital data practices. By focusing predominantly on the data office teachers’ practices this paper privileges how new forms of professionalism are emerging around those educators most closely entangled in digital data assemblages and makes more tentative claims about how processes of de- and re-professionalisation may apply to other teachers.

Analysis was approached from a sociomaterial perspective, drawing on ‘after-Actor-Network Theory’ approaches as sensitizing concepts to recognize the materiality (including digital and data infrastructures) and relationality of both teacher professionalism and the research process (Fenwick et al., 2011). This involved tracing the technical and material components of the education data assemblage, as well as the people and practices engaged in data processes, and employing ethnographic tactics to attend to the cultural features of sociotechnical systems (Seaver, 2017; Star, 1999). Research data is seen as one element in a relational assemblage of knowledge production, that also includes theoretical resources and the researcher’s own embodied experiences, involving, in Jackson and Mazzei’s terms, “using theory to think with data and data to think with theory” (2012: 261) to create new knowledge. The current analysis involved a return to the original data collected, ‘plugged in’ to new theoretical concepts of data infrastructuring, digital intermediaries and teacher professionalism.

Findings

Intensification of teachers’ data labour: technical, social and relational

My time spent in the data office showed a significant amount of “invisible work” that went into devising, establishing and maintaining educational data systems (Star, 1999: 377). Despite the use of at least two commercial data management software platforms, direct data work took up a lot of staff time, including the creation and operation of a bespoke pupil performance monitoring and intervention system. This system was based on frequent, formal pupil assessment, with ‘data drops' reporting grades for every pupil in every subject and sent to the data office six times per year, where it was collated, processed and circulated in various reports, and displayed in the data office. This wall-sized display comprised rows of printed postcards, each including pupils’ names and photographs alongside snippets of data. This data wall, Sarah explained, was the result of a bespoke algorithm designed by the data office teachers themselves, which ranked pupils by their relative priority to be assigned to “intervention classes” designed to boost their performance. Each data drop then triggered a further cascade of tasks for the data office team and senior leadership staff. Indeed, there was a flow chart drawn on one of the many whiteboards on the walls of the data office reminding everyone of the sequence of tasks throughout the data drop cycle and those responsible for them.

Even this flow chart smoothed out the messier reality of the labour involved in making these data systems work. At each data drop, some teachers were late with data entry and had to be chased, missing data had to be accounted for in reports, and the data office teachers continually worked to persuade colleagues to engage with and respond to the resulting reports. There was also a significant amount of manual labour involved in maintaining data systems, involving mundane technologies of spreadsheets, phone calls, handwritten notes and printed documents. A full-time administrator was employed in the data office to carry out essential data work, including data entry and data cleaning, maintaining pupil records, inputting handwritten data received from teachers and parents via notes or phone calls, transferring and translating data between different platforms, compiling reports from information across different databases and printing and circulating paperwork to teachers, pupils and parents.

The reality of the data infrastructure in Ridgewood School was thus a long way from the vision of fully automated data flows outsourced to commercial platforms.. Rather, the data office team dedicated significant amounts of human, manual and relational labour, care and attention to maintaining the infrastructure to make data ‘work’, intensifying rather than reducing human labour (Piattoeva and Saari, 2018; Selwyn, 2021).

Bespoke and shifting data infrastructures

The data drop system was not an off-the-shelf commercial solution but had been developed in-house by the data office teachers. In order to make data actionable in the ways they wanted, the data office team had devised their own algorithm, determining pupils’ eligibility for intervention classes. Similarly, the visualisation of pupils’ ranking in the data postcard wall had been devised by the data office teachers as a way of making pupils’ status highly visible to themselves and other teachers. Sarah explained how creating and applying their own algorithm meant they were ready to adapt to changing school targets and priorities by adjusting data weightings and input measures if necessary to produce different priority pupils.

As well as positioning their data systems as able to respond to changing priorities, the data office team also critically reflected on the operation and effects of their data systems and made dynamic adjustments in response to concerns they identified. For example, Sarah and a visiting head teacher from another school reflected on the “optimum frequency” for assessing pupils that balanced their need for up-to-date pupil performance data with considerations of how quickly changes in pupils’ learning might be expected to show up in their assessment data, and the burden of assessment on pupils and teachers. Similarly, Joe, an English teacher who worked closely with the data office critically reflected on how he did not wish pupils to “think of themselves as a number on a piece of paper” as this could discourage them from engaging with feedback that he felt would help them improve. To address this concern, he went as far as initiating an action research project in which pupils only received qualitative, formative feedback. The visiting head teacher and Joe were closely associated with the data office and shared similar perspectives on the value of a data-driven approach to schooling. Their position as ‘insiders’ appeared to enable open and somewhat critical discussions of the limitations, difficulties, and possible alternative approaches to data systems.

In contrast to critiques of delegated and automated decision-making, where users are unaware of how decisions are derived from data, the data office team took great pride in explaining to other teachers how and why their system worked in the way that it did. By adopting a bespoke approach to data infrastructuring, the data office team had greater control over how data flowed, what it meant, and what it made actionable, and were able to dynamically adapt data systems to respond to changing priorities and educational concerns. They were also, at least to some extent, able to engage with and adapt in response to critiques of the system. Rather than customising commercial platforms as “prosumers”, (Hartong, 2016; Williamson, 2016), here, the data office team have even more direct control over data systems and may be better understood as educational data intermediaries in their own right – devising and dynamically updating data infrastructures in response to changing educational priorities and concerns.

Extending and expanding data infrastructures

As well as devising, maintaining and adapting existing data infrastructures, the data office team were actively expanding the extent and remit of data systems throughout the school. For example, pupils’ future performance forecasts (an input measure to the pupil data wall) were generated through teachers’ subjective processes of prediction rather than a standardised data-driven analysis. In response, the data office required teachers to show how their forecasts were derived from pupils’ most recent assessment data, in conjunction with patterns of previous cohorts’ progression. In this way, what had been a matter of teacher professional judgement became a matter of infrastructural process which could be both standardised and auditable. A further example of data infrastructure expansion was expressed by attempts to create a ‘closed loop’ model that encompassed a full cycle of data analysis and intervention, in order to evaluate the impact of interventions on pupils. A final aspect of data infrastructure expansion was the move to develop entirely new data systems to bring new areas of school life into the data infrastructure, including systems for monitoring staff absences and processes for assigning pupils to optional courses in their final years.

Data infrastructures can never encompass or anticipate every question or decision that arises, as the complexity of social life necessarily exceeds the systems that attempt to capture it (Piattoeva and Boden, 2020; Piattoeva and Saari, 2018). At the same time, the critiques, loopholes and opportunities that exceed the data infrastructure act as triggers for the continued expansion of the scope and remit of data systems in an “infinitely expandable” data infrastructure (Gulson et al., 2022: n.p.). In this way, the data infrastructure constantly adapts and expands.

The work of extending and expanding data infrastructures to capture an increasing volume of school life was crucial to how the data office teachers saw their role. Sarah, the Head of Improving Achievement, commented,

“how I see the work of this office is, actually, we find a problem, something that’s not being done very well, we find a process and a system to make it be done better, we give that system back to that person and that person then does that better”.

In this way, they positioned themselves, not just as data analysts, but as data intermediaries, engaged in data systems development to bring greater efficiency and accuracy to a wide range of educational and institutional problems. To achieve this required more than the application of data analysis skills to educational questions, it required bringing together their professional understanding of education and schooling as teachers in conjunction with data skills and logics to forge a new form of edu-data-systems professional expertise.

Expanding the application of data infrastructures through this systems-engineering approach to education also expanded the sphere of the data office teachers’ agency, influence and power. Through their work as data intermediaries and edu-data systems engineers, the data office defined how educational activities could be measured, represented and valued, and in this way became increasingly powerful players, not just through the technical operation of data systems or the application of data analytic techniques, but taking control of ever more areas of school life by bringing them in to data governance infrastructures.

Making and re-making data relations

Data infrastructuring works to bring data into relation with other elements; it is these relations that give meaning and effects to data rather than the data ‘itself’ (Hartong, 2021; Lewis and Hartong, 2022). The relations wrought through the data office teachers’ work actively shaped and re-shaped the specific meanings of the pupil performance data they worked with, with some data taking on quite different meanings at different points and in different sets of relationships. For example, one key data point in the data drop system was a pupil’s Minimum Expected Grade (MEG). This was an individual pupil’s target grade, calculated as a fixed amount of progress made over a pupil’s secondary school career. The percentage of pupils achieving MEGs was an important accountability measure, and so pupils’ progress data was closely monitored using spreadsheets which automatically colour coded data to show whether or not pupils were “on track”. In my interviews with classroom teachers Joe and Sophie, they were critical of the validity of MEGs, noting that not all children progressed at the same rate and that prior assessments were not necessarily reliable. Yet, when reports of children at risk of missing targets based on this data were created, Sophie and Joe both accepted the need to act upon this data. Similarly, the data drop system and the postcard wall itself were further examples of relations within the data assemblage creating particular meanings and effects from data. Pupil performance data had a provisional meaning at the time of the assessment, then it was stripped of its originating context as a single number was entered into monitoring and tracking spreadsheets. Then, context was added back in via the addition of further data, including accountability targets, demographic data, school priorities, intervention classes, pupil photographs, to produce certain pupils as a ‘priority’ for intervention.

Perhaps surprisingly, the data office teachers were also able to recontextualise the use of MEGs by forging a different set of data relations to show how they could be seen as an invalid accountability target. Faced with a potentially negative judgement from a school inspector, the data office teachers obtained and analysed national pupil progress data to show that ‘expected’ progress targets for pupils with lower grades on entry to secondary school was the exception rather than the norm (Education Datalab, 2015); and that therefore a school like theirs, with a high proportion of lower-attaining pupils on entry, was statistically unlikely to meet this accountability measure, and the targets could therefore hardly be termed ‘expected’. The data office team were successful in using this relational recontextualization to change the meaning of pupil progress data and secure a positive inspection result. Pupil data can, in this way, be seen as internally inconsistent – whether progress is ‘as expected’ or ‘on track’ depends on the specific relations it is brought into, and thereby able to “travel routes unforeseen” (Piattoeva 2021). In this way, the data infrastructuring work of the data office teachers was an act of de- and re-contextualisation in which data can come to mean and do new and different things, through cutting away relations with the complex world and forging new relations, including with other data, systems and practices to make data mean and do different things (Piattoeva, 2021)

Discussion

Emerging edu-data-systems expertise and de/re-professionalism

The data office teachers’ work as intermediaries, involved in building, expanding and adapting data infrastructures that shaped data flows in their school, goes beyond adding a layer of technical data expertise over an educational context. As Sarah said, “I don’t think it’s a coincidence that we’re both maths teachers”, referring to both their statistical skills and their teaching experience. Data infrastructures do not simply arrive in schools ‘ready to go’, but require significant care, attention and labour to work and take effects (Lewis and Hartong, 2022; Piattoeva and Saari, 2022; Selwyn, 2021; Star, 1999). The data office teachers invested substantial technical, administrative and relational labour in devising, maintaining, expanding, critiquing, experimenting with and adapting data systems, as they attempted to integrate their educational concerns and understandings into data infrastructuring practices. Rather than delegating educational decisions to software platforms or performatively complying with accountability metrics, this was an attempt to integrate, negotiate and find ways to work within wider regimes of data governance.

By integrating their understanding of schooling with their enthusiasm and skills in data and systems engineers, they developed a new kind of education professionalism – as education-data-systems experts, able to exercise agency as intermediaries in the flow, meaning and effects of data within both the local context of their school and beyond. Sarah recognized how their specific combination of educational and technical expertise might also enable them to act in a more entrepreneurial fashion, training other schools to make use of their data systems. As Sarah said,

“I do actually believe that you could sell all these programs as a package, but I would never sell them to a school that wasn’t prepared to allow us to train them, because I don’t want this to be abused – you can’t buy it unless you’re going to use it properly, y’know?”

Sarah understood the value of their bespoke data systems as necessarily accompanied by training to ensure it is embedded with an educational understanding. In this way, Sarah and Chris position themselves as a new kind of entrepreneurial educational professional, helping schools make data work for them.

This is not to suggest that processes of datafication at work operated entirely in the service of supporting teachers’ professional expertise and agency. Indeed, the data practices do seem to have de-professionalised some teachers by locating legitimate educational judgements primarily within data (Grant, 2022). The research presented here concentrates on the data office teachers, and there is less evidence of the experiences of teachers outside their ambit. However, Sophie, one of the classroom teachers interviewed, described how data collection and reporting were simply an accepted and routine part of her role that she gave little further thought to; consistent with other studies that found that most teachers are ambivalent towards or uninterested in data (Selwyn et al., 2022; Star, 1999). Whilst the data office teachers and their closely associated colleagues were able to critique the validity of data and the impacts of data practices on pupils and teachers, they dismissed questioning and resistance from other teachers as demonstrating limited appreciation and understanding of data, suggesting that those teachers may have been de-professionalised even as those more associated with data were developing new forms of professionalism. In this way, new divides around edu-data expertise can be seen to emerge, between the ‘data do-ers’ and the ‘data done-to’ (Selwyn et al., 2015), with some teachers becoming de-professionalised as more ‘holistic’ forms of educational knowledge are devalued, while at the same time other teachers benefit from emerging new forms of edu-data professionalism.

Teachers’ reproduction, reconfiguration, and resistance of data governance

The data office teachers’ work in data infrastructuring enabled them to exercise a significant amount of power, agency and influence throughout their school. Positioning themselves not just as data analysts, but as educational systems engineers, they developed, maintained and expanded data infrastructures in ways that shaped increasing domains of their colleagues’ work. This expansion encompassed teachers’ forecasting practices, evaluation of intervention outcomes, monitoring staff absence and entrepreneurial activity, and gave the data office teachers a central and increasing influence over colleagues within and beyond their own school. This power was recognised through the resources and status allocated to them, including dedicated office space, a full-time administrative support staff member, and positions on senior management and leadership committees. The data office teachers’ close association with, control over, and their position as intermediaries within data infrastructures gave them agency and power within their own school (Hardy, 2015) as well as positioning them as influencers in wider networks through networked and entrepreneurial activity (Arantes and Buchanan, 2022).

The data drop system devised and implemented by the data office teachers was a way of enacting networked data governance in the local school context. National accountability targets such as the percentage of pupils achieving MEGs were core elements in data wall algorithm, and the emphasis on anticipating and intervening in pupils’ future performance was a direct response to the high stakes of these measures in inspections. In these ways, the data office teachers and systems worked as intermediaries to connect State targets with school practice. Yet these data practices were not just a conduit through which national accountability targets could enter and take effect within the school. These flows of data power had to be made to work, with the local meanings and effects of data substantially shaped by the systems the data office teachers established. The data office teachers and their systems made data take effects in specific ways, through devising and adapting bespoke data infrastructures, dealing with exceptions, raising and responding to critiques, in ways that enabled them to enact professional agency and apply their own combination of edu-data expertise. Rather than simply being agents of an externally imposed performative accountability regime or technology companies (Arantes and Buchanan, 2022), the data office teachers were able to enact agency within this wider data governance assemblage, finding ways to make data work and make sense within their local context.

While the data office teachers did not, on the whole, exercise their agency for wholesale resistance or critique of data governance regimes, there were moments when they were able to use their education data expertise to challenge the operation of performative data power, as when they used their own data analysis to successfully challenge a school inspector on ‘expected’ progress targets. Again, this was not a comprehensive rejection of data governance per se, but, from their position as education data experts within the wider data assemblage, they were able to exercise professional responsibility and agency to further the interests of their school, by becoming attuned to the sociomaterial relations in which they were involved (Fenwick, 2016). Of course, the ability to challenge data power in this way arises precisely through the data office teachers’ privileged position as data intermediaries and their combined data and educational expertise within these governing systems; challenge was only possible because it did not reject or problematise governing data infrastructures in their entirety. As intermediaries, the data office teachers were able to work within the wider data assemblage to both critique and operationalise data governance regimes. Analysing their role challenges a sharp distinction between compliance with and internalisation of performative data logics on the one hand, and on the other, a fully educative form of professional agency and expertise that critiques and rejects data (Hardy and Lewis, 2016). As educational data intermediaries, the data office teachers appear to be, at one and the same time, both operationalising a performative accountability regime, while at the same time critically reflecting upon it and attempting to adapt it to address more educative concerns.

Conclusion

Through their work as data intermediaries, intimately involved in data infrastructuring, the data office teachers in this study developed a new form of education-data-systems professionality, which integrated their understanding and experience of education and schooling with data logics and systems engineering approaches. This was, of course, an uneven pattern: the professional data assemblages of which the data office teachers were part, worked to de-professionalise some teachers, but at the same time supported a re-professionalisation of those teachers who were able to develop new forms of expertise and agency in relation to data. This expertise, alongside their position as data intermediaries, gave the data office teachers substantial agency and influence within and beyond their school, through monitoring, evaluating and directing colleagues’ work and through the entrepreneurial extension of their edu-data infrastructures. Their operationalisation of data governance regimes through devising and expanding data infrastructures further reinforced existing organisational and governing power flows.

The data office teachers were not just conduits of governmental data power. They also exercised substantial agency through data infrastructuring, shaping the scope, extent and remit of data systems within their schools and making data work in their own school context. Their professional educational and data expertise also enabled them to critique data governance infrastructures from within, problematising some aspects of datafication including the burden of assessment, the validity of data and unrealistic accountability targets. These were limited challenges that sought to make data work in ways that worked in the interests of school leadership and management rather than a wholesale rejection of governing data power or resistance of the performative effects of datafication. Teachers, even as data intermediaries or edu-data-systems professionals, have little scope to be able to resist such forces. However, by focusing on the active role of the data office teachers as data intermediaries, a more complex intertwining of critical, educative and performative data governance emerges.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Lyndsay Grant

Note

References

Anderson

Cohen

(2015) Redesigning the identities of teachers and leaders: a framework for studying new professionalism and educator resistance. Education Policy Analysis Archives 23: 85.

Arantes

Buchanan

(2022) Educational data advocates: emerging forms of teacher agency in postdigital classrooms. Learning Media and Technology 48: 1–21.

Ball

(2003) The teacher’s soul and the terrors of performativity. Journal of Education Policy 18(2): 215–228.

Ball

(2015) Education, governance and the tyranny of numbers. Journal of Education Policy 30(3): 299–301.

Biesta

Priestley

Robinson

(2015) The role of beliefs in teacher agency. Teachers and Teaching 21: 624–640.

Bradbury

Roberts-Holmes

(2017) Creating an Ofsted story: the role of early years assessment data in schools’ narratives of progress. British Journal of Sociology of Education 38: 943–955.

Clutterbuck

(2023) The role of platforms in diffracting education professionalities. Tertium Comparationis 29(1): 73–92.

Dabisch

(2023) The practices of data-based governance: German school supervision, professionalism and datafied structurations. Tertium Comparationis 29(1): 48–72.

Decuypere

(2021) The topologies of data practices: a methodological introduction. Journal of New Approaches in Educational Research 10(1): 67–84.

10.

Education Datalab (2015) Seven Things you Might Not Know About Our Schools. London, UK: Education Datalab. Available at: https://educationdatalab.org.uk/2015/03/seven-things-you-might-not-know-about-our-schools/.

11.

Evetts

(2011) A new professionalism? Challenges and opportunities. Current Sociology 59(4): 406–422.

12.

Evetts

(2013) Professionalism: value and ideology. Current Sociology 61(5–6): 778–796.

13.

Fenwick

(2016) Professional Responsibility and Professionalism: A Sociomaterial Examination. Abingdon: Routledge. Available at: https://www.google.co.uk/books/edition/_/KEZ-CwAAQBAJ?hl=en&gbpv=1&pg=PA1973.

14.

Fenwick

Edwards

(2011) Considering materiality in educational policy: messy objects and multiple reals. Educational Theory 61(6): 709–726.

15.

Fenwick

Edwards

(2016) Exploring the impact of digital technologies on professional responsibilities and education. European Educational Research Journal 15: 117–131.

16.

Fenwick

Edwards

Sawchuck

(2011) Emerging Approaches to Educational Research: Tracing the Sociomaterial. London, UK: Routledge.

17.

Fenwick

Mangez

Ozga

(2014) Governing knowledge: comparison, knowledge-based technologies and expertise in the regulation of education. In: Fenwick

Mangez

Ozga

(eds) World Yearbook of Education 2014. London, UK: Routledge, 3–10.

18.

Grant

(2022) Reconfiguring education through data: how data practices reconfigure teacher professionalism and curriculum. In: Hepp

Jarke

Kramp

(eds) New Perspectives in Critical Data Studies: The Ambivalences of Data Power. Cham, Switzerland: Palgrave Macmillan, 217–239. https://link.springer.com/chapter/10.1007/978-3-030-96180-0_10.

19.

Grek

(2009) Governing by numbers: the PISA ‘effect’ in Europe. Journal of Education Policy 24(1): 23–37.

20.

Gulson

Sellar

(2019) Emerging data infrastructures and the new topologies of education policy. Environment and Planning D: Society and Space 37: 350–366.

21.

Gulson

Sellar

Webb

(2022) Infrastructure: interoperability, datafication, and extrastatecraft. In: Algorithms of Education How Datafication and Artificial Intelligence Shape Policy. Minneapolis, USA: University of Minnesota Press. Available at: https://manifold.umn.edu/read/algorithms-of-education/section/2687a46c-8ff6-4f28-832e-16763af3ca52#ch04 (accessed 7 September 2022).

22.

Hardy

(2015) Data, numbers and accountability: the complexity, nature and effects of data use in schools. British Journal of Educational Studies 63: 467–486.

23.

Hardy

Lewis

(2016) The ‘doublethink’ of data: educational performativity and the field of schooling practices. British Journal of Sociology of Education 38: 671–685.

24.

Hartong

(2016) Between assessments, digital technologies and big data: the growing influence of ‘hidden’ data mediators in education. European Educational Research Journal 15(5): 523–536.

25.

Hartong

(2021) The power of relation-making: insights into the production and operation of digital school performance platforms in the US. Critical Studies in Education 62(1): 34–49.

26.

Hartong

Decuypere

(2023) Platformed professional(itie)s and the ongoing digital transformation of education. Tertium Comparationis 29(1): 1–21.

27.

Holloway

Brass

(2018) Making accountable teachers: the terrors and pleasures of performativity. Journal of Education Policy 33(3): 361–382.

28.

Jackson

Mazzei

(2012) Thinking with Theory in Qualitative Research: Viewing Data across Multiple Perspectives. London, UK: Routledge.

29.

Jarke

Breiter

(2019) Editorial: the datafication of education. Learning, Media and Technology 44(1): 1–6.

30.

Kerssens

Pangrazio

Nichols

(2022) How Google took over the classroom: the role of intermediaries in the googlization of three education systems. Digital Education Governance beyond International Comparative Assessments: Complex Histories, Contested Presents, and Contingent Futures. Edinburgh, Scotland: The University of Edinburgh. Available at: https://media.ed.ac.uk/playlist/dedicated/79280571/1_kcusxmak/1_33jfm915.

31.

Lasater

Albiladi

Davis

, et al. (2020) The data culture continuum: an examination of school data cultures. Educational Administration Quarterly 56: 533–569.

32.

Lewis

Hartong

(2022) New shadow professionals and infrastructures around the datafied school: topological thinking as an analytical device. European Educational Research Journal 21(6): 946–960. DOI: 10.1177/14749041211007496.

33.

Lewis

Holloway

(2019) Datafying the teaching ‘profession’: remaking the professional teacher in the image of data. Cambridge Journal of Education 49: 35–51.

34.

Manolev

Sullivan

Slee

(2019) The datafication of discipline: ClassDojo, surveillance and a performative classroom culture. Learning, Media and Technology 44(1): 36–51.

35.

Moss

(2017) Assessment, accountability and the literacy curriculum: reimagining the future in the light of the past. Literacy 51(2): 56–64.

36.

Nespor

(1997) Tangled up in School. Politics, Space, Bodies and Signs in the Educational Process. Mahwah, NJ: Lawrence Erlbaum Associates.

37.

Ozga

(2009) Governing education through data in England: from regulation to self-evaluation. Journal of Education Policy 24(2): 149–162.

38.

Ozga

(2016) Trust in numbers? Digital Education Governance and the inspection process. European Educational Research Journal 15(1): 69–81.

39.

Perrotta

Gulson

Williamson

, et al. (2021) Automation, APIs and the distributed labour of platform pedagogies in Google Classroom. Critical Studies in Education 62: 97–113.

40.

Piattoeva

(2021) Numbers and their contexts: how quantified actors narrate numbers and decontextualization. Educational Assessment, Evaluation and Accountability 33: 511–533. DOI: 10.1007/s11092-021-09363-x.

41.

Piattoeva

Boden

(2020) Escaping numbers? The ambiguities of the governance of education through data. International Studies in Sociology of Education 29(1–2): 1–18.

42.

Piattoeva

Saari

(2018) The infrastructures of objectivity in standardized testing. In: Maddox

(ed) International Large-Scale Assessments in Education: Insider Research Perspectives. London, UK: Bloomsbury.

43.

Piattoeva

Saari

(2022) Rubbing against data infrastructure (s): methodological explorations on working with(in) the impossibility of exteriority. Journal of Education Policy 37(2): 165–185.

44.

Roberts-Holmes

(2014) The ‘datafication’ of early years pedagogy: ‘if the teaching is good, the data should be good and if there’s bad teaching, there is bad data. Journal of Education Policy 30(3): 302–315.

45.

Seaver

(2017) Algorithms as culture: some tactics for the ethnography of algorithmic systems. Big Data & Society 4(2): 205395171773810.

46.

Sellar

(2015) Data infrastructure: a review of expanding accountability systems and large-scale assessments in education. Discourse: Studies in the Cultural Politics of Education 36(5): 765–777.

47.

Selwyn

(2021) The human labour of school data: exploring the production of digital data in schools. Oxford Review of Education 47: 353–368.

48.

Selwyn

Henderson

Chao

(2015) Exploring the role of digital data in contemporary schools and schooling - ‘200,000 lines in an Excel spreadsheet. British Educational Research Journal 41(5): 767–781.

49.

Selwyn

Cumbo

Pangrazio

(2022) Data classes: an investigation of the people that ‘do data’ in schools. In: Pangrazio

Sefton-Green

(eds) Learning to Live with Datafication: Educational Case Studies and Initiatives from Around the World. Oxfordshire, UK: Routledge. Available at: https://bridges.monash.edu/articles/chapter/Data_classes_an_investigation_of_the_people_that_do_data_in_schools/18950555/2 (accessed 26 July 2022).

50.

Souto-Otero

Beneito-Montagut

(2016) From governing through data to governmentality through data: artefacts, strategies and the digital turn. European Educational Research Journal 15(1): 14–33.

51.

Spina

(2020) Data Culture and the Organisation of Teachers’ Work: An Institutional Ethnography. Oxfordshire, UK: Routledge. Available at: https://doi.org/10.4324/9780429056192.

52.

Star

(1999) The ethnography of infrastructure. American Behavioral Scientist 43(3): 377–391.

53.

Williamson

(2015) Governing software: networks, databases and algorithmic power in the digital governance of public education. Learning, Media and Technology 40: 83–105.

54.

Williamson

(2016) Digital education governance: data visualization and predictive analytics as ‘real-time’ policy instruments. Journal of Education Policy 31(2): 123–141. DOI: 10.1080/02680939.2015.1035758.