Making students visible: Transparency and the politics of digital assessment

Abstract

This article examines how students understood and responded to transparency in digital assessment through a case study of Cadmus, a platform that embeds guidance within the assessment process while also making aspects of student activity visible to staff. Drawing on surveys, classroom discussion, and reflective analysis from a postgraduate politics module, the article shows that digital transparency is experienced in different ways. Some students valued the structure and guidance the platform provided. Others were uneasy about being made visible through it and felt constrained by the way it shaped their work. The article argues that these responses matter because digital assessment platforms do more than deliver instructions: they also shape how students experience authority and trust. Cadmus is therefore approached as a site where institutional values and power relations are encountered and questioned, rather than simply a neutral, technical instrument. The article concludes that educators should treat these moments of student unease and critique as opportunities for reflexive learning and should design digital assessment in ways that make platform politics discussable while foregrounding student agency and trust.

Introduction: When the platform becomes the pedagogy

I don’t like the idea of it watching me while I write.

These concerned words were uttered by a student in our postgraduate International Politics seminar. We had just finished reviewing an assessment brief when I explained that their upcoming assignment would be submitted through Cadmus, a digital assessment platform designed to scaffold academic integrity and support inclusive learning by embedding student guidance and offering staff visibility over aspects of students’ working processes. I was expecting questions about deadlines or how to approach the unfamiliar task. Instead, a hand went up to ask: ‘Is this a new form of surveillance?’

That moment pivoted the classroom. What began as a logistical announcement became a provocation for a broader conversation about visibility and power in digital learning. What does it mean, we asked, to write within an environment that records traces of process? What changes when tools presented as fair and clarifying are experienced as coercive or difficult to interpret?

As digital platforms increasingly mediate students’ learning, educators must attend to their political implications, especially as universities expand their use of digital infrastructures. This article offers a case study based on student feedback and surveys, alongside my reflections as module instructor during a Cadmus pilot in a UK higher education institution. Its purpose is to clarify what ‘transparency’ can mean when implemented through platform design and to show how students interpret that transparency in practice. In this article, transparency refers to structured visibility: the way platforms make aspects of assessment expectations and student engagement visible within an interface, while also making students’ writing practices and engagement traces legible to educators and institutions. I argue that this form of visibility is never neutral because it redistributes who can see what, under what conditions, and with what consequences.

I first outline why transparency in assessment is relational and political, then analyse how students used Cadmus and how they described its constraints. I end by arguing that platforms can reshape assessment culture, and that educators can respond by making platform politics discussable rather than treating the tool as neutral. In this sense, the article contributes to the growing body of work in Politics that takes teaching and learning seriously as a form of scholarly reflection within the discipline (Elliott et al., 2025).

The transparent classroom is not neutral

Transparency is often invoked in higher education policy and practice as an unqualified good (Khalil et al., 2023). Within assessment discourse, it is commonly associated with clarity and equity because ‘transparency breeds trust’ (Carless, 2009: 86) and explicit expectations are assumed to reduce hidden disadvantage. Scholarship on digital assessment platforms often draws on this ethos, emphasising their promise for access and inclusion through embedded guidance and formative feedback within the assessment workflow. Onwuegbuchulam et al. (2024) found that embedding academic skill-building prompts and referencing guidance within the assessment environment was valuable for students unfamiliar with academic conventions or working in a second language. Similarly, Riddoch-Contreras (2024) reflected on using Cadmus on a second-year undergraduate module and argued that digital tools can streamline guidance and reduce ambiguity, enabling students to focus more clearly on disciplinary thinking and argument construction.

These are clear benefits, but they fail to capture how transparency works as a political relation. As scholars have argued, transparency is not ‘a static attribute’ (Gonsalves and Lin, 2024: 14). It is a discursive tool operating within fields of power, meaning that it organises relationships by deciding what is visible, to whom, and on what terms. This attention to students’ experience connects with O’Neill’s (2024) argument that students’ lived experiences can reveal how institutional and curricular structures are organised and whose assumptions they privilege. Trowler and Cooper (2002: 222) describe ‘teaching and learning regimes’ that shape how assessment criteria are interpreted and enacted through cultural and disciplinary norms that exceed what rubrics can capture. Hence, Gonsalves and Lin (2024: 11) challenge ‘techno-rational’ approaches to assessment in which students are positioned as passive receivers of information, expected to decode and internalise assessment norms independently. This model neglects the sociomaterial and relational dimensions of learning. Across different traditions of learning theory, including supported learning (Vygotsky, 1978), experiential learning (Kolb, 1983), and sociomaterial perspectives (Fenwick, 2015), learning depends on interaction and interpretation. Transparency, therefore, cannot be reduced to providing clearer instructions. Students need support in understanding how expectations apply, and educators must exercise judgement in supporting different learning trajectories.

Digital assessment platforms promise to make expectations clear by embedding assessment criteria, guidance, checklists, and referencing support into a student interface. This promise rests on a particular vision of what academic success looks like and how it should be pursued (Bearman and Ajjawi, 2018; Gonsalves and Lin, 2024). By structuring the workflow through prompts and steps, platforms can also impose behavioural expectations that automate judgement and reduce space for student variation. Ball’s (2003) critique of performativity and Biesta’s (2010) concern with measurement cultures are relevant here: assessment systems can shift from supporting learning towards managing performance within centrally sanctioned frameworks of ‘quality’ and ‘assurance’. In this context, transparency can become a governance technology, framing learning as something that should be visible and trackable.

Transparency takes on even sharper implications when it is entangled with surveillance. In a context shaped by concern about academic integrity and the rise of generative AI, platforms increasingly advertise capacity to track keystrokes and flag pasted content. Nor is this unique to Cadmus: systems like Moodle record logins and access patterns, generating data infrastructures that students rarely consider in their daily practice. Within platform-mediated transparency, the student becomes visible to institutional actors, while key features of the platform’s classifications and the institutional purposes attached to the data remain difficult for students to see.

This produces an asymmetry of visibility. Foucault’s account of the panopticon and Ball’s work on performativity offer useful lenses for understanding how visibility can reorganise behaviour and reshape educational norms (Ball, 2003; Foucault, 1995). In contexts where students experience observation, they may modify their behaviour, internalising the sense that processes are being watched and judged (Dai et al., 2025: 660). Platforms can therefore influence behaviour by encouraging students to perform compliance with expected workflows or discouraging practices that might trigger scrutiny. This is especially likely in a policy environment where measurement and comparison become routine ways of establishing worth and quality (Walls, 2020: 292). Time spent on a platform and patterns of engagement can be interpreted as signals of effort or risk. However, this data is ambiguous: is the student who spent several hours on the platform a more dedicated and ‘better’ student than the one who spent less time? Are they struggling more or working inefficiently? Should the educator even be presupposing such questions?

The stakes increase in contexts where institutional monitoring is tied to external compliance regimes. At many institutions, staff are required to track attendance and engagement data to fulfil reporting obligations to the Home Office for students on visas. Under such conditions, the boundary between educational support and institutional monitoring becomes blurred. A tool introduced as supportive can become entangled with institutional accountability and immigration governance. This does not determine what students will think; but it does help to explain why visibility can feel politically charged, especially for international students whose status is already governed through monitoring. Against this backdrop, students’ questions about Cadmus prompted closer attention to how the platform was interpreted and experienced in practice.

Context and method: Cadmus as a site of assessment politics

I initially agreed to pilot the platform with attention focused on prospective inclusion benefits and clarity. During implementation, I came to see how digital assessment platforms facilitate and structure particular forms of academic engagement, through which institutional aims and commercial supplier logics can be encoded, communicated, experienced, and resisted. The purpose of my study therefore shifted from testing whether Cadmus ‘worked’ towards exploring how students interpreted the pedagogical logics embodied in the platform, and how those interpretations shaped classroom dynamics and assessment relationships.

The study was situated within an International Politics postgraduate module at a research-intensive university in the United Kingdom (UK). The cohort was diverse, comprising home and international students, many of whom were navigating UK higher education conventions for the first time. The assessment task was a comparative review of two academic books, designed to develop evaluative judgement and disciplinary writing. Its distinctive feature was delivery through Cadmus, where students were invited to write within the platform while accessing embedded supports such as referencing guidance and integrity prompts. For example, when a substantial portion of text was pasted into the platform, Cadmus queried authorship and prompted students to consider paraphrasing in line with academic integrity norms.

As the module instructor, I introduced the platform, framed the assessment, and responded to students’ questions as they arose. This position shaped how the interaction unfolded and how student responses are interpreted. Student comments and survey responses are therefore treated as situated within our ongoing pedagogical relationship, rather than as detached statements that speak for themselves. My own reflections also form part of the evidence, since the significance of the classroom moment emerged through my interpretation of it as a point of political learning. This approach is consistent with bell hooks’ (1994) account of engaged pedagogy, which recognises teaching as a relational practice through which authority and learning are negotiated in context.

To investigate students’ experiences, I used a two-stage exploratory survey: pre-assignment (n = 19) on familiarity with academic integrity and confidence in assessment expectations, and post-assignment (n = 7) on how Cadmus shaped writing and understanding of the task, including open-ended reflections. Responses were analysed through close reading of recurring concerns and interpreted in dialogue with work on transparency and assessment regimes (Bearman and Ajjawi, 2018; Gonsalves and Lin, 2024; Trowler and Cooper, 2002). Although the sample size is small, the study focuses on how students understood and responded to transparency as it was experienced through the platform, rather than making general claims about Cadmus. Ethical approval was obtained, and participation was voluntary and anonymised.

Student experiences of platform-mediated assessment

Platform analytics indicated varied engagement practices. Forty percent of students worked entirely off-platform, while one-third used Cadmus from start to finish, and the remainder adopted a hybrid approach. All students accessed the Cadmus manual, and most students accessed feedback after their essay had been marked, often returning to it multiple times. The average time spent on the platform was 4 hours and 45 minutes, suggesting that even students who preferred to write externally still used Cadmus as a supplementary space for planning.

For some students, the platform’s embedded guidance functioned as intended. Tangible benefits were highlighted in relation to the notes tab and checklist, and in relation to step-by-step prompts that supported structuring. These responses suggest that, for some students, Cadmus provided a scaffold that supported independent learning and reduced uncertainty, aligning with the positive accounts in the works by Onwuegbuchulam et al. (2024) and Riddoch-Contreras (2024).

For other students, responses that began as comments on usability and workflow opened into concerns about autonomy and interpretation. In earlier work on the ‘implementation staircase’, Reynolds and Saunders (1987) and Saunders et al., (2015) describe how policy intentions are interpreted and enacted differently as they move through levels of governance. A similar interpretive dynamic appeared here. As the educator, I initially saw the platform as a tool designed to support students. Some students, however, experienced it as a system that shaped how they worked and made their activity visible to others. The findings below are therefore organised around two connected experiences: how students used platform-based guidance to make sense of expectations, and how they experienced limits on control through the platform’s design. Trust links these experiences, since transparency is meaningful only when students believe the system works in their interests and that their actions will be interpreted fairly (Carless, 2009; Khalil et al., 2023).

Navigating expectations: Guidance and the limits of platform instruction

Some students welcomed Cadmus’ scaffolding features, which aimed to clarify assessment expectations and promote academic integrity. In the pre-assignment survey, one student worried about ‘structuring the work’ and ‘managing time’. Several international students flagged unfamiliarity with UK academic writing and referencing norms, and with the unfamiliar assignment, including one student who wrote, ‘I’ve never do [sic] this before’, and another who identified ‘understanding module requirements and marking expectations’ as their key concern. In relation to the potential value of embedded supports, one student gave the highest possible rating (5 on a 1 to 5 Likert-type scale) for the anticipated helpfulness of the tools, suggesting confidence that structured prompts and referencing guidance might be useful.

Even here, students’ expectations were not uniform. Several students expressed scepticism about whether a platform could deliver inclusivity. One wrote ‘not particularly’ in response to a question about inclusivity features, while another offered a flat ‘no’. These brief responses signalled reluctance to accept transparency as something that can be delivered as a technical fix, echoing critiques of techno-rational assessment models (Gonsalves and Lin, 2024).

This became more apparent in reflections on the intellectual demands of the assignment itself. Several students indicated that while Cadmus offered structured support, it did not address the kind of discipline-specific judgement they needed in order to interpret the task. One student wrote, ‘I did not quite understand the scope of the 2-book review and whether we needed to use loads of references, engage with other literature, etc or not. This was not explained in Cadmus either’. Another explained,

my concerns were of a different kind. I was unsure how scientifically rigorous [sic] the essay must be (should I write a book review including a research protocol or sampling strategy, or can I take a more ‘casual’ approach?). Cadmus provided little guidance on these issues as they are unique to each individual.

These responses underscore that platform guidance cannot substitute for dialogue and the relational work of teaching described by sociomaterial learning perspectives (Fenwick, 2015) and by theories of experiential and supported learning (Kolb, 1983; Vygotsky, 1978).

They also point to a specific political implication. Platform-based transparency can place responsibility for ‘getting it right’ on students as individual decoders of prompts and criteria, even where students are asking for conversation about how judgement should be applied in context. This concern aligns with Trowler and Cooper’s (2002) argument that criteria sit within regimes of meaning that students must learn to interpret. It also aligns with Bearman and Ajjawi’s (2018) critique of ‘myths of transparency’: what appears visible can still be difficult to interpret without shared interpretive practices.

Negotiating visibility and control: Workflow constraints and the experience of being seen

Alongside requests for clarification, many students described constraint, especially in relation to the interface and workflow. Some students adapted to the structured environment. Other students resisted it, drafting in Word or Google Docs and pasting into Cadmus for submission. These decisions reflected discomfort with being channelled into a system that shaped how writing could be produced and revised.

One student explained, ‘Although I tried to use Cadmus, I found that some functions, such as adding comments, changing font size and formatting . . . were not available. This made the process less comfortable, therefore I used Word’. Another wrote, ‘It was confusing and a bit stressful to have to adapt to a different writing format’. Here, the platform’s design offered support for standardisation and comparability, while students experienced a loss of control over their writing practices. These comments connect to the argument that platforms can encode a model of ‘good work’ as a standardised process, rather than as a plural set of practices shaped by disciplinary and personal routines (Bearman and Ajjawi, 2018; Trowler and Cooper, 2002).

Visibility intensified these concerns. Even when students did not describe the platform as surveillance, their comments suggested discomfort with working in a system that records aspects of their writing process. In this context, guidance and oversight became difficult to separate. The central issue was trust: trust in how activity data might be read, and trust in whether the platform was designed with students’ interests in mind (Carless, 2009; Khalil et al., 2023). Once visibility is built into the assessment environment, students may question whether the platform exists to support learning or to manage institutional risk, particularly in response to concerns about academic integrity and generative AI (Ball, 2003; Dai et al., 2025).

Pedagogical pivot: Teaching politics through the platform

The moment a student asked whether Cadmus was ‘watching’ them marked a turning point in the seminar. Until that point, both I and our guest technician from Cadmus, visiting to provide an operational tutorial, had introduced it in functional terms. We explained it as a mechanism to support structuring and referencing, and as a system that prompts integrity reflection when text is pasted. The student’s question unsettled that framing.

What followed was a lively discussion about power in digital education. Students asked who sets the rules and who has access to engagement data, and they raised questions about how that information might be interpreted. They linked platform-based assessment to wider practices such as algorithmic decision-making and biometric monitoring. For some students, these concerns connected directly to institutional monitoring requirements linked to visa compliance, highlighting how data practices in education can feed into broader systems of governance.

This moment aligned closely with bell hooks’ (1994: 156) account of ‘liberatory pedagogical practice’, in which authority shifts so that student experience and critique are recognised as legitimate sources of knowledge. Here, students shifted the agenda, and my role became one of supporting the inquiry and treating their concerns about visibility as part of the learning. The platform became something we could question together, which also made my own role more visible, since I had introduced the tool and shaped how it was understood in the classroom. In this sense, liberatory practice emerged through engagement with the platform and the conditions of participation it created. It also resonates with Smith’s (2025) discussion of students as partners, actively contributing to the learning process, and Freire’s (1970) emphasis on dialogue and co-inquiry, where teachers and students investigate shared conditions together. This unplanned turn, from ‘how to use the platform’ towards ‘what the platform does to us’, positioned students as co-inquirers and enacted a form of assessment literacy consistent with Bearman and Ajjawi’s (2018) view that transparency is never neutral and with Gonsalves and Lin’s (2024) critique of models that cast students as passive recipients.

Assessment literacy thus became a form of political practice. Students examined how academic evaluation is organised and how authority is built into platform design, paying attention to how institutional values are communicated through an interface. This article does not seek to settle whether increased policing is justified in the so-called ‘AI era’, or whether platforms provide necessary safeguards for academic integrity. Those debates are ongoing. The contribution here is different: it shows how a platform can change the assessment relationship by offering guidance while also making student activity more visible, and how students can make sense of that change in political terms. Even when integrity goals are legitimate, the experience of being monitored can affect students’ sense of agency and belonging. Making space to discuss that experience is therefore both a pedagogical responsibility and an ethical task in a politics classroom.

Conclusion: Designing for reflexive assessment

This article began with a student’s discomfort about writing within a platform that records traces of process. It ends with an implication for how educators think about transparency in digital assessment. Platform-mediated transparency should be treated as a political relationship rather than as a technical upgrade. Tools like Cadmus can scaffold learning and reduce uncertainty, especially for students who are new to academic conventions or working in a second language (Onwuegbuchulam et al., 2024; Riddoch-Contreras, 2024). They can also reshape assessment cultures by standardising how students work and by increasing how visible their activity is to educators (Bearman and Ajjawi, 2018; Gonsalves and Lin, 2024; Trowler and Cooper, 2002).

Trust remains central. But trust is not produced solely by making expectations explicit through written guidance or platform prompts. It develops when students understand how transparency works and how their activity will be interpreted, particularly where data can be read in different ways (Carless, 2009; Khalil et al., 2023). In a pilot context, students’ confidence in the instructor can help sustain that trust, because the platform is presented as an experiment within an established relationship. Wider roll-out can change these conditions, as the tool becomes more distant from local teaching practices and more closely tied to institutional governance aims, including measurement cultures and external compliance requirements (Ball, 2003; Biesta, 2010; Walls, 2020).

The case also points to a pedagogical opportunity. When the platform becomes the pedagogy, educators can treat that moment as a site of learning. This involves making platform politics discussable: explaining what data is collected, who can see it, how it might be interpreted, and what it cannot show. It also involves treating student critique as legitimate rather than as resistance to academic integrity or clearer guidance. In doing so, educators can cultivate politically attentive learners who can recognise how authority is exercised through systems and interfaces. In a digital university where assessment infrastructures increasingly shape student experience, that form of political literacy is part of what a politics education can provide.

Importantly, this is not about discarding technology or rejecting transparency but about asking how assessments can be designed to centre agency and trust, including through co-created assessment practices that emphasise reciprocity rather than compliance. If we are willing to be unsettled, our students may become our best critics, challenging us to rethink the systems we endorse. In those moments, we can ask together: What does transparency demand of us as students and educators – and what should we demand from our institutions and platforms in return?