Skip to content

4.5 Reframing data ethics in research methods education

Jump to bottom
javieraatenas-pixel edited this page Jun 18, 2026 · 2 revisions

Content adapted from:

Introduction

In most research methods courses, ethics appears as a procedural checkpoint—typically reduced to forms, approvals, and informed consent. Atenas et al. challenge this limited view, arguing that ethics must extend beyond “securing informed consent” toward a deeper understanding of how data practices are embedded in power and society. Markham (2006) provides a crucial conceptual shift, arguing that ethics is not something external to research, but something enacted through it:

ethics is “produced, reinforced, or resisted through practice”

This reframing transforms ethics from a static body of rules into a dynamic way of doing research. To teach research methods ethically means teaching students that every methodological decision—what to measure, how to collect, how to interpret—is already an ethical act.

Markham et al. (2018) deepen this argument in the context of big data, noting that contemporary research cultures often assume that knowledge derived from large datasets is inherently objective—thereby obscuring the human decisions embedded in methods. Ethics is not something we apply to research—it is something we produce through methodological choices. Teaching methods therefore means teaching ethical enactment.

flowchart LR
    %% Nodes
    A[Research Question]
    B[Method Choice]
    C[Data Practices]
    D[Interpretation]
    E[Knowledge Production]
    F[Ethical reasoning embedded]
    A --> B
    B --> C
    C --> D
    D --> E

    %% Ethical reasoning links
    A --> F
    B --> F
    C --> F
    D --> F
    E --> F

    %% Styling (pastel colours + black text)
    style A fill:#FFD6E8,stroke:#333,color:#000
    style B fill:#D6EAF8,stroke:#333,color:#000
    style C fill:#D5F5E3,stroke:#333,color:#000
    style D fill:#FCF3CF,stroke:#333,color:#000
    style E fill:#E8DAEF,stroke:#333,color:#000
    style F fill:#FADBD8,stroke:#333,color:#000
Loading

Embedding Ethics Across the Research Lifecycle

Most curricula fail to embed ethics across the research cycle. Markham (2006) and Markham et al. (2018) reinforce that ethics is processual, unfolding across multiple decisions, while emphasising that data practices involve “multiple moments, decisions, actions, and operations” where harm can occur. As shown in table below Analysis of Research Methodology Courses and Data Science Programmes which showcases an overview of research methodology courses and data science programmes showing the prevalence of research ethics, informed consent, data ethics, and ethics readings across academic levels, methodological orientations, and data science curricula.

Category Course Unit UG Master's PhD Quantitative Qualitative Data Science Programmes
Academic Level Ethics of research and informed consent 42/118 58/81 24/51 – – –
Academic Level Data ethics 12/118 27/81 3/51 – – –
Methodology Type Ethics of research and informed consent – – – 65/114 59/136 –
Methodology Type Data ethics – – – 27/114 15/136 –
Ethics Readings Include ethics readings – – – 54/114 97/136 –
Data Science Programmes Ethics of research and informed consent – – – – – 12/80
Data Science Programmes Data ethics – – – – – 25/80

Adapted from Atenas, Havemann, and Timmermann (2023).:

Ethical issues do not arise at one moment—they emerge continuously at each stage where human judgment meets data.

flowchart TD
    %% Nodes
    A[Define Research Problem]
    B[Design Study]
    C[Collect Data]
    D[Process Data]
    E[Analyse]
    F[Communicate]

    A1["Who defines problem?"]
    B1["Whose perspective shapes design?"]
    C1[Context of participation]
    D1[Data transformations]
    E1[Interpretive bias]
    F1[Impact on audiences]
    A --> B --> C --> D --> E --> F

    %% Reflective questions
    A --> A1
    B --> B1
    C --> C1
    D --> D1
    E --> E1
    F --> F1

    %% Styling (pastel colours + black text)
    style A fill:#FFD6E8,stroke:#333,color:#000
    style B fill:#D6EAF8,stroke:#333,color:#000
    style C fill:#D5F5E3,stroke:#333,color:#000
    style D fill:#FCF3CF,stroke:#333,color:#000
    style E fill:#E8DAEF,stroke:#333,color:#000
    style F fill:#FADBD8,stroke:#333,color:#000

    style A1 fill:#FFF5F8,stroke:#333,color:#000
    style B1 fill:#F4F9FD,stroke:#333,color:#000
    style C1 fill:#F3FBF7,stroke:#333,color:#000
    style D1 fill:#FEF9E7,stroke:#333,color:#000
    style E1 fill:#F7F1FB,stroke:#333,color:#000
    style F1 fill:#FDEDEC,stroke:#333,color:#000
Loading

Teaching Through Ethical Principles as a lived practice

The integration of ethical principles into research methods education has long been regarded as a cornerstone of responsible scholarly practice. However, as Atenas et al. (2023) demonstrate, such principles are frequently presented in a fragmented or superficial manner—often limited to procedural concerns such as informed consent and data protection. This reductionist approach risks positioning ethics as an external constraint upon research, rather than as a constitutive element of methodological practice.

A more robust pedagogical approach emerges when ethical principles are reinterpreted as lived, situated, and enacted forms of reasoning. Markham (2006) provides a pivotal reorientation in this regard, arguing that ethics is not simply applied to research, but rather “produced, reinforced, or resisted through practice”. Extending this position, Markham, Tiidenberg, and Herman (2018) caution against the technocratic assumption that ethics can be operationalised through static checklists, particularly in data-intensive environments characterised by complexity and opacity. Instead, they advocate for an understanding of ethics as emergent, relational, and embedded within ongoing methodological decisions.

From Principles to Practice

Within this reframing, ethical principles—such as respect for autonomy, privacy, fairness, and non-maleficence—should not be treated as prescriptive rules to be followed mechanistically. Rather, they function as heuristic devices that orient researchers towards critical questioning.

Atenas et al. (2023) conceptualise these principles as action-guiding, emphasising their role in enabling students to interrogate the ethical implications of data practices across the research lifecycle. Yet, as Markham (2006) suggests, the ethical significance of these principles cannot be fully apprehended in abstraction; their meaning is realised only through engagement with concrete research contexts, where competing values, uncertainties, and constraints must be negotiated.

From a pedagogical perspective, this implies that teaching ethical principles should prioritise interpretive engagement over declarative knowledge. Students should not merely be able to define ‘privacy’ or ‘fairness’, but must learn to ask:

  • What constitutes privacy in this particular context?
  • Whose definition of fairness is being applied?
  • How do these principles interact or conflict with one another?

Such questions foreground the inherently context-sensitive and contested nature of ethical reasoning, particularly in data-driven research.

Principles in Tension

An important implication of this approach is the recognition that ethical principles are often in tension rather than harmonious. For instance, efforts to ensure fairness in algorithmic systems may require the collection of sensitive demographic data, thereby potentially infringing upon privacy. Similarly, maximising transparency in research processes may conflict with the need to protect participants from harm.

Atenas et al. (2023) acknowledge these complexities, noting that data ethics involves navigating “moral problems arising from data… in view of developing morally desirable solutions”. Markham et al. (2018) further elaborate that such tensions are intensified within large-scale, digitally mediated environments, where the consequences of methodological decisions are often diffuse and unpredictable.

Teaching ethical principles, therefore, requires moving beyond the presentation of stable norms towards cultivating students’ capacity to balance competing ethical demands. This involves developing what might be termed ethical judgement: the ability to weigh principles against one another in light of specific circumstances, and to justify decisions reflexively.

Ethics as Relational and Situated

A further dimension of ethical practice concerns its fundamentally relational character. Markham (2006) emphasises that ethical decisions are always embedded within relationships—between researchers and participants, between data and the individuals they represent, and between knowledge production and its wider societal consequences.

Markham et al. (2018) extend this insight in the context of digital research, highlighting how contemporary data practices often obscure these relationships. Data are frequently abstracted, decontextualised, and repurposed, leading to a detachment between the researcher and the lived realities of those whose data are being analysed.

In this sense, teaching ethical principles requires re-situating data within their social and human context, so ethical principles become tools for reconnecting data practices with their human implications. Students must be encouraged to recognise that datasets are not neutral artefacts, but rather:

  • constructed through particular methodological choices
  • shaped by existing power structures
  • imbued with social and cultural meaning

Pedagogical Implications: Towards Lived Ethics

If ethical principles are to function as lived practice, pedagogical design must actively support their enactment. This involves creating learning environments in which students can engage with authentic, complex, and ambiguous scenarios, rather than simplified or idealised cases.

Atenas et al. (2023) advocate for the integration of case studies, participatory approaches, and critical reflection within research methods teaching. In alignment with Markham et al. (2018), such approaches should enable students to experience ethics as something that is continually negotiated within practice.

  • require students to articulate how multiple principles intersect and conflict
  • ask them to justify their methodological choices in ethical terms
  • encourage reflexive consideration of their own positionality as researchers

Through such activities, ethical principles are transformed from static content into dynamic processes of inquiry and reflection.

flowchart TB
    %% Nodes
    A[Ethical Principles Framework]

    B[Autonomy]
    C[Privacy]
    D[Justice]
    E[Non-maleficence]
    F[Bias Awareness]
    G[Power Analysis]

    B1[Consent & agency]
    C1[Contextual privacy]
    D1[Equity outcomes]
    E1[Prevent harm]
    F1[Recognise systemic bias]
    G1[Challenge structural inequality]
    A --> B
    A --> C
    A --> D
    A --> E
    A --> F
    A --> G

    B --> B1
    C --> C1
    D --> D1
    E --> E1
    F --> F1
    G --> G1

    %% Styling (pastel colours + black text)
    style A fill:#F8E1F4,stroke:#333,color:#000

    style B fill:#FFD6E8,stroke:#333,color:#000
    style C fill:#D6EAF8,stroke:#333,color:#000
    style D fill:#D5F5E3,stroke:#333,color:#000
    style E fill:#FCF3CF,stroke:#333,color:#000
    style F fill:#E8DAEF,stroke:#333,color:#000
    style G fill:#FADBD8,stroke:#333,color:#000

    style B1 fill:#FFF5F8,stroke:#333,color:#000
    style C1 fill:#F4F9FD,stroke:#333,color:#000
    style D1 fill:#F3FBF7,stroke:#333,color:#000
    style E1 fill:#FEF9E7,stroke:#333,color:#000
    style F1 fill:#F7F1FB,stroke:#333,color:#000
    style G1 fill:#FDEDEC,stroke:#333,color:#000
Loading

Ethics as Situated and Contextual Practice

A foundational contribution of Markham’s work is the insistence that ethics cannot be treated as universal, fixed, or independent of context. Rather, ethical decision-making is fundamentally situated, relational, and contingent upon specific social, cultural, and technological conditions. As Markham (2006) argues, ethical reasoning emerges through engagement with particular research contexts, rather than through the application of abstract rules.

This proposition is reinforced in later work, where Markham, Tiidenberg, and Herman (2018) explore how digital research environments—characterised by scale, speed, and opacity—further complicate the possibility of stable ethical frameworks. Atenas et al. (2023) similarly emphasise that ethical issues in data practices must be understood within •techno-centric environments and intersecting hierarchies of power•.

The Limits of Universal Ethics

Traditional ethical frameworks in research methods often rely on universal principles that are presumed to apply across contexts—such as autonomy, beneficence, and justice. Whilst these principles provide an essential normative foundation, both Markham (2006) and Atenas et al. (2023) highlight the limitations of relying on them as fixed standards.

Markham (2006) challenges the notion that ethics can be resolved through prior codification, arguing that such approaches risk overlooking the complexities of lived research contexts. In digitally mediated environments, this problem becomes even more pronounced. As Markham et al. (2018) observe, data practices frequently detach information from its original context, complicating the relationship between data and the individuals it represents. This disconnection makes it increasingly difficult to determine what constitutes appropriate ethical conduct.

Context, Meaning, and Interpretation

A key feature of situated ethics is the recognition that meaning itself is context-dependent. Data does not possess intrinsic meaning; rather, its significance is constructed through the contexts in which it is produced, analysed, and interpreted.

Markham (2006) emphasises that ethical decisions must take into account the multiple layers of interpretation that mediate the relationship between researcher, data, and participant. This interpretive dimension is particularly salient in digital research, where data may be repurposed across different contexts without the knowledge or consent of those who originally produced it.

Markham et al. (2018) highlight the ethical implications of such practices, noting that the abstraction and aggregation of data can obscure the lived experiences of individuals, thereby reducing complex social realities to decontextualised data points. This process raises critical questions:

  • Does publicly available data equate to ethically usable data?
  • How should researchers interpret consent in contexts where data is continuously generated and repurposed?
  • What responsibilities do researchers have towards individuals who may never be aware of their participation?

Relational Ethics in a Datafied Society

Situated ethics is also inherently relational, in that ethical decisions are shaped by the relationships between researchers, participants, and broader communities. Markham (2006) argues that ethical responsibility arises from these relationships, rather than from abstract obligations alone. Contemporary data practices disrupt traditional relational models of research. Markham et al. (2018) suggest that in large-scale digital environments, relationships between researchers and participants are often indirect or non-existent. Data may be collected without direct interaction, analysed at scale, and used for purposes far removed from its original context. This creates what might be described as a relational gap, in which ethical responsibility becomes more diffuse and difficult to locate.

Atenas et al. (2023) address this challenge by advocating for approaches grounded in an ethics of care, which emphasises attentiveness to relationships, vulnerability, and interdependence. Such an approach encourages researchers to consider not only individual participants, but also the communities and social systems within which data practices are embedded.

Ethics Under Conditions of Uncertainty

A further implication of situated ethics is the acknowledgement of uncertainty as an inherent feature of research practice. As Markham et al. (2018) note, the consequences of data practices are often unpredictable, particularly in complex digital ecosystems where data can be recombined, repurposed, and reinterpreted in unforeseen ways. This uncertainty challenges traditional models of ethical decision-making, which assume that potential risks and harms can be identified and mitigated in advance. Instead, ethical practice must be understood as an ongoing process of anticipation, monitoring, and adaptation, including:

  • anticipating impacts before research begins
  • monitoring effects during the research process
  • evaluating consequences after dissemination
flowchart TD
    %% Nodes
    A[Research Design]
    B[Anticipate impacts]
    C[Data Collection & Use]
    D[Monitor ethical implications]
    E[Analysis & Interpretation]
    F[Evaluate social consequences]
    G[Reflect & revise practice]

    H[Individual ethics: consent, privacy]
    I[Collective ethics: equity, justice]
    J[Power analysis: bias, representation]
    K[Societal impact: inequalities, governance]
    A --> B --> C --> D --> E --> F --> G
    G --> A

    %% Ethical layers
    A --> H
    C --> I
    E --> J
    F --> K

    %% Styling (pastel colours + black text)
    style A fill:#FFD6E8,stroke:#333,color:#000
    style B fill:#D6EAF8,stroke:#333,color:#000
    style C fill:#D5F5E3,stroke:#333,color:#000
    style D fill:#FCF3CF,stroke:#333,color:#000
    style E fill:#E8DAEF,stroke:#333,color:#000
    style F fill:#FADBD8,stroke:#333,color:#000
    style G fill:#EAF2F8,stroke:#333,color:#000

    style H fill:#FFF5F8,stroke:#333,color:#000
    style I fill:#F3FBF7,stroke:#333,color:#000
    style J fill:#F7F1FB,stroke:#333,color:#000
    style K fill:#FDEDEC,stroke:#333,color:#000
Loading

Clone this wiki locally