How do cancer research scientists deal with machines and consumables? Exploring research ethics from an inductive ethnographic perspective

Several social scientists and communication scholars have conducted ethnographic studies in laboratories to provide authentic insights into scientific knowledge construction as a social process. From a social constructionist perspective, and drawing on actor-network theory (Latour, Callon), understanding science in the making requires tracing networks of scientists, instruments, materials, and institutions and their interactions. Lab environments, while central to innovation, raise ethical implications that affect decision-making, necessitating ongoing attention. Bioethics offers principles (respect for autonomy, beneficence, non-maleficence, justice) to guide practice, yet the responsible and ethical use of machines and consumables in labs presents unique challenges. This paper closely observes practices and interactions within a Belgian cancer research lab to examine ethical implications of knowledge construction. Adopting an inductive ethnographic approach, it aims to advocate for responsible and reflective scientific practices from a transdisciplinary stance. Trust emerged from the data as a central ethical concern, underscoring the value of ethnography and participant observation for revealing previously unexamined aspects of scientific practice and highlighting the need for a holistic understanding of science in the making.

Trust is framed as a fundamental ethical principle underpinning research integrity and credibility. Hardin (2002) conceptualizes trust as a rational, strategic choice based on assessments of credibility and reliability of researchers, institutions, machines, and processes, supported by institutional structures (transparency, accountability, peer review) and social norms. International statements (European Code of Conduct for Research Integrity 2017; Singapore 2010; Montreal 2013; Hong Kong 2019) foreground three principles: research integrity, responsible conduct of research, and responsible use of resources and technologies. Prior scholarship highlights threats to integrity from hyper-competitiveness, inadequate training, problematic metrics, and biases in peer review/publishing (Mejlgaard et al., 2020) and argues that reliance on norms and compliance is insufficient to warrant public trust (Yarborough, 2014, 2021). Efforts to embed ethics in lab practice include strategies integrating Ethics Parallel Research (EPR), Social Labs, and the Responsible Research and Innovation (RRI) framework to anticipate implications, foster inclusion, and support normative evaluation (Buedo et al., 2023; Burget et al., 2017). Despite growing guidance, translating ethics into day-to-day lab routines remains challenging (Laas et al., 2022; Resnik et al., 2023). Trust also extends to public perceptions, making transparent, effective science communication essential (Besley et al., 2018). The present study examines trust between lab scientists and machines/consumables and how this trust unfolds to the public, noting potential reputational risks and erosion of confidence if trust is unjustified.

An ethnographic research design was employed within an oncology (fundamental cancer) research lab in Belgium. The lab comprises ten postdoctoral and doctoral researchers, technicians, and a principal investigator (PI) with a strong publication record (255 publications, several patents, H-index 81 in 2022). Data were collected over 18 months via direct participant observation, individual and collective semi-structured interviews, and detailed field notes. Interviews probed: (1) processes to verify results; (2) trust in machines; (3) trust in products from companies; (4) proportion of operations performed by other parties; (5) diagramming stages/operations in current research. Individual interviews were analyzed synchronically; a collective interview six months later was analyzed diachronically. Transcription followed the Jefferson system (Atkinson & Heritage, 1984). Pseudonyms protected identities: Paul (PI); Stephan, Daniel, Brian, Dilara (postdocs); Cédric, Gaelle (PhD researchers); Dalila, Marie (technicians); Joanne (master’s student). The lab uses a range of machines (freezers, hoods, incubators, centrifuges, western blot apparatus, microscopes, spectrophotometers) and consumes certified products (media, trypsin, powders, gels, pipettes, plates, tips, filters, etc.) sourced from external providers whose materials can affect cell behavior and research outcomes. Data analysis was qualitative and inductive, organized around ethical principles salient to lab environments: research integrity, responsible conduct of research, and responsible use of resources. Research questions: RQ1: How do researchers use lab machines and consumables in their practices? RQ1a: What are the ethical implications of such use? Ethical approval was obtained; informed consent was secured.

- Researchers widely trust machines but differ in understanding how they function and whether they measure intended variables (validity vs. reliability). Trust is often grounded in assumptions: design quality, provider reputation, prior use by others, routine maintenance, and vendor QC protocols, rather than independent validation. Participants rarely distinguish reliability (consistent operation) from validity (measuring the intended construct). Examples include: Cédric's assertion that machines are designed to do what they do; Paul's reliance on maintenance and external validation; Joanne's deference to community use and manufacturer competence. In a spectrophotometer discussion, participants described technique but not validation of measuring the targeted property, eventually acknowledging machines may measure unintended signals (e.g., contamination). - Consumables are used extensively—several participants described using “everything” readymade. Researchers generally cannot independently verify composition or purity for many reagents and rely on provider QC, documentation, and reputation. Paul acknowledged being “blinded” for certain substances, acting on trust that products perform as described. Daniel initially underestimated dependence on providers but, after mapping workflows, recognized extensive reliance. Gaelle cited industry QC and GMP but confirmed the lab does not check alterations in products. - Prior to interviews, participants reported never discussing these “philosophical” issues. During individual interviews, all but one postdoc did not consider this trust risky; in the collective interview, all acknowledged potential risks (e.g., skewed results, blinding, phenotype changes). - Risks summarized (Table 1): (1) Excessive trust in machines without validating they measure the intended parameters; (2) Excessive trust in consumables without ensuring they match provider specifications. Rationales offered include cost, competence of manufacturers, advice from peers, and provider QC/GMP, but these constitute subjective value judgments rather than direct evidence. - Scope and propagation: Practices occur continuously across experiment types and consumables. Participants indicated similar practices in prior labs across multiple countries (USA, France, Belgium, Netherlands, Spain). If errors exist in widely used machines/consumables, they may be widespread and hard for a single lab to detect. - Ethical implications: Such unexamined trust threatens research integrity, compromises responsible conduct, and raises concerns regarding the responsible use of resources and technologies. Ongoing situational awareness and ownership of research elements are lacking in current structures. Relevant data points: 18 months of ethnographic engagement; lab with ten postdoctoral and doctoral researchers plus technicians and a PI; during individual interviews only one postdoc viewed the trust as risky; collective interview led all to acknowledge risks; participants used phrases like “everything” or “100%” to describe reliance on readymade inputs.

The findings address RQ1 by detailing how researchers rely heavily on machines and readymade consumables to execute nearly all protocols, often assuming proper function and composition based on provider reputation, peer usage, or routine maintenance rather than validating that instruments measure intended constructs or that consumables conform to specifications. Regarding RQ1a, this reliance introduces ethical vulnerabilities across the three principles highlighted by major integrity codes: (1) Research integrity is jeopardized when validity is presumed; measurements may be consistent yet misaligned with intended phenomena, risking skewed results and false confidence. (2) Responsible conduct of research is compromised when accountability for critical elements is shifted to manufacturers/providers without appropriate verification or awareness of limitations. (3) Responsible use of resources and technologies is questioned when labs invest in tools and materials whose performance and composition are not critically evaluated, potentially wasting resources and propagating errors. The lab’s recognition, during the collective interview, that trust can lead to “blinding” or phenotype changes illustrates how ethnographic reflection can surface unnoticed assumptions shaping knowledge construction. Given the transnational standardization of lab machines and consumables, any systematic issues may propagate across the scientific community, amplifying the need for transparent validation practices and reflexive engagement with tools and materials. These insights align with literature emphasizing that institutional norms and compliance training alone are insufficient to secure deserved trust and public confidence; integrating ethics into routine lab practice and communication remains essential.

This inductive ethnography contributes a grounded account of how cancer researchers’ unexamined trust in machines and consumables can undermine research integrity, responsible conduct, and responsible use of resources. By revealing a taken-for-granted dependence on black-box instruments and readymade inputs—and by documenting how researchers’ views evolved through reflective dialogue—the study underscores the value of ethnographic methods for exposing hidden assumptions in knowledge construction. The paper advocates for increased critical awareness and situational ownership over research elements, including clearer distinctions between reliability and validity and more deliberate evaluation of consumables’ specifications. Potential future directions include: (i) comparative, multi-lab ethnographies to assess generalizability and map variation across settings; (ii) development and assessment of practical safeguards and reflexive routines for validating machine outputs and verifying critical reagent properties; (iii) embedding ethics within laboratory practice using approaches highlighted in the literature (e.g., Ethics Parallel Research, Social Labs, RRI) to foster anticipation, inclusivity, and accountability; and (iv) enhancing transparent documentation and communication practices to strengthen deserved trust within the scientific community and with the public.

The study focuses on a single cancer research lab with a relatively small number of potential participants, limiting generalizability to broader populations of cancer research labs.