Keywords

1.1 Judging the Ethics of Research

When asked to judge the ethical issues involved in research or any evidence-gathering activity, any research ethicist worth their salt will (or should) reply, at least initially: ‘It depends’. This is neither sophistry nor evasive legalism. Instead, it is a specific form of casuistry used in ethics in which general ethical principles are applied to the specifics of actual cases and inferences made through analogy. It is valued as a structured yet flexible approach to real-world ethical challenges. Case study methods recognise the complexities of depth and detail involved in assessing research activities. Another way of putting this is to say: ‘Don’t ask me to make a judgement about a piece of research until I have the details of the project and the context in which it will or did take place.’ Understanding and fully explicating a context is vital as far as ethical research (and evidence-gathering) is concerned, along with taking account of the complex interrelationship between context and method (Miller and Dingwall 1997).

This rationale lies behind this collection of case studies which is one outcome from the EU-funded PRO-RES Project.Footnote 1 One aim of this project was to establish the virtues, values, principles and standards most commonly held as supportive of ethical practice by researchers, scientists and evidence-generators and users. The project team conducted desk research, workshops and consulted throughout the project with a wide range of stakeholders (PRO-RES 2021a). The resulting Scientific, Trustworthy, and Ethical evidence for Policy (STEP) ACCORD was devised, which all stakeholders could sign up to and endorse in the interests of ensuring any policies which are the outcome of research findings are based upon ethical evidence (PRO-RES 2021b).

By ‘ethical evidence’ we mean results and findings that have been generated by research and other activities during which the standards of research ethics and integrity have been upheld (Iphofen and O’Mathúna 2022). The first statement of the STEP ACCORD is that policy should be evidence-based, meaning that it is underpinned by high-quality research, analysis and evidence (PRO-RES 2021b). While our topic could be said to be research ethics, we have chosen to refer more broadly to evidence-generating activities. Much debate has occurred over the precise definition of research under the apparent assumption that ‘non-research projects’ fall outside the purview of requirements to obtain ethics approval from an ethics review body. This debate is more about the regulation of research than the ethics of research and has contributed to an unbalanced approach to the ethics of research (O’Mathúna 2018). Research and evidence-generating activities raise many ethical concerns, some similar and some distinct. When the focus is primarily on which projects need to obtain what sort of ethics approval from which type of committee, the ethical issues raised by those activities themselves can receive insufficient attention. This can leave everyone involved with these activities either struggling to figure out how to manage complex and challenging ethical dilemmas or pushing ahead with those activities confident that their approval letter means they have fulfilled all their ethical responsibilities. Unfortunately, this can lead to a view that research ethics is an impediment and burden that must be overcome so that the important work in the research itself can get going.

The alternative perspective advocated by PRO-RES, and the authors of the chapters in this volume, is that ethics underpins all phases of research, from when the idea for a project is conceived, all the way through its design and implementation, and on to how its findings are disseminated and put into practice in individual decisions or in policy. Given the range of activities involved in all these phases, multiple types of ethical issues can arise. Each occurs in its own context of time and place, and this must be taken into account. While ethical principles and theories have important contributions to make at each of these points, case studies are also very important. These allow for the normative effects of various assumptions and declarations to be judged in context. We therefore asked the authors of this volume’s chapters to identify various case studies which would demonstrate the ethical challenges entailed in various types of research and evidence-generating activities. These illustrative case studies explore various innovative topics and fields that raise challenges requiring ethical reflection and careful policymaking responses. The cases highlight diverse ethical issues and provide lessons for the various options available for policymaking (see Sect. 1.6. below). Cases are drawn from many fields, including artificial intelligence, space science, energy, data protection, professional research practice and pandemic planning. The issues are examined in different locations, including Europe, India, Africa and in global contexts. Each case is examined in detail and also helps to anticipate lessons that could be learned and applied in other situations where ethical evidence is needed to inform evidence-based policymaking.

1.2 The Case for Cases

Case studies have increasingly been used, particularly in social science (Exworthy and Powell 2012). Many reasons underlie this trend, one being the movement towards evidence-based practice. Case studies provide a methodology by which a detailed study can be conducted of a social unit, whether that unit is a person, an organization, a policy or a larger group or system (Exworthy and Powell 2012). The case study is amenable to various methodologies, mostly qualitative, which allow investigations via documentary analyses, interviews, focus groups, observations, and more.

At the same time, consensus is lacking over the precise nature of a case study. Various definitions have been offered, but Yin (2017) provides a widely cited definition with two parts. One is that a case study is an in-depth inquiry into a real-life phenomenon where the context is highly pertinent. The second part of Yin’s definition addresses the many variables involved in the case, the multiple sources of evidence explored, and the inclusion of theoretical propositions to guide the analysis. While Yin’s emphasis is on the case study as a research method, he identifies important elements of broader relevance that point to the particular value of the case study for examining ethical issues.

Other definitions of case studies emphasize their story or narrative aspects (Gwee 2018). These stories frequently highlight a dilemma in contextually rich ways, with an emphasis on how decisions can be or need to be made. Case studies are particularly helpful with ethical issues to provide crucial context and explore (and evaluate) how ethical decisions have been made or need to be made. Classic cases include the Tuskegee public health syphilis study, the Henrietta Lacks human cell line case, the Milgram and Zimbardo psychology cases, the Tea Room Trade case, and the Belfast Project in oral history research (examined here in Chap. 10). Cases exemplify core ethical principles, and how they were applied or misapplied; in addition, they examine how policies have worked well or not (Chaps. 2, 3 and 5). Cases can examine ethics in long-standing issues (like research misconduct (Chap. 7), energy production (Chap. 8), or Chap. 11’s consideration of researchers breaking the law), or with innovations in need of further ethical reflection because of their novelty (like extended space flight (Chap. 9) and AI (Chaps. 13 and 14), with the latter looking at automation in legal systems). These case studies help to situate the innovations within the context of widely regarded ethical principles and theories, and allow comparisons to be made with other technologies or practices where ethical positions have been developed. In doing so, these case studies offer pointers and suggestions for policymakers given that they are the ones who will develop applicable policies.

1.3 Research Design and Causal Inference

Not everyone is convinced of the value of the case study. It must be admitted that they have limitations, which we will reflect on shortly. Yet we believe that others go too far in their criticisms, revealing instead some prejudices against the value of the case (Yin 2017). In what has become a classic text for research design, Campbell and Stanley (1963) have few good words for what they call the ‘One Shot Case Study.’ They rank it below two other ‘pre-experimental’ designs—the One-Group Pretest–Posttest and the Static-Group Comparison—and conclude that case studies “have such a total absence of control to be of almost no scientific value” (Campbell and Stanley 1963, 6). The other designs have, in turn, a baseline and outcome measure and some degree of comparative analysis which provides them some validity. Such a criticism is legitimate if one prioritises the experimental method as the most superior in terms of effectiveness evidence and, as for Campbell and Stanley, one is striving to assess the effectiveness of educational interventions.

What is missing from that assessment is that different methodologies are more appropriate for different kinds of questions. Questions of causation and whether a particular treatment, policy or educational strategy is more effective than another are best answered by experimental methods. While experimental designs are better suited to explore causal relationships, case studies are more suited to explore “how” and “why” questions (Yin 2017). It can be more productive to view different methodologies as complementing one another, rather than examining them in hierarchical terms.

The case study approach draws on a long tradition in ethnography and anthropology: “It stresses the importance of holistic perspectives and so has more of a ‘humanistic’ emphasis. It recognises that there are multiple influences on any single individual or group and that most other methods neglect the thorough understanding of this range of influences. They usually focus on a chosen variable or variables which are tested in terms of their influence. A case study tends to make no initial assumptions about which are the key variables—preferring to allow the case to ‘speak for itself’” (Iphofen et al. 2009, 275). This tradition has sometimes discouraged people from conducting or using case studies on the assumption that they take massive amounts of time and lead to huge reports. This is the case with ethnography, but the case study method can be applied in more limited settings and can lead to high-quality, concise reports.

Another criticism of case studies is that they cannot be used to make generalizations. Certainly, there are limits to their generalisability, but the same is true of experimental studies. One randomized controlled trial cannot be generalised to the whole population without ensuring that its details are evaluated in the context of how it was conducted.

Similarly, it should not be assumed that generalisability can adequately guide practice or policy when it comes to the specifics of an individual case. A case study should not be used to support statistical generalizations (that the same percentage found in the case will be found in the general public). But a case study can be used to expand and generalize theories and thus have much usefulness. It affords a method of examining the specific (complex) interactions occurring in a case which can only be known from the details. Such an analysis can be carried out for individuals, policies or interventions.

The current COVID-19 pandemic demonstrates the dangers of generalising in the wrong context. Some people have very mild cases of COVID-19 or are asymptomatic. Others get seriously ill and even die. Sometimes people generalise from cases they know and assume they will have mild symptoms. Then they refuse to take the COVID-19 vaccine, basically generalising from similar cases. Mass vaccination is recommended for the sake of the health of the public (generalised health) and to limit the spread of a deadly virus. Cases are reported of people having adverse reactions to COVID-19 vaccines, and some people generalise from these that they will not take whatever risks might be involved in receiving the vaccine themselves. It might be theoretically possible to discover which individuals WILL react adversely to immunisation on a population level. But it is highly complex and expensive to do so, and takes an extensive period of time. Given the urgency of benefitting the health of ‘the public’, policymakers have decided that the risks to a sub-group are warranted. Only after the emergence of epidemiological data disclosing negative effects of some vaccines on some individuals will it become more clear which characteristics typify those cases which are likely to experience the adverse effects, and more accurately quantify the risks of experiencing those effects.

Much literature now points to the advantages and disadvantages of case studies (Gomm et al. 2000), and how to use them and conduct them with adequate rigour to ensure the validity of the evidence generated (Schell 1992; Yin 2011, 2017). At the same time, legitimate critiques have been made of some case studies because they have been conducted without adequate rigor, in unsystematic ways, or in ways that allowed bias to have more influence than evidence (Hammersley 2001). Part of the problem here is similar to interviewing, where some will assume that since interviews are a form of conversation, anyone can do it. Case studies have some similarities to stories, but that doesn’t mean they are quick and easy ways to report on events. That view can lead to the situation where “most people feel that they can prepare a case study, and nearly all of us believe we can understand one. Since neither view is well founded, the case study receives a lot of approbation it does not deserve” (Hoaglin et al., cited in Yin 2017, 16).

Case studies can be conducted and used in a wide range of ways (Gwee 2018). Case studies can be used as a research method, as a teaching tool, as a way of recording events so that learning can be applied to practice, and to facilitate practical problem-solving skills (Luck et al. 2006). Significant differences exist between a case study that was developed and used in research compared to one used for teaching (Yin 2017). A valid rationale for studying a ‘case’ should be provided so that it is clear that the proposed method is suitable to the topic and subject being studied. The unit of study for a case could be an individual person, social group, community, or society. Sometimes that specific case alone will constitute the actual research project. Thus, the study could be of one individual’s experience, with insights and understanding gained of the individual’s situation which could be of use to understand others’ experiences. Often there will be attempts made at a comparison between cases—one organisation being compared to another, with both being studied in some detail, and in terms of the same or similar criteria. Given this variety, it is important to use cases in ways appropriate to how they were generated.

The case study continues to be an important piece of evidence in clinical decision-making in medicine and healthcare. Here, case studies do not demonstrate causation or effectiveness, but are used as an important step in understanding the experiences of patients, particularly with a new or confusing set of symptoms. This was clearly seen as clinicians published case studies describing a new respiratory infection which the world now knows to be COVID-19. Only as case studies were generated, and the patterns brought together in larger collections of cases, did the characteristics of the illness come to inform those seeking to diagnose at the bedside (Borges do Nascimento et al. 2020). Indeed case studies are frequently favoured in nursing, healthcare and social work research where professional missions require a focus on the care of the individual and where cases facilitate making use of the range of research paradigms (Galatzer-Levy et al. 2000; Mattaini 1996; Gray 1998; Luck et al. 2006).

1.4 Devil’s in the Detail

Our main concern in this collection is not with case study aetiology but rather to draw on the advantages of the method to highlight key ethical issues related to the use of evidence in influencing policy. Thus, we make no claim to causal ‘generalisation’ on the basis of these reports—but instead we seek to help elucidate ethics issues, if even theoretical, and anticipate responses and obstacles in similar situations and contexts that might help decision-making in novel circumstances. A key strength of case studies is their capacity to connect abstract theoretical concepts to the complex realities of practice and the real world (Luck et al. 2006). Ethics cases clearly fit this description and allow the contextual details of issues and dilemmas to be included in discussions of how ethical principles apply as policy is being developed.

Since cases are highly focussed on the specifics of the situation, more time can be given over to data gathering which may be of both qualitative and quantitative natures. Given the many variables involved in the ‘real life’ setting, increased methodological flexibility is required (Yin 2017). This means seeking to maximise the data sources—such as archives (personal and public), records (such as personal diaries), observations (participant and covert) and interviews (face-to-face and online)—and revisiting all sources when necessary and as case participants and time allows.

1.5 Cases and Policymaking

Case studies allow researchers and practitioners to learn from the specifics of a situation and apply that learning in similar situations. Ethics case studies allow such reflection to facilitate the development of ethical decision-making skills. This volume has major interests in ethics and evidence-generation (research), but also in a third area: policymaking. Cases can influence policymaking, such as how one case can receive widespread attention and become the impetus to create policy that aims to prevent similar cases. For example, the US federal Brady Law was enacted in 1993 to require background checks on people before they purchase a gun (ATF 2021). The law was named for White House Press Secretary James Brady, and his case became widely known in the US. He was shot and paralyzed during John Hinckley, Jr.’s 1981 assassination attempt on President Ronald Reagan. Another example, this time in a research context, was how the Tuskegee Syphilis Study led, after its public exposure in 1971, to the US Department of Health, Education and Welfare appointing an expert panel to examine the ethics of that case. This resulted in federal policymakers enacting the National Research Act in 1974, which included setting up a national commission that published the Belmont Report in 1976. This report continues to strongly influence research ethics practice around the world. These examples highlight the power of a case study to influence policymaking.

One of the challenges for policymakers, though, is that compelling cases can often be provided for opposite sides of an issue. Also, while the Belmont Report has been praised for articulating a small number of key ethical principles, how those principles should be applied in specific instances of research remains an ongoing challenge and a point of much discussion. This is particularly relevant for innovative techniques and technologies. Hence the importance of cases interacting with general principles and leading to ongoing reflection and debate over the applicable cases. At the same time, new areas of research and evidence generation activities will lead to questions about how existing ethical principles and values apply. New case studies can help to facilitate that reflection, which can then allow policymakers to consider whether existing policy should be adapted or whether whole new areas of policy are needed.

Case studies also can play an important role in learning from and evaluating policy. Policymakers tend to focus on practical, day-to-day concerns and with the introduction of new programmes (Exworthy and Peckam 2012). Time and resources may be scant when it comes to evaluating how well existing policies are performing or reflecting on how policies can be adapted to overcome shortcomings (Hunter 2003). Effective policies may exist elsewhere (historically or geographically) and be more easily adapted to a new context instead of starting policymaking from scratch. Case studies can permit learning from past policies (or situations where policies did not exist), and they can illuminate various factors that should be explored in more detail in the context of the current issue or situation. Chaps. 2, 3 and 5 in this volume are examples of this type of case study.

1.6 The Moral Gain

This volume reflects the ambiguity of ethical dilemmas in contemporary policymaking. Analyses will reflect current debates where consensus has not been achieved yet. These cases illustrate key points made throughout the PRO-RES project: that ethical decision-making is a fluid enterprise, where values, principles and standards must constantly be applied to new situations, new events and new research developments. The cases illustrate how no ‘one point’ exists in the research process where judgements about ethics can be regarded as ‘final.’ Case studies provide excellent ways for readers to develop important decision-making skills.

Research produces novel products and processes which can have broad implications for society, the environment and relationships. Research methods themselves are modified or applied in new ways and places, requiring further ethical reflection. New topics and whole fields of research develop and require careful evaluation and thoughtful responses. New case studies are needed because research constantly generates new issues and new ethics questions for policymaking.

The cases found in this volume address a wide range of topics and involve several disciplines. The cases were selected by the parameters of the PRO-RES project and the Horizon 2020 funding call to which it responded. First, the call was concerned with both research ethics and scientific integrity and each of the cases addresses one or both of these areas. The call sought projects that addressed non-medical research, and the cases here address disciplines such as social sciences, engineering, artificial intelligence and One Health. The call also sought particular attention be given to (a) covert research, (b) working in dangerous areas/conflict zones and (c) behavioral research collecting data from social media/internet sources. Hence, we included cases that addressed each of these areas. Finally, while an EU-funded project can be expected to have a European focus, the issues addressed have global implications. Therefore, we wanted to include cases studies from outside Europe and did so by involving authors from India and Africa to reflect on the volume’s areas of interest.

The first case study offered in this volume (Chap. 2) examines a significant policy approach taken by the European Union to address ethics and integrity in research and innovation: Responsible Research and Innovation (RRI). This chapter examines the lessons that can be learned from RRI in a European context. Chapter 3 elaborates on this topic with another policy learning case study, but this time examining RRI in India. One of the critiques made of RRI is that it can be Euro-centric. This case study examines this claim, and also describes how a distinctively Indian concept, Scientific Temper, can add to and contextualise RRI. Chapter 4 takes a different approach in being a case study of the development of research ethics guidance in the United Kingdom (UK). It explores the history underlying the research ethics framework commissioned by the UK Research Integrity Office (UKRIO) and the Association of Research Managers and Administrators (ARMA), and points to lessons that can be learned about the policy-development process itself.

While staying focused on policy related to research ethics, the chapters that follow include case studies that address more targeted concerns. Chapter 5 examines the impact of the European Union’s (EU) General Data Protection Regulation (GDPR) in the Republic of Croatia. Research data collected in Croatia is used to explore the handling of personal data before and after the introduction of GDPR. This case study aims to provide lessons learned that could contribute to research ethics policies and procedures in other European Member States.

Chapter 6 moves from policy itself to the role of policy advisors in policymaking. This case study explores the distinct responsibilities of those elevated to the role of “policy advisor,” especially given the current lack of policy to regulate this field or how its advice is used by policymakers. Next, Chap. 7 straddles the previous chapters’ focus on policy and its evaluation while introducing the focus of the next section on historical case studies. This chapter uses the so-called “race for the superconductor” as a case study by which the PRO-RES ethics framework is used to explore specific ethical dilemmas (PRO-RES 2021b). This case study is especially useful for policymakers because of how it reveals the multiple difficulties in balancing economic, political, institutional and professional requirements and values.

The next case study continues the use of historical cases, but here to explore the challenges facing innovative research into unorthodox energy technology that has the potential to displace traditional energy suppliers. The wave power case in Chap. 8 highlights how conducting research with integrity can have serious consequences and come with considerable cost. The case also points to the importance of transparency in how evidence is used in policymaking so that trust in science and scientists is promoted at the same time as science is used in the public interest. Another area of cutting-edge scientific innovation is explored in Chap. 9, but this time looking to the future. This case study examines space exploration, and specifically the ethical issues around establishing safe exposure standards for astronauts embarking on extended duration spaceflights. This case highlights the ethical challenges in policymaking focused on an elite group of people (astronauts) who embark on extremely risky activities in the name of science and humanity.

Chapter 10 moves from the physical sciences to the social sciences. The Belfast Project provides a case study to explore the ethical challenges of conducting research after violent conflict. In this case, researchers promised anonymity and confidentiality to research participants, yet that was overturned through legal proceedings which highlighted the limits of confidentiality in research. This case points to the difficulty of balancing the value of research archives in understanding conflict against the value of providing juridical evidence to promote justice. Another social science case is examined in Chap. 11, this time in ethnography. This so-called ‘urban explorer’ case study explores the justifications that might exist for undertaking covert research where researchers break the law (in this case by trespassing) in order to investigate a topic that would remain otherwise poorly understood. This case raises a number of important questions for policymakers around: the freedoms that researchers should be given to act in the public interest; when researchers are justified in breaking the law; and what responsibilities and consequences researchers should accept if they believe they are justified in doing so.

Further complexity in research and evidence generation is introduced in Chap. 12. A case study in One Health is used to explore ethical issues at the intersection of animal, human and environmental ethics. The pertinence of such studies has been highlighted by COVID-19, yet policies lag behind in recognising the urgency and complexity of initiating investigations into novel outbreaks, such as the one discussed here that occurred among animals in Ethiopia. Chapter 13 retains the COVID-19 setting, but returns the attention to technological innovation. Artificial intelligence (AI) is the focus of these two chapters in the volume, here examining the ethical challenges arising from the emergency authorisation of using AI to respond to the public health needs created by the COVID-19 pandemic. Chapter 14 addresses a longer term use of AI in addressing problems and challenges in the legal system. Using the so-called Robodebt case, the chapter explores the reasons why legal systems are turning to AI and other automated procedures. The Robodebt case highlights problems when AI algorithms are built on inaccurate assumptions and implemented with little human oversight. This case shows the massive problems for hundreds of thousands of Australians who became victims of poorly conceived AI and makes recommendations to assist policymakers to avoid similar debacles. The last chapter (Chap. 15) draws some general conclusions from all the cases that are relevant when using case studies.

1.7 Into the Future

This volume focuses on ethics in research and professional integrity and how we can be clear about the lessons that can be drawn to assist policymakers. The cases provided cover a wide range of situations, settings, and disciplines. They cover international, national, organisational, group and individual levels of concern. Each case raises distinct issues, yet also points to some general features of research, evidence-generation, ethics and policymaking. All the studies illustrate the difficulties of drawing clear ‘boundaries’ between the research and the context. All these case studies show how in real situations dynamic judgements have to be made about many different issues. Guidelines and policies do help and are needed. But at the same time, researchers, policymakers and everyone else involved in evidence generation and evidence implementation need to embody the virtues that are central to good research. Judgments will need to be made in many areas, for example, about how much transparency can be allowed, or is ethically justified; how much risk can be taken, both with participants’ safety and also with the researchers’ safety; how much information can be disclosed to or withheld from participants in their own interests and for the benefit of the ‘science’; and many others. All of these point to just how difficult it can be to apply common standards across disciplines, professions, cultures and countries. That difficulty must be acknowledged and lead to open discussions with the aim of improving practice. The cases presented here point to efforts that have been made towards this. None of them is perfect. Lessons must be learned from all of them, towards which Chap.15 aims to be a starting point. Only by openly discussing and reflecting on past practice can lessons be learned that can inform policymaking that aims to improve future practice. In this way, ethical progress can become an essential aspect of innovation in research and evidence-generation.