Only experimental research and face-to-face interviews are subject to ethical concerns.

Philosophers have contemplated what constitutes ethics for centuries (notably examples include Immanuel Kant and John Stuart Mill), with considerable debate over the value of utilitarian approaches (best result for the majority) versus absolutist views of actions as either inherently right or wrong (the act is judged rather than outcome). Such debates continue to this day, but regardless of ongoing disagreements, the need for the establishment of general ethical standards for research involving humans became increasingly apparent in the mid 20th century. In particular, medical experiments by the Nazi's and the Willowbrook study (deliberate infection of people with learning difficulties with hepatitis) emphasised the need for tight controls over research.

It was only after horrific cases like these that the scientific community began to consider guidelines for research, resulting in the Nuremberg Code in 1947, the Declaration of Helsinki in 1964 and the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research in 1978.7 Such developments led to the establishment of ethics committees and research governance to regulate biomedical and social science research. Broadly speaking, research ethics have come to be defined as the moral problems encountered in connection with scientific or other academic research by the researcher, their subjects or their social environment.8

At the basic level, in order for a research project to be ethical, the dignity, rights, safety and well-being of participants must be the primary consideration.6 The main aim of an ethics committee is to ensure that this is the case. More specifically, the ESRC (the main funding body for social research in the UK) outlines several main points for doing ethical social research: (1) research should be designed, reviewed and undertaken to ensure integrity and quality; (2) research staff and subjects must be informed fully about the purpose, methods and intended possible uses of the research, what their participation in the research entails and what risks, if any, are involved; (3) the confidentiality of information supplied by research subjects and the anonymity of respondents must be respected; (4) research participants must participate in a voluntary way, free from any coercion; (5) harm to research participants must be avoided; and (6) the independence of research must be clear, and any conflicts of interest or partiality must be explicit.9 Ethics committees in the UK are responsible for ensuring that these criteria are met as well as taking into account the interests, safety and well-being of researchers.6

Ethics committees also take into consideration the principle of justice, requiring that the benefits and burdens of research be distributed fairly among all groups and classes in society, taking into account age, gender, economic status, culture and ethnic considerations.6 For example, ethics bodies may point out to funding bodies and researchers the relative neglect, or the over-burdening, of particular groups of research participants.

Before a study is undertaken, researchers should gain formal ethics approval from the appropriate local ethics committee. This may involve submission to multiple ethics committees as, for example, Universities and NHS Trusts may want to independently assess a study. This process of seeking ethics approval involves submitting information on the nature of the project, including the aims and objectives of the project, project time frame, methods of recruiting, participants, recruitment forms, information letters and so one. This process may take some time as often ethics committees have a backlog of projects to approve and may only meet on a monthly basis. Moreover, it is possible, particularly if it is the researcher's first attempt at gaining ethics approval, changes will need to be made and the application resubmitted. This is not unusual and researchers should approach this as a learning process rather than a rejection of the legitimacy of the project.

Privacy and confidentiality are the most obvious, and very important, ethical concerns in social research. Your ethics application should explain very clearly how you will store the project data (if you are university-based, your organisational will have a set policy), including ensuring that participants’ details and their responses cannot be accessed by other parties. Generally, ethics committees ask for data to be kept on a secure computer network (or in a locked cabinet) in the locked office. A good example of why this is important is an interview the current author did with a lecturer whose employer did not realise he had metastatic prostate cancer. He was worried that if they discovered this, his career prospects would be adversely affected. Keeping his details confidential was clearly incredibly important for his relationship with his employer.

Another general concern is the way in which researchers go about recruiting patients. Given the power dynamics between the researched and the researcher, ethics committees will be quite stringent about the ways in which researchers are allowed to recruit. For example, non-clinical researchers may not be allowed to enter a clinic and approach patients. Researchers may have to negotiate with the hospital to get access to patient information and send letters to patients (this may even involve the hospital sending letters on your behalf to protect patient information). This process is about ensuring that patients are not approached and pressured into participation, as well as being a result of increasingly stringent laws regarding the sharing of personal data. An example of this is the author's current study on CAM and cancer, whereby posters advertising the study were placed in oncology wards rather than cancer patients being approached face-to-face in the waiting room. If a patient expressed interest in the study, they were given, by the clinical staff, an information sheet, a ‘consent to be contacted form’ and a stamped self-addressed envelope to send back to the author's university. Even if recruitment does not involve entering a hospital or NHS institution, ensuring that participants have the option of not being approached face-to-care for participation is very important.

Contacting potential participants by mail (ensuring you explain where you got their contact details from and why) and then by phone after a week or so is a good strategy as it means people have had time to think about the project and their potential participation. Often social research involves recruiting gatekeepers who provide access to many other participants (e.g. breast cancer campaigners or patient support group leaders). This gaining of trust is not a license to loosen one's approach to patient recruitment. Rather, high ethical standards should be maintained throughout the project, especially when patients have a high degree of trust due to either the position of the researcher (e.g. being their clinician) or the individual gatekeeper's support for the project.

Gaining informed consent is a crucial aspect of any research project. This does not merely involve a signature on a form. Rather, the process of gaining consent involves numerous elements including (a) a lay-accessible description of the project; (b) an opportunity to ask questions; (c) the opportunity to discuss involvement with friends and family; and (d) description of participants’ rights at each point in the project. Participants should be given a description of the project to keep (including funding source, aims and objectives, and the nature of their involvement) and contact details for either the principal researcher (ideally someone other than the person asking for consent) or the ethics body, which oversaw the approval given to the project. This gives the participant someone to contact if they have questions about the project or if they have a concern about a particular issue related to the project (Fig. 1).

It is vital that the process of getting consent involves the participant having a clear idea about what they are consenting to and when their involvement will begin and end.10 This also means addressing the issue of patients’ expectations of the project versus the actual nature of the research. In the authors’ current research, it was vital that cancer patients understood before giving consent that the study was not a ‘trial’ and that their involvement would not give them access to ‘alternative’ or ‘experimental’ treatments. Nor would the researcher be able to give advice regarding the ‘best’ alternative cancer treatments available during or after the interview (this was often a strong agenda behind participation). This was particularly important given the title of the project – ‘CAM and cancer’ – and the desire of patients to ‘try anything’ when faced with a cancer diagnosis. Informed consent is not merely a process of getting the consent form signed, rather, maintaining consent should be an ongoing part of any research project.

Part of being ethical in recruiting is being aware of a potential participant's ability to exercise choice around whether or not to give consent.10 A good example of this is the current authors’ research in the hospice context with terminally ill cancer patients on their use of CAM. Although these patients are very near death, the chances of them refusing to take part in a study is slim given the desire to ‘give something back’ before they die. In such contexts, the researcher needs to take into account the patient's own situation and whether they really have the option not to take part in the project. If the answer is no, the researcher may need to reassess the burden of being involved in the research versus the benefit to the patient and society as a whole. As a general rule, err on the side of safety and exclude those participants where there is doubt. This same point has been emphasised in relation to minority groups (e.g. women or ethnic minorities) who, due to their collective history, may be prevented from exercising their autonomy.7

Ethnographic research has been the subject of much debate, particularly when participants do not know they are being observed. As well as an absence of informed consent, ethnographic observations can provide the researcher with information that may be personal and potentially damaging for participants (e.g. prognosis in the case of a patient or professional negligence in the case of a clinician). Thus, as is the case for other research methods, an essential aspect of observational research is the act of protecting the anonymity of the participants by presenting observations in reports in such a manner that identities remain undisclosed.

Ethical considerations continue throughout the research process and in particular, are crucial when data are being collected. Although the most cited ethical concerns are confidentiality and informed consent, the way a researcher interacts with patients and collects data are also important in terms of doing ethical research. A good example is that of interviewing cancer patients. The qualitative interview (or observation of clinician/client interactions) is a context where patients often talk about sensitive issues, and in the case of cancer patients, there is often discussion of fear of death, relationships with partners and other incredibly sensitive issues. In this context the researcher is in a real position of power as the patient trusts the researcher to both respond sensitively to their experience, and furthermore, that the information being provided is actually important for the research project (i.e. not putting the patient through something that is unnecessary). Thus researchers should constantly review the degree to which patients are willingly providing information (rather than being drawn along be the researcher), and the degree to which the discussion is appropriate for the research questions. Interviews can be emotional for participants and researchers have a responsibility to limit dialogue to that which is relevant to the project. However, it is also common practice for qualitative researchers to allow the discussion to progress as determined by the participant.1 Thus, a fine balance is needed between exploring the individual patient's experience and ensuring that participants are providing useful and relevant data. More care is clearly needed for very ill patients.

Moreover, the researcher has a responsibility not to over-direct the respondent when employing qualitative methodologies. This means not imposing predetermined issues on the interviewee,11 but rather, working through the participant's experiences—what we call a ‘guided conversation’.12 In this way the data remains true to the experience of the interviewee rather than an imposition of preconceived ideas by the researcher. Thus, being ethical is also about being a good methodologist and ensuring that the data is of high quality and thus, useful for informing policy and practice.

There are also important ethical concerns when writing up and disseminating data. The obvious ethical practice is to avoid divulging the identity of the participants or the institutions where they were recruited from. There are exceptions to these general guidelines, including situations where participants desire to be named in publications so they feel ‘ownership’ over their data.13 In such cases it is imperative that participants realise the exact nature of the way in which their accounts will be presented and in which contexts (i.e. medical journals or public seminars). Moreover, it is usually inappropriate to name any individual mentioned by the participants (i.e. those mentioned in interview excerpts or in ethnographic observations). As illustrated in Fig. 2, divulging the name of clinicians being discussed in the interview excerpt, in a publication, could harm the specialist and or the patient:

The Latin phrase often used in ethical discussions – primum non nocere or ‘first do no harm’ – also applies to the individuals (in case of Fig. 1, a medical specialist) being discussed by the participants. Although, occasionally it may be important to identify the individual the participant is referring to (e.g. their practitioner), make sure this is vital to the point being made, particularly, if it involves negativity toward the individual.

Another important ethical practice is being true to the data as a whole and ensuring that participants’ accounts are represented accurately.16 Whether in relation to statistics from a survey or data from qualitative interviews, there is substantial room for tweaking findings according to the ideas or preconceptions of the researcher (or exaggerating the generalisability of one excerpt) rather than representing the data as a whole. This has been labelled by some as the ‘juicy quotes’ phenomena in qualitative research, whereby a researcher presents single quotes as representing the whole sample, whereas they are merely the most emotive or extreme. Biased reporting of research is unethical.7 Similarly, in quantitative studies, statisticians can manipulate (or select out) statistics to reinforce a particular perspective or choose not to report negative or contradictory results. Being ethical is also about being true to the data collected, accurately representing the accounts of the participants and ensuring that the limitations or shortcomings of the data collected are made visible.16

Given that much research is now funded by external bodies (rather than internal funding from universities) there may be significant pressure put on researchers to get positive findings, or to water-down findings before publication. It is now common practice for publications to be vetted by the funder of the project before they are submitted for review. Although this is perhaps unavoidable to a degree, ethical research means avoiding as much as possible external influence, and secondly, divulging such influences in publications and other dissemination activities.

The safety of the researcher is often forgotten when considering research ethics. Some forms of social research (particularly ethnographies) involve unavoidable risks, but largely, with most methodologies one can utilise particular strategies that ensure researchers are not at risk. For example, there is always some degree of risk to the researcher whenever they are outside a protective environment. Interviewing mentally ill people, for example, is obviously potentially dangerous. Interviewing or surveying people in their own homes also introduces an element of risk to the personal safety of researchers. Frequently researchers go to interview respondents alone and sometimes out of office hours. The nature of research is such that often they have minimal information about respondents before calling on them, perhaps only their name, address and telephone number and that they fall within the scope of the research. A good strategy is to always let a colleague (ideally someone who is also working on the same project to avoid confidentiality issues) know that you are visiting a patient (name, locations, phone number), and secondly, you should always have a means of contacting emergency services. Finally, if you have any reason to believe that a respondent (or anyone else present at the interview location) might present a risk of aggression/violence, then arrange to be accompanied. Universities should already have safety policies in place for staff.18