Conduct research ethically
Ethical issues are assuming increasing importance in research, with most research proposals – even at undergraduate level – needing to be subject to their university's ethics committee and follow a particular code.
On this page, we look first at what constitutes ethics in the context of management and social research, before exploring the various issues that come up throughout the research, from gaining access to an organisation to disseminating the results.
Ethics is about the definition of what is good, and why people choose one form of good over another – in other words, it is about values and value systems.
Research is about moving forward the frontiers of knowledge, and, in the case of management research, hopefully also improving human welfare by helping businesses to run better.
Research ethics, therefore, is about how the researcher goes about the business of advancing human knowledge without doing harm to others. Certain value judgements are made, for example that people have a right to privacy.
Interest in research ethics has grown considerably over the last ten years: the boundaries as to what constitutes acceptable practice have tightened, and most universities will tend to have ethics committees which scrutinise proposals, particularly those which involve the manipulation of people. Candidates for higher research degrees may be expected to prepare an ethics protocol.
Many of the ethical codes developed for research have been done in the context of medicine and science, and are concerned with experimental treatment, control groups, etc. The peculiar characteristic of management research is that it is about people and the way that they interact with their environment. Ethical issues are therefore likely to be more complex, involving subjects in a social rather than an individual setting, and use of a range of instruments based on observation, interviews and questionnaires.
Above all there is a need to balance the behaviour of the researcher and the needs of the research with the rights of the subject. As a minimum the researcher should avoid harm and treat others as they would want to be treated. Preferably, however, where research is done in an organisation, part of the objective should be to benefit the organisation and its employees. Research should be based on the principle of doing greater good than harm.
The researcher needs to adopt an ethical approach from every stage of the research from the initial contact to the dissemination. We shall be looking at requirements in more detail in subsequent pages. First, though, a brief look at ethics committees and some examples of codes.
Most research proposals will need to go through the university’s ethics committee, which will consider such questions as potential harm, sensitive data, full disclosure, anonymity, data confidentiality, need for debriefing, use of any hazardous materials, etc. The committee will also look for a coherent and well thought out design.
Note that ethics committees are fully aware of the dilemmas that the researcher may be faced with, for example the need for confidentiality and concern over harm to respondents needs to be balanced with the potential good of the research. They will want to know, however, that the researcher is aware of the dilemma, and the ways in which he or she proposes to minimise negative effects on participants.
The UK body, the Economic and Social Research Council (ESRC), which funds most social and some management research, has recently implemented a Research Ethics Framework and now requires proposals to take ethical issues into account.
The ESRC has six key principles:
- Research should be designed, reviewed and conducted in such a way as to ensure integrity and quality.
- Both staff and subjects should be fully informed about the purpose, methods and uses of the research, together with any risk. Where this is not possible, the ESRC has very precise guidelines.
- Confidentiality and anonymity must be respected.
- Participation should be voluntary, without coercion.
- Harm to participants should be avoided.
- Research should be independent, and any conflict of interest or partiality should be explicit.
Responsibility for ethics lies with the principal researcher and for review with the institution with which the researcher is associated.
Codes of ethics
There have been various attempts to codify the ethical issues involved in research. An early one is Glass (1966), whose principles are:
- cherish truthfulness;
- avoid aggrandisement at the expense of your fellow researchers;
- defend freedom of scientific enquiry;
- disseminate findings fully.
Russ-Eft et al. (1999) further developed the ethics codes in relations to management science research, which became the basis of the guidelines of the Association of Human Research Development. Their principles involve a professional and responsible attitude towards research, using appropriate means of data collection including adjustment for demographic factors associated with ethnicity, disability etc., informed consent, avoidance of deception except where legitimate for purposes of research, and careful interpretation of findings.
Many of the social science disciplines have their own guidelines, and the relevant websites and their links are listed below.
- Academy of Management
- American Psychological Association
- Academy of Human Resource Development
- British Psychological Society
- British Sociological Association
- Social Research Association
The guidelines of one of the main societies, the Academy of Management, are summarised below:
The Academy's ethical guidelines stipulate the following in respect of research:
- In support of the advancement of managerial knowledge, members should exercise "prudence in research design", and should design, analyse, research and report their findings rigorously.
- Informed consent should be obtained with research subjects, using easily understandable language.
- Privacy and confidentiality should be maintained.
- In reporting on research, members should:
- never fabricate data;
- report findings fully, without misrepresentation;
- present any qualifications to research;
- leave their research open to verification;
- take responsibility for errors, with appropriate steps to address, e.g. retraction, correction;
- report the sources of their sponsorship;
- acknowledge contribution of others in data collection.
- In publishing, members should:
- always be explicit in citing the work of others;
- provide appropriate credit to all authors, for example if a work is substantially based on a student dissertation, the student should appear as principal author;
- submit an article to one journal at time and not submit to a second journal unless the policy permits multiple submissions; if data has been submitted elsewhere, this publication should be cited;
- if journal editors, operate without any ideological or personal favouritism; ensure the confidentiality and anonymity of the review system; ensure that all accepted manuscripts are actually published; and review and respond in a timely manner;
- if peer reviewing, maintain the confidentiality and anonymity of the process, declaring any conflict of interest.
Source: Academy of Management Code of Ethics
Glass, B. (1966), Science and Ethical Values, Oxford University Press, Oxford, UK.
Ethics & access
One of the most difficult aspects of research is one that comes up front: gaining access to the organisation. Businesses are, after all, about profit rather than advancing the frontiers of knowledge, so finding synergy between your research and a willing organisation can be difficult. It is at times like this when it's very important to have an ethical perspective as it can be all too easy to distort the truth or change the project in an unhelpful way.
First, we shall explore in a general way some of the issues around gaining access, which is also dealt with in "How to... use secondary data and archival material".
The first step is to get a foot in the door by using your contacts – preferably someone with whom you have a connection, for example, through your professional association, or an organisation to which you belong such as a church or a social club. If you are studying part time while working, you may have a ready-made scenario, but you should take care that your place of work really can provide for you what you are looking for (see below).
Once you have decided on a contact to pursue, the next thing is to give them an initial call or send them an e-mail. It is important to present a succinct summary of what you are trying to achieve, together with what it will involve for the organisation – what data you will be collecting and over how long a period, to whom you will need to speak, whether you will be able to maintain confidentiality, etc.
You are more likely to be granted access if you can offer the organisation a specific benefit which matches their business needs – for example if you want to research ways of making the supply chain more effective and they have long been aware that this is an area of weakness.
Be sure that you have the right person who can take the project forward internally; if not, ask to be put through to someone who can act as gatekeeper.
Once you have an agreement, put something in writing so that both sides are clear about the project's scope and your role.
Gaining organisational access can be very time-consuming and you should make sure that you have allowed for this in your research schedule.
Access & ethics – avoiding coercion
Because gaining access can be tricky, there is plenty of scope on both sides for coercion.
On your side as the researcher: you should avoid the temptation to plead or be over persuasive. Not only does this violate the right of the other not to be subject to coercion, it is unlikely to gain the willing participation which you are after.
On the organisation's side: they may manipulate you into agreeing to something which is more in their best interests than in that of your research. This may involve you modifying your initial proposal in such a way as to fail to reach the academic outcome you are looking for. All sorts of things can affect the research, for example the sponsor:
- is more interested in finding an answer to a particular question which may be too slight, too specific, and will not allow sufficient scope for grounding in the theoretical background;
- may offer too small a sample, either because a larger one isn't available or because they restrict access to the people they will allow you to meet.
More seriously (because it will be harder to retrieve the situation as more time will have been lost) is when the scope of the project changes because of events in an organisation, or because people change their minds about allowing access to the data.
It is very important that you look at the impact of every change on your research. I once did research into the use of a virtual learning environment in a particular university. Two different pieces of software were going to be trialled, and I was advised just to concentrate on one group of learners. Unfortunately, the piece of software which they were due to use proved incompatible with other systems in the university, and so the group fell back on discussion groups. This meant that it was only possible to look at discussion in a much less sophisticated context, which in turn affected the richness of the findings.
Finally there is the danger that the organisation may not like your findings, and try and control them or dictate outcomes.
Serving two masters
Whatever the vicissitudes of your relationship with your organisation, you need to be aware that you have two different audiences for your report – the organisation and your examiners or academic public. These two audiences will have different expectations.
- The organisation will be interested in practical solutions – how can things make a difference to the business.
- The examiners/academic public will, depending on the level, be interested in your ability to produce a well crafted piece of academic work, and perhaps push forward the frontiers of knowledge.
You are often advised to produce a separate report for the organisation – we shall look at this in greater depth in section 5: "Reporting and disseminating research".
You also need to be aware of the difference between a researcher and a consultant:
- A researcher assumes objectivity and collects data with a view to confirming a hypothesis or looking at findings.
- A consultant gives advice and helps to solve a problem.
Ethical issues may raise themselves as you try and balance these roles: are you getting too sucked into the role of a consultant and thereby losing your objectivity? Are you, simply by doing your research, raising expectations which you cannot fulfil?
Imagine, for example, if you were researching morale in a company which had been restructured. Might there not be a risk that people would use you as a way of articulating to the management their unhappiness?
Ethics & research design
The research design stage, if carefully executed, should anticipate the likely problems in data collection and determine the instruments used; it should likewise bring up the major ethical issues and propose ways of tackling them.
Most ethics committees will scrutinize your research design before giving permission to proceed. It is therefore of great importance that you look at ethical issues as carefully as you look at your data collection methods, and consider ways to resolve them.
We shall look first at two key principles of research ethics before examining issues inherent in particular methods.
Privacy & confidentiality
The right to privacy is the cornerstone of research ethics. You may be the recipient of some highly sensitive information, and people have the right not to have that information disclosed in the wrong circumstances and to withhold information should they so wish.
Imagine what the concerns of your research participants would be in the following circumstances:
- you were carrying out research into people's career aspirations in a particular organisation, and whether or not that organisation met those aspirations with its employee development programme,
- you were looking at the approach of senior management to succession planning.
It doesn't take much imagination to see that in the above circumstances, participants would not want to be identified, and might be cautious about answering certain questions.
The right to privacy thus involves the right:
- not to participate in the research;
- not to be subject to undue pressure to participate;
- to remain anonymous;
- to be approached at times which are convenient (and ideally agreed beforehand);
- not to be questioned for longer than the time previously agreed, i.e. if you have agreed with someone previously that the interview will last 30 minutes, you should not prolong it;
- not to be questioned in a confrontational manner, or subject to questions that give rise to undue stress;
- not to have the scope of the interview widened beyond that already agreed.
Informed consent means that both the participants and the gatekeepers are fully aware of the scope of the research, the exact nature of their participation, and the use to which the data will be put. The following table provides a checklist of the requirements for informed consent. The principle was first developed in the Nuremberg Code, in response to the medical experiments of the Second World War.
Note that if you are interviewing participants under 18, you will need written consent.
|Nature of the research||What is its purpose; who is doing it; who is funding or sponsoring; who will participate i.e. what is the sampling frame; what is the progress of the research hitherto?.|
|What is involved in taking part?||What type of data will be sought; how (e.g. through interview, observation, etc.); how much time will be required and when?|
|What are the implications of taking part and the rights of the participants?||Assurance that participation is voluntary and that participants have the right to say no; they should have the right to control whether or not what they say is recorded; they have the right to withdraw from an interview; what are the risks and benefits of participating; assurance that they will remain anonymous and that data will be confidential.|
|To what use will the reported data be put?||Who will have access to the data? How will the research be disseminated? What are the safeguards on anonymity and confidentiality of participants? What will happen to the data after completion?|
Adapted and summarised from Saunders, M., Lewis, P. and Thornhill, A. (2003), Research Methods for Business Students, Pearson Education.
Informed consent is linked to the principle of honesty in research, and the importance of not deceiving. There may well be times when it is not possible to reveal, or at least not totally, the actual purpose of the research as so doing would skew the data. There are a number of ways around this as far as informed consent is concerned: you can be truthful, but vague and imprecise, or you can rely on the reputation of the educational organisation with which you are working or studying; you can get informed consent afterwards. It is very important in such a case to debrief participants fully after the research is completed.
One experiment involved going up to people in trains and asking them to give up their seats, on the grounds that the interlocutor was someone in authority. The vast majority of participants gave up their seat without complaint. The experiment, which would on ethical grounds probably not be allowed nowadays, was looking at people's reactions to authority. If participants had known this they would probably have responded very differently!
Ethical issues of particular methods
Particular research methods have their own concerns, which we examine below.
Qualitative research – simply by its probing nature and its ability to provide depth – raises ethical issues of privacy, confidentiality and boundaries.
In interviews, the researcher must secure a balance between probing and being pushy and over zealous. Certain issues may be raised by one participant which the interviewer may wish to pursue with others. Care should be taken that in so doing identities are not revealed: it is often better to steer the conversation in that direction rather than quote, even anonymously.
In observation, certain areas should be out of bounds, for example private telephone calls. There are moral problems with covert observation, when participants are not told that they are being observed. This is dealt with further in "How to... use ethnographic methods and participant observation". Debriefing the observed is particularly important.
Quantitative research is considered to have fewer moral problems, as the questions in a survey (for example) are generally less probing. However, there are still issues of privacy, confidentiality and anonymity of data.
If you are carrying out an experiment, you have to consider the effect on your control group of denying them a treatment which may be beneficial. Imagine, for example, that you are trying out the effectiveness of a new educational programme on a particular group. What about the effect on the control group, who are denied this programme?
Data collection & protection
In this section, we look at some further problems with data collection, and at related legislative issues.
When you are actually collecting data, objectivity and anonymity are of the highest importance.
Data must be collected fully and accurately, with everything being recorded. Avoid being selective in what you note.
When you record data from interviews, preserve anonymity by giving your participants a letter or number code. Some questionnaires may ask the respondents to provide information for which the participants may not wish to identify themselves, in which case you can ask them to provide you with their own code (in case you need to go back to them).
Particular problems arise from data collected via the Internet. Online surveys and e-mail are often used as a way of gathering information, however, the following behaviour is considered unethical:
- spamming, i.e. sending e-mails to a large number of people or organisations,
- collecting data through tracking devices – you will not have gained the participants' consent,
- conducting research without identifying the sponsor,
- hacking into an organisation's website,
- contacting people repeatedly – an invasion of their privacy,
- forwarding e-mails to another person – as it is possible to identify the author, it compromises anonymity.
Having to reply in writing by e-mail may also be considered time-consuming and therefore intrusive. So in addition to compromising privacy, quality and reliability of data may also be affected.
"The value of online surveys" by Joel R. Evans and Anil Mathur, Internet Research, Vol. 15 No. 2, 2005 pp. 195-219, looks at the pros and cons of obtaining data online.
Owing to advances in technology and the ability to store data in digital databases long after the use for which it was originally intended, there are ethical problems inherent in data collection. The researcher needs to be aware of the legislative restrictions to obtaining and storing data. European member states are governed by the European Directive 95/46/EC. The UK has the UK Data Protection Act 1998. Many other countries have similar legislation.
Full details of the UK Data Protection Act can be seen on http://www.opsi.gov.uk/ACTS/acts1998/19980029.htm. Its principles are summarised below:
The Data Protection Act 1998
The Act requires that personal data must be:
- lawfully and fairly processed;
- obtained for lawful and explicit processes;
- adequate, relevant and not excessive to the purpose;
- accurate and up to date;
- be kept no longer than necessary;
- processed in accordance with the rights granted to the subjects by the Act;
- be held securely;
- be transferred only to countries outside the jurisdiction of the EC which afford similar rights to those within.
There are certain exemptions for research, but these are limited. Where data are not processed in a way to cause distress or damage to a subject, then personal data may be:
- processed further for research, although subjects may need to be informed;
- stored indefinitely;
- exempt from the provision that data subjects may request information.
There is a particular category of personal data, known as sensitive personal data, which relates to such matter as ethic origin, political or religious opinions, trade union membership, sexual orientation or criminal record. You will need explicit consent to obtain this.
Reporting & disseminating research
You have safely obtained your data, and your sponsoring organisation has not interfered so your results will honestly reflect the findings and satisfy academic requirements. You may be inclined to breathe a sign of relief; however, your moral obligations to your public, your publishers and your sponsoring organisation have not ended. This is where we come to consider the ethics of how you present your research, and how you disseminate it.
Writing the report
Honest representation of data
It is part of your integrity as a researcher that you analyse your data effectively and report appropriate conclusions. There are cases of outright fraud which involve fabricating data; less extreme, but still unethical, are cases where data have been misrepresented, or some data have been conveniently "forgotten", so that false or misleading conclusions are drawn. In some cases, this may be due to pressures from the sponsoring organisation which wishes for a different outcome; however it is very important that you remain objective and do not allow yourself to be influenced.
Privacy, as we have seen, is a critical aspect of research ethics and it is one that must be maintained at the writing-up stage. Particularly if you are reporting interviews or conversations with individuals, it is very important that you don't identify the individual in any way – for example, "a female manager" in a department which does not have many may potentially compromise the person concerned.
Organisations as well as individuals require privacy and may require you to sign a disclosure agreement which will prevent their names appearing in your report: you can get over this by referring to them as Company A etc., but you will need to discuss the implications with your supervisor if appropriate.
The requirements of academic publication often necessitate only brief recording of findings in summary form, so there is no opportunity to cite individuals. For example, John Wood and Tricia Vilkinas in "Characteristics associated with success: CEOs' perspectives" (Leadership & Organization Development Journal, Vol. 26 No. 3, pp. 186-196.) describe how their analysis revealed seven categories of characteristics; further information on the findings is given in tables outside the body of the article.
Margaret Linehan, in "Networking for female managers' career development" (Journal of Management Development, Vol. 20 No. 10, pp. 823-829) refers to people whom she quotes as chief accountant, computer company, human resources manager, computer company: this gives an idea of their status and employment sector without compromising anonymity.
Acknowledging the ideas of others
Of the 73 academic misconduct cases which have been investigated in the UK over the past 3 years (Baty, P. (2006), "Plagiarists face clampdown", The Times Higher Education Supplement, December 8), more than a few involve plagiarism or failure to acknowledge use of material from other sources. You should always provide citations when referring to other people's ideas, whether you are quoting directly or indirectly, or whether you are just paraphrasing.
Disseminating the research
Once you have written the report, there are wide possibilities as to how you disseminate it. A number of quite diverse ethical issues arise here, depending on your audience.
The business audience
If you are doing a report to fulfil the requirements of an academic degree, you may well have two audiences for your report: your examiners and your business sponsor.
To please the former, there is the small matter of satisfying them that you can master the craft of research. The latter will be interested in how your findings can help the business, and herein lies a possible ethical dilemma in that your report may negatively impact on all or part of the business.
You should remember that you have a "duty of care" to the business and its employees. If the report is negative, you should not just dump it on someone's desk and run away; perhaps you could arrange to discuss your findings with a few key people. Be aware of the danger of its getting into the wrong hands and being used against certain individuals; if you think that there is potential for this to happen, then a warning at the data collection stage may not go amiss.
The academic audience
Quite different ethical issues are at stake when you are dealing with an audience of fellow academics. People are unlikely to be unduly sensitive about your findings; dangers and temptations are more likely to arise from a pressure to "publish or perish". Some people adopt a portfolio approach to publishing, re-casting an academic article into a practitioner one for the trade press. There is nothing wrong with this; however, submitting the same article to more than one academic journal is, because it wastes the editor's time and that of the reviewers, and, should the article be published in more than one journal it will take up valuable space.
In the 2004 article, "Self-interest and scholarly publication: the dilemma of researchers, reviewers, and editors", International Journal of Educational Management, Vol. 18 No. 6, pp. 33 -341.), Raymond Calabrese and Brian Roberts analyse a case history of a fictional academic who submits the same article to four journals simultaneously, uses the data collected by students in another article claiming sole authorship, and then proceeds to "mix and match" both sets of data in three further articles.
The authors argue that such behaviour has a negative impact on the mutual trust upon which academic society is based.
If you have made extensive use of other people's data, then those people should be cited as authors. It is usual in the case of multi-authored articles to list authors according to the size of their contribution. Thus in the case of a joint article which is largely based on a research degree, the student's name should come first.
Should you use data which has already been published (by you) elsewhere, you should cite the work concerned.
Editors and reviewers also have moral obligations of fairness, confidentiality and impartiality. These have been identified by the Academy of Management which were summarised at the end of section 1: "What are the ethics of research?".