OrientationObservationIn-depth interviewsDocument analysis and semiologyConversation and discourse analysisSecondary Data
SurveysExperimentsEthicsResearch outcomes
Conclusion
10.3 Confidentiality, anonymity and privacy Confidentiality, anonymity and the privacy of participants in the research should always be considered.
10.3.1 Confidentiality Confidentiality involves ensuring that the data collected about an individual or group is used only for the research and that the material is not made available to third parties. In rare cases this may even require action to ensure confidentiality in the face of official demands for the data, such as a court subpoena.
Mario Brajuha, for example, undertook participant observation as a waiter in a restaurant on Long Island, New York. He risked imprisonment when he refused to violate confidentiality following suspected arson in the restaurant where he had been working. The police demanded his field notes for evidence, as did two suspects in the case wanted the notes for their legal defence. Brajuha refused to releasxe the notes and the controversy only ended when the suspects died two years later and the prosecutor's office abandoned its effort to obtain the notes (Brajuha & Hallowell, 1986).
In another case, with a more detrimental outcome for the researcher, Rik Scarce, a graduate student refused to submit his field notes about radical environmentalists after one of the groups he was studying vandalised a university laboratory. He refused to tell a grand jury what he knew about the group and as a result was jailed for several months (Monaghan, 1993).
One way to secure confidentiality of data, even in the face of a court subpoena, is to ensure that the research data are linked to a code and that the code is separately linked to the respondent. The link file can then be easily deleted or encrypted in some way that makes the link back to the data very difficult to interpret. For example, to guarantee Italian prisoners' confidentiality, masking techniques were used, such as giving each subject a pseudonym and removing any details that would indicate which prison they were from (MacDonald, 1998).
The ethical standards of the American Psychological Association (APA, 2016) suggests that in reports or presentations of research, investigators do not disclose confidential or personally identifiable information concerning their subjects unless the person has given written permission (or unless there is some other ethical or legal authorisation to do so) and that confidential information should be disguised to avoid any potential harm.
10.3.2 Anonymity Confidentiality, thus, usually includes ensuring anonymity of subjects in the publication of the research findings. This normally means that data is only published as aggregates for groups of people identified by general type. Or it may mean, that when individual comments are quoted, the designation is to a general characteristic of the respondent (such as, male; female manager; over 40, female, shopfloor worker) but not so specifically that the individual could be identifiable from the general description (female, ethnic minority Board member).
In some cases, respondents agree to being identified in research publications. It is important to be very clear with respondents what such agreement involves and that once published there is no turning back to anonymity.
10.3.3 Privacy A third element is privacy. This involves ensuring that researchers do not intrude upon respondents' private life and affairs, without consent from the research subject. This also includes being mindful of publication of results that may lead to the research subjects' privacy being compromised and subsequently invaded by third parties following publication: unless, of course, this is the intention of exposé research, which is more likely to be the province of media journalists.
An often-cited example of research with questionable ethical tactics that invaded privacy is Humphreys's (1970) study of casual homosexual encounters, Tearoom Trade. Humphreys was interested in gaining understanding not only about practising homosexuals but about heterosexuals who briefly engaged in homosexual encounters. In addition to observing encounters in public toilets in parks (so-called 'tearooms'), Humphreys developed a way to gain access to detailed information about the subjects he covertly observed.
Acting as a voyeuristic lookout, Humphreys was able both to observe the encounters and to identify participants' car license plates. Humphreys used the license plate numbers to locate the participants home address through the local department of motor vehicles. Humphreys subsequently disguised himself and deceived the people he'd been observing by pretending he was conducting a survey in their neighbourhood. The result was that Humphreys managed to collect extensive information about each of the subjects.
Shortly after the publication of Humphreys work in 1970, there was a considerable outcry against the invasion of privacy, misrepresentation of research identities, and deception commonly being practised during the course of research. Many of the controversies that revolve around Humphreys research remain key ethical issues today. Paramount among these issues are the justifications that the subject matter was of critical importance to the scientific community and at it simply could not have been investigated in any other manner…. In the case of Humphreys study there are many researchers who maintain that the social, legal, and psychological policy changes that have resulted far outweigh any minor invasions of privacy. This is not to suggest that there are not other researchers who argue that the research was unethical no matter how great the benefits have been (Berg, 1998, p. 35).
Suler (2000) has pointed out that now that so much interaction is via cyberspace, issues of privacy are exacerbated:
In this age of expanding access to information, a critical ethical responsibility is recognizing the right to privacy. In chat rooms, e-mail channels, and message boards, it's temptingly easy to gather information without people having any idea they are part of a research project. Even if conceptualized as archival research, naturalistic studies, or participant-observation, a published article or public presentation of the project easily could violate the privacy of an individual or group. It could cause them harm. To prevent these deleterious effects, the investigator needs a solid grounding in ethical thinking.
The converse of privacy in cyberspace interactions is verification of the identity of the research subject and even whether it is a real or imagined persona.
When many online people present minimal, partial or imaginary aspects of their identity, you can't always verify who is who. What should be known about a person's identity before they become a subject? How do you know for sure you are always working with the people you think you are working with? Do imaginary screen names and online persona qualify as confidential information, or should the researcher's report disguise what already looks something like a disguise? It's critical to know whether a subject is a minor, but how do you verify the age of the person? Researchers too can easily hide the fact that they are researchers, or even pretend to be someone else, which is probably acceptable in an environment where everyone else is doing the same thing. Or is it? All of these questions indicate the various twists and turns in online research, so consultation with experts is important. But who are they? Is cyberspace research significantly different from traditional research that it warrants new standards of expertise? Some say yes. (Suler, 2000)
In all research situations, ethical considerations will focus initially on the research subjects. However, researchers should not only take account of the interests and needs of the subjects but also the interests of the wider society. This is the argument made by, on the one hand, researchers/ or journalists exposing malpractice and, on the other, by researchers undertaking covert research aimed at benefitting society, such as those who defended Humphrey's (1970) approach.
This raises the issue of consent (see Section 10.5).
