Abstract
This paper identifies and challenges some implicit and explicit beliefs expressed by some human subject researchers, namely: the degree to which they can cause emotional harm, and to which they are obliged to ensure the emotional wellbeing of their informants during, or subsequent to involvement in, a research project. The beliefs are sometimes expressed as incontestable fact - simply 'how things are', and 'how they ought to be'.
However, there appear to be many internal contradictions in what some claim, as well as inconsistency between those who write about the nature of research relationships. Moreover, many of these claims are unsupported by evidence or argument.
For example, claiming that the researcher carries the principal responsibility for the emotional outcome of an interview, or that the researcher has greater power to elicit information than the participant has to withhold it, may be underestimating the informant's capacity for self-determination. And even if researchers' personal skills were in some way superior to those of informants, this can never lead directly to the moral claim that researchers should take alternative, preventive, protective, or remedial action. Confusion in this area may be derived from the ambiguity of the word 'responsibility', which can refer to both causal agency or a moral injunction.
Among the issues raised, four in particular call for further discussion. a) What beliefs about relationships are implied by these writers, and are such models healthy? b) What counts as ethical reasoning? c) How are we to decide what needs our ethical attention? d) What could a sociology of research ethics tell us about how particular values are formulated and sustained with the research community?
1. BACKGROUND TO THE PROBLEM
1.1 Introduction
The stimulus for this paper was a chapter by Measor and Sikes (1992) in which they discuss aspects of the relationship between the researchers and the teachers whose lives and careers they were studying. They write that, in life history work, "because the material is intimate, it means that the potential for harm is greater" and "it is intimacy which raises some of the sharpest ethical questions"(Measor and Sikes 1992:210). Such thoughts were not new to me; I had heard much the same from fellow students on a research methodology course. But, just as I found with those students, there seemed to be some confusion.
Measor and Sikes (1992:211) claim that people being researched should be "treated as ... autonomous", and also that "researchers have an obligation to protect people from being managed and manipulated in the interests of research."
First, these two statements seemed to contradict one another. Believing participants can be "managed and manipulated" does not appear compatible with a belief in their autonomy, which I take to mean the capacity for self-determination. Measor and Sikes never define what they mean by autonomy, so perhaps they do see these two claims as compatible. But such ambiguities and apparent contradictions permeate writing about the ethics of research relationships.
Second, any claim made about an obligation that researchers might have towards participants is worth exploring. Obligation is about 'shoulds'. The word 'should' always implies a higher aim. It is about doing something in order that a particular thing which is valued is achieved or safeguarded: 'You should do x so that condition y is fulfilled.' Expressed in this way, statements about obligations are proposals about appropriate means for particular ends. Part of this paper will explore the ends towards which these 'shoulds' might be directed, and whether they are as uncontentious as might first appear.
1.2 Anxieties in research literature
In his research on police deviance, Punch (1989:189) wrote: "I could not escape the realisation that deceit and dissemblance were part of the research role and I did not feel ethically comfortable with that insight." What might be meant by being 'ethically comfortable'? How much can we rely on our feelings to guide our ethical assessment of a situation? What about projection, transference, codependence, all those common human activities that can get us tangled in someone else's personality so that we lose sight of who is responsible for what?
The language of those who express unease about the researcher-participant relationship is rich with emotive terms: 'rape model of research' (Reinharz 1979:95), 'social striptease' (Ball 1983:95), 'coaxing' and 'seducing' (Plummer 1988:16), and 'exploitation' (Measor and Sikes 1992:214). Punch (1994:93) expresses anxiety throughout his discussion, and says how he "became uncomfortably aware of the manipulative element in the relationships built in the field." Although he pulls back from saying what researchers should do, he seems to imply that there is something not right about leaving "the researched behind in the field setting" considering they can experience "acute feelings of abandonment and betrayal (Punch 1994:93)." What is one to do though? Never leave? Of course, the parties may be able to devise a way of parting which is appropriate for both of them, but is this a moral imperative? If so, what is the source of that imperative?
After interviewing 31 mother-students about their experiences of gaining a higher education, Edwards (1993: 193) wrote: "... some of the women I talked to were either in, or close to, tears during parts of their interviews, and certainly I found this disturbing and wondered what on earth I was doing to them and to me." Whatever, Edwards' particular beliefs about relationships, some interesting issues arise from this. The researcher might express concern and provide practical support (on being granted permission by the participant). But I suspect that the researcher's discomfort may be more a focus on self (the researcher) than the other (participant). In other words, the researcher's anxiety may centre around some issue within herself1; perhaps her own societal or familial conditioning that tears should not be shown, or guilt because of a belief that she has caused that person to be unhappy.
Lee (1993:106) issues a note of caution: "Brannen argues that faced with such distress interviewers may want to help but should strongly question their motives for doing so. Such feelings on the part of the interviewer, she suggests 'often have more to do with helping the helper than those who are in need' (Brannen 1988:559). All that may be possible in these situations is for the interviewer to undertake the difficult task of enduring and sharing the pain of the respondent." With this in mind, 'protection' may actually mean the researcher protecting herself from unwelcome reactions by the participant, even though the observable behaviour may seem to be 'looking after' the participant.
All of the above examples of beliefs about the power researchers hold over participants' feelings and behaviour, and the moral obligations for researchers are encapsulated in the following question from Sieber (1992:44): "How can investigators protect subjects from the pain of invaded privacy?" Her book (Planning ethically responsible research: a guide for students and internal review boards) holds many assumptions: that pain is unacceptable, that participants must therefore be protected from it, and that it is the moral obligation of researchers to do the protecting.
1.3 Some claims made in the social research community
All those quoted so far in this paper have published their concerns after conducting their research. To what extent do the anxieties expressed after fieldwork effect the fieldwork itself, and to what extent are we seeing attempts at politically correct statements which do not reflect the inner state or actions of the researchers during their fieldwork? I have no evidence to be able to answer that. However, it seems reasonable to believe that such stated beliefs are reflected in some way in the field. For example, here is a statement by someone in a qualitative research e-mail discussion group:
Treating others with respect means that we do not act in an oppressive manner. This means that we do not manipulate, cajole, or coerce participants in order to obtain data. Because our relationship with participants is not parallel, our obligation in this is greater than it would be in a parallel relationship (Dawson 1996).
Clearly, under certain circumstances, this researcher does feel responsible for the participants' emotions; she asserts certain obligations on the part of the researcher without stating the origins of these; and it looks like she believes that the researcher has abilities superior to those of the participant. And all of this is stated as though it is fact, rather than opinion. It might be useful if such claims, which, I suspect, are not unusual in the social research community, could be supported by evidence and argument.
Another contributor to the discussion list wrote: "If two people were having a dialogue and one got distressed as a result, then there is no 'obligation' on either party to deal with that person's distress (Frohmader 1996a)." This was described as a 'natural' situation. The writer then continued: "However, in a research setting, the situation is not natural - the researcher has 'constructed' that situation for their own purposes (ie their research) and they have asked the participants to 'help' them to achieve this (Frohmader 1996a)". Later, she wrote:
It is not difficult to find similar claims in research literature that the researcher has greater power to elicit information than the participant has to withhold it.
It seems that such claims as these lead to certain 'shoulds' about researchers' behaviour. For example Sieber and Stanley (1988:54) claim that one should balance freedom of enquiry against "obligation not to harm persons or institutions in society". It would make little sense to issue such a moral injunction unless it was deemed possible to inflict harm; and, presumably, harm can only be inflicted if a participant is not sufficiently protected. At this stage I have been careful not to indicate who might be responsible for affording that protection - a supervising body, the researcher, or the participant, or some combination of these.
Measor and Sikes (1992:226) declare that "we should not initiate situations that we are not prepared to see through to their potential conclusion." In the context of their argument, this is clearly a moral 'should', referring to what is right, proper and good for the participants, rather than simply to what might be useful for the research. A research methodology text states: "The principle that subjects ought not to leave the research situation with greater anxiety or lower levels of self-esteem than they came with is a good one to follow (Cohen and Manion 1994:371)." If this is not a moral precept, then it is clearly driven by one. Cohen and Manion (1994:377) also state: "The ... principle ... of equal respect, demands that we respect the equal worth of all people. This requires us to ... regard [people] as free and rational ..." This is puzzling. If people are free and rational, then how could researchers not treat them as such?
1.4 Common escape routes
Although it seems almost fashionable to show concern about exploiting participants, it is also not uncommon to find the same writers letting themselves off the hook: Ball (1985) says that in extreme cases, the participant can terminate the interview; Butt et al. (1992:93) protect themselves against any charge of 'practising therapy without a licence' by stating that they simply did not intend to, and that "individuals have a sense as to personal readiness for self-disclosure regarding particularly personal topics"; Sieber (1992:47) asserts that "an individual's right to privacy from research enquiry is protected by the right to refuse to participate in the research," and one can only assume that the individual's autonomy extends to the interview itself, meaning that he could withdraw or refuse to answer.
This is the view taken by Measor and Sikes (1992:230). They express considerable anxiety about manipulating their participants by the judicious sharing of details about themselves to give the impression of reciprocity and break the ice in the interview; but they then conclude: "informants are not fearful victims who open up their lives and souls because they are told or asked to. People have boundaries and strategies to protect themselves in research situations." As for the researcher's side of the situation, as Lee (1993:110) suggests "... interviewers have few sanctions which they can deploy should a respondent decide not to accept their definition of a situation." Perhaps it is useful to remember that asking for the disclosure of personal information is not the same as getting it; the choice to disclose remains undeniably with the participants.
Another angle found in the literature is that one should not be too concerned about the disturbance that can arise in peoples' lives through engagement as research participants because of the therapeutic element of self-disclosure to a willing listener, or the opportunity to explore problematic situations. In some life history work with teachers, Kompf (1992) noted how some participants initially saw the process as quite threatening, to the extent that one teacher even burned the manuscript ritualistically page by page at the end of the research project. He reassured himself by adding that, apparently, 'most' teachers expressed how the process had made them see themselves afresh, and gain new insights into what their past meant to them, and this often meant that they gained a new professional impetus.
Sieber and Stanley (1988) refer to the famously controversial experiment by Milgram in which he deceived research participants so that they believed they were inflicting lethal electric shocks on people. Many commentators believed that the research was deeply disrespectful and damaging to the participants. However, Sieber and Stanley (1988:52) point out that: "... when Milgram (1977) conducted a one year follow-up with his research subjects, he found that less then 1% of his subjects regretted that they had participated in the research and most were glad that they had. Although this result, in part, may reflect a rationalisation by the subjects, it is important to note that they did not feel harmed." Sieber and Stanley's statement is not quite accurate: what Milgram established was that these participants reported that they did not feel harmed. Is this sound evidential basis for establishing a) whether participants are harmed, or b) what counts as being harmed?