HP has highlighted the methodological and ethical dilemmas that arise when conducting research among socially marginalised groups, as well as the problems of independent researchers who must grapple with these issues without institutional support. Her preparations getting to know her informants, their cultural practices, social organisation and life-styles before launching into formal interviews are laudable.

HP was candid enough to inform potential informants that no concrete benefit would accrue to them out of the research, since the data would mainly be used for policy formulation. But she also said she would do her best to help them get access to health services, if needed. Further, she did help some informants in the course of the research. In effect, potential participants were promised a concrete benefit: help accessing health services.

Such assistance may help fulfill the goal of equity because participants are receiving something in return for their co-operation. There also appears to be a natural synergy between the data HP collected and the kind of assistance she offered. But making such offers in advance may become an inducement to take part in the research. The many expectations expressed by research participants, which HP describes as frustrating, could result from her initial offer of assistance, which she made in all good faith.

The researcher is surprised at being approached by one of her interviewees for help filing a harassment case. Such expectations escalate when researching lower-income marginalised groups over an extended period of time, and it may not be enough to restate boundaries. This problem is more severe for the independent researcher. In such situations, the researcher can offer information and network with other organisations which the person in need can contact. For instance, HP could have referred this person to the legal aid cell of an NGO in the city. This could also have been done when demands were made for livelihood options and loans. In the case of medical information on feminisation, the researcher need not have hesitated to provide the information. Informants know that one person cannot solve all their problems. As in the case of other relationships, integrity should generally override other considerations.

HP was very mindful of following principles of confidentiality and privacy. But her reasons for not seeking written informed consent are not convincing. It is a misconception that written informed consent is required only when medical tests are performed on participants, or that non-literacy and multiple languages prevent one from collecting written informed consent. A copy of the informed consent form containing a contact address and the sponsoring organisation’s name can be given to informants to keep. A well drafted consent form not only aims at ensuring that the researcher is following conventional ethical guidelines but also gives informants a way of contacting the researcher after the research is over, should they wish to do so. Documentation is an important way of formalising relationships and ensuring accountability, which carries equal meaning for both literate and non-literate persons: the only difference is that extra effort has to be made to effectively communicate information in the case of the latter.

HP admits that no formal comprehension tests were conducted to assess her informants’ level of understanding about the research and informed consent process. The purpose of the exercise is lost if we don’t know informants’ level of comprehension.

Accessing informants to participate in the research appears to have been a difficult task, given the closed nature of transgender communities. HP had to strategise in approaching potential informants, while at the same time not alienating their gurus. It appears that at one level the researcher tried to bypass gatekeepers (gurus) by contacting potential research subjects independently, but at another level she could not escape the gurus’ influence. Negotiating such close-knit hierarchical networks was undoubtedly very trying. By excluding two persons who wanted to participate in the research due to their guru’s refusal because she anticipated risk to them, she chooses to honour the principle of non-maleficence over personal autonomy. It is likely that these persons knew the risks of going against the dictates of their guru and still wanted to participate. The fact that the researcher expresses uncertainty about the course of action she chose highlights the complicated nature of this particular ethical dilemma.

Ethical dilemmas are not always clear cut, since they implicate personal emotions and value judgements of researchers. While carrying out methodologically viable and ethically sound research, researchers need to learn to deal with their own perceptions of the right and good as also their anxieties of rejection, fear and guilt.