The role of a qualitative researcher is to collect, analyze and interpret data, and report findings for the purpose of increasing understanding about the phenomenon under study [1]. However, within the context of an interview, it may become difficult to maintain that primary identity or role of researcher. This problem has been well documented in the nursing literature on qualitative research (e.g. [16, 17]). In a study of family caregiving conducted by nurse-researchers, multiple roles and relationships between the researchers and the participants emerged, including: stranger-stranger, researcher-participant, friend-friend, nurse-client, and guest-host [13].

Nurse-researchers may experience role conflict when deciding how to introduce themselves to participants. Interviewers are aware that the initial introduction will influence the participant's perception of them [14, 16, 18]. For example, if they identify themselves as 'nurse-researchers', then participants with positive past experiences with nurses may feel comfortable sharing intimate information to which others may not have been privy [19]. Conversely, if participants' experiences have been negative then they may purposefully omit information, especially if they are aware that the nurse has a professional responsibility to act on any disclosed information. If participants are unfamiliar with the process of research, then they may view the interviewer in the more familiar role of 'nurse' and attempt to focus the interaction towards their clinical concerns [3, 20]. In a study to explore the influence of professional role on qualitative interviews, Richards and Emslie [21] compared researcher-participant interactions in two different qualitative studies about middle-aged adults’ perceptions of heart disease. One study was conducted by a sociologist who introduced herself as a researcher with no medical background; the second study was conducted by a General Practitioner (GP), who identified herself as a non-practicing physician. In the study conducted by the sociologist, participants expressed views that were more critical of physicians, whereas in the interviews conducted by the GP, some participants were apologetic about expressing negative views about physicians and several took the opportunity in the interview to ask the physician medically related questions. Richards and Emslie [21] conclude that disclosure of the researcher’s professional role and the participant’s perceptions of that role have the capacity to influence the qualitative data shared. On the other hand, concealing one's role as a clinician may create both personal and ethical concerns for some researchers [5, 8].

An extra layer of complexity exists if the researcher conducts research interviews with his/her own patients. Hamberg and Johansson [7], family physicians who interviewed their patients about long-term musculoskeletal pain, found that interview situations were filled with tension and that power asymmetry with participants was extreme. Both Britten [22] and Archbold [20] advise that clinicians should not interview their own patients because they may feel compelled to participate, fearing that refusal might jeopardize the care or treatment they are receiving.

The ability to effectively communicate and interpret data on multiple levels is a skill that contributes to both effective in-depth interviewing and the delivery of professional health care [5]. However, even though the goal of a research interview is to generate knowledge, the act of participating in this process may have therapeutic benefits for the interviewee. For some people, especially those who are vulnerable and marginalized, the experience of telling their story has been described as empowering, cathartic, healing and an opportunity for self-reflection [23, 24].

A different situation exists when a nurse-researcher decides to intervene or is asked to provide health information during an interview. In a study about the transition to motherhood, Oakley [15] identified that over the course of 178 interviews, she was asked 878 questions and that 76% of these questions were requests for health information. Several authors argue that while health care providers are socialized to care and provide service, the primary objective of a research interview is to collect data and not to offer intervention [10, 19].

The impact of providing clinical interventions during in-depth interviewing can be interpreted differently depending upon the researcher's philosophical assumptions about the nature of the researcher-participant relationship. From a post-positivist perspective, Hutchinson and Wilson [25] caution that interventions made during an interview threaten the validity or objectivity of the data. Information provided by the nurse-researcher may influence the participants’ responses, change the focus of the interview, discourage the participant from openly sharing more information or prematurely terminate the interview [5, 10]. In spite of this, refusing to answer clinical questions or concerns may negatively affect the interview [22]. One strategy to deal with these issues is to defer requests for information until the end of the interview [22]. When a need for further intervention is identified, then the nurse-researcher should refer the participant to another health care professional [10].

In interpretative or critical paradigms the subjective nature of the researcher-participant relationship is highlighted. Therefore, intervening during an interview is seen as enhancing rather than threatening the validity of the data [15, 26]. Rather, it is this reciprocal exchange of information that builds rapport and trust so that more meaningful information is ultimately shared [11, 27]. Offering of health-related information is also seen as a positive reward that offsets the burden of participating in long interviews that are often physically and emotionally exhausting [15, 28]. Wilde [4] hypothesizes that intervening may even open up new areas for inquiry and exploration. Additionally, qualitative researchers who apply a feminist methodological lens to the research process identify that the research process can be utilized as a vehicle to empower change at individual and system levels [11]. In a feminist grounded theory study to explore family health promotion practices by mothers with a history of intimate partner violence, Ford-Gilboe, Wuest and Merritt-Gray [29] identified that in addition to providing information on community resources at the end of the interview, they also empowered the women to access those resources and in some cases advocated directly on the behalf of the study participant.

Regardless of the researcher's paradigmatic stance, there are certain circumstances when the clinician has a legal and ethical responsibility to intervene and perhaps even to stop an interview. If a serious conflict should develop, then the need to provide a therapeutic intervention should take precedence over the need to collect data [30]. For example, if during the course of an interview, the researcher learns or suspects that a child may be in need of protection, the interviewer must report this information to a local child welfare agency. The nurse-researcher also has a professional responsibility to intervene if there is an immediate threat to an individual's health and safety, such as respiratory distress or cardiac arrest and she/he has the professional skills to cope with such a threat. The nurse-researcher may also feel a professional responsibility to intervene if she is interviewing an individual in the community who has limited access to health and social services, or who may not have the skills to follow through with a referral to another health care professional.

To facilitate the implementation of ethical research practices in qualitative research, Hewitt [18] has developed a comprehensive list of the components of ethical researcher-participant relationships. This framework provides researchers with a description of the factors of the research relationship that should be addressed in order to ethically acknowledge bias, maintain rigor, establish rapport maintain respect for participant autonomy and confidentiality and how to avoid exploitation.

In qualitative research it is imperative that the researcher explicitly identifies the paradigmatic approach in which the research design is situated. The assumptions associated with a particular paradigm will define how a researcher interacts with participants. Related questions that need to be raised include: What is the form and nature of reality? What is the nature of the relationship between the nurse-researcher and the participant?

It is important to explore with the participant their perceptions and experiences with past care from health and social service providers. The participant's perception of the agency sponsoring the research should also be examined. The nature and quality of their previous experiences with staff from that agency also needs to be understood. Particularly in program evaluation studies, if the nurse-researcher is a part of the institution where the participant is receiving care then honest opinions may be suppressed [33].

It is impossible to reduce all inequalities of power, but researchers must protect participants' rights to anonymity, confidentiality and reduce any psychological, physical or social risks associated with participating in an interview [24]. To decrease the imbalance of power, some authors advocate for mutual self-disclosure, or the sharing of personal values, beliefs or opinions to emphasize shared experiences and minimize differences [4, 15]. When the researcher expresses opinions there is a risk that the participant will choose to agree with the researcher's conclusions, but Hamberg and Johansson [7] suggest that sharing opinions and interpretations actually opens up opportunities for the participants to react, protest or modify their responses. If the researcher chooses to disclose personal information, how much is appropriate? Self-disclosure and the intimate content of many interviews may turn the researcher-participant relationship into a friend-friend relationship [3, 15]. Does this enhance the quality of the data or does it create an increased risk of exploitation [11]? How does the development of a friendship affect the reporting of research findings?

To minimize role conflict, researchers need to be able to clearly define and articulate their roles to participants. Careful consideration must be given to how one presents him or herself to the participant: is it as a graduate student, researcher or nurse-researcher? The participant's beliefs about the role will influence what information is disclosed [14]. For example, if identified as a nurse-researcher, what is the participant's perception of a researcher and of a nurse? Different factors may influence an individual's decision to interact with a nurse versus a researcher. If the participant perceives the nurse-researcher in the more familiar role of 'nurse', May [34] argues that this is an obstacle in obtaining informed consent. However, it has been discussed that it is often a challenge in interactions between nurse-researchers and study participants to suppress acknowledgment of the ‘nurse’ identity [35, 36]. Colbourne and Sque [35] conclude therefore that the researcher’s identity as a nurse should be revealed but that the impact of this revelation on the data be reflected upon.

The decision to intervene will be determined by the underlying assumptions of the research design and by who identifies the need for intervention and the immediacy of the need. If a life-threatening situation arises, then the nurse- researcher must intervene immediately; other interventions could probably be left until the end of the interview. Sponsoring agencies should be aware that clinician-researchers might provide clinical interventions during an interview. If this is the situation, it is the researcher’s responsibility to be aware of the agency’s policies and procedures, have a method of reporting interventions if necessary and ensure that they have adequate insurance coverage for their professional interventions.

If the decision has been made that interventions will not be offered within the interview, then the researcher should be knowledgeable of community resources to which to refer the participant or be able to leave a package of pertinent information with the participant. If there is the potential that the content of the interview will trigger a negative emotional response from the participant, then the researcher has an ethical duty to have a counselor or other services available for follow-up. It would be important to identify if the sponsoring agency has the resources available to respond to an increase in workload in such situations.

Any decisions to intervene should be documented in the researcher's field notes and described in research publications [27]. Additional questions for reflection could include: Did the intervention open up alternate areas for inquiry [4]? What effect will intervening have on the researcher's relationship with other clinicians or the participant's relationship with his/her primary health care provider? Does the researcher have an obligation to follow-up and evaluate the effectiveness of the intervention?