Interviewers possess a certain power because they have both taken and been given authority to ask questions (and to expect answers) that most people do not get to ask the subject. The subject can claim or reclaim power by refusing to answer questions, or by lying. The researcher can reclaim some of that power by the fact that getting to serve as an expert can be deeply rewarding to the subject, who will not want to relinquish the role of expert. The interviewer tends to have the power of a prestigious institution such as a university behind her or him, which can shift power to the interviewer, especially in research projects in which there is already a power social or economic differential between researcher and subject. However, sometimes the power that the interviewer holds because of her or his association with an institution can be countered by the power of a subject, such as a subject who is a high-ranking police officer as in the study cited above.

The more aware a researcher is about these power dynamics that exist in the relationship between researcher and subject (as expressed through whatever methodological tools are chosen) the more valid and reliable the data will be that are gathered. The quality of research always depends on the quality of the research design, and in the case of qualitative research in the humanities, part of the research design...

Moving beyond looking at interviews as a collaborative process (albeit one in which the two sides may have different levels of power and commitment to the process) to looking at interviewing as a form of performance.
Such a dramaturgical analysis hearkens back to Erving Goffman, but Denzin extends this analysis by adding in elements of postmodern theory. He argues that both people involved in an interview have a far greater degree of control over how the other person sees them than they realize. Even more than this, each person involved in the interview has a far greater degree of control over the nature of their own reality than they tend to realize.

An ethnographic interview thus, according to Denzin, is a set of parallel performances in which each person is actively arranging their experiences to make their own world meaningful to themselves within the context of their social and cultural world even as they also try to mold the other person's behavior and assessment of self. Seen in this respect, each interview has clear agonistic features to it that, carefully and insightfully managed, can allow for deeply insightful and rigorous research.