This article offers an account of a doctoral insider research project that became problematic. The project was investigating mathematics teaching in a university in the UK, and by contrasting the research account with research diary entries pertaining to two interviewees, different interpretations of the interview data are evident. These differences offer illustrations of four aspects of the problematics of interviewing as follows: (1) personal relations and expectations position everyone in the interview; (2) the motivation for the research affects what the researcher learns; (3) the same material generates accounts that emphasize different things; and (4) things happen in people's heads during the interviews that are not recorded. The argument developed by considering the examples is that authorial voice is constructed out of decisions regarding the data together with considerations regarding the researcher's position. The validity of insider research requires reflexive consideration of the researcher's position, and this is especially pertinent in the case of research undertaken by practitioner researchers on professional doctorates. (Contains 6 figures.)