When I arrived at Nakivale Refugee Settlement, Uganda, in August 2013, I came with a mattress, a blanket, a bucket and some items needed for my seven-week stay in the settlement. My colleague and I each rented rooms in a small hostel run by a Congolese refugee family, and I looked forward to an ethnographic approach to our fieldwork. I had many expectations for the real-life perspectives I would learn from interviewing and observing the refugees there but was unprepared for the dilemmas my presence in the settlement would engender. The research aim was to learn how interactions between the refugees and the community that received them shaped the livelihoods of both. However, during fieldwork, I often found myself in a context in which the refugees perceived me as part of an international community with authority to bring forward their cases. Less than a day after our arrival, rumour in Nakivale spread like wildfire: “Norway is back to select more refugees for resettlement”. I had not anticipated how my qualitative interviews could create the refugees’ expectations for resettlement to a third country (or help their current situation in the settlement) or how they consequently would openly share vulnerable stories of sexual violence without my probing.

The fieldwork challenges described in this chapter have methodological implications for how to analzse collected data (discussed in Bjorkhaug, 2017). Here I argue that, in settings like in Nakivale where sexual exploitation may easily become the dominating subject even when this is not our study's focus or aim, we have to be prepared to read and interpret how actors employ such stories to navigate in a competitive terrain, reconstruct the meaning of their public stories in the wider context and thus understand the power politics of narrative(s) across different groups. This chapter's contribution is that it discusses how researchers can implement fieldwork, as well as self-care, in a complex context by addressing three dilemmas: (1) how the research is situated in a wider context we cannot necessarily influence, (2) how participants can be strategic in their responses to the researcher and, therefore, potentially influence data collection, and (3) the researcher's role in the field when we listen to others’ distress and find ourselves struggling to provide a response we are not prepared to give.