interfered with the interviewing process and at times I even forgot they were around (barring the loud voices and shouting at times). As I was there every day and made sure I dealt with issues of informed consent and confidentiality consistently, they eased to my presence and even volunteered to participate in the research after the first couple of days. My youthful appearance further made women more relaxed and more open towards me. Caring for my participants and being attendant to avoiding harm was an ongoing process. Thinking of research ethics as a continual process of collaboration opens up opportunities to dissolve the (mis)conception that ethics approval means ethical research (Halse and Honey, 2005). While I sought not to retraumatize my research participants (Sikweyiya and Jewkes, 2012), the emotive stories of women and continued reflection upon the ethical implication of research in detention had rippling effects to my mental health. Given the lack of information about it and my lack of experience, I was rapidly overwhelmed. At the end of each day, I was exhausted. The lack of natural light for many hours on consecutive days affected my physical health. As soon as I could leave the field, I rushed with feelings of relief and excitement to the analytical phase of the project.25 Somewhat optimistically, I assumed that committing women’s words on paper would matter (i.e., make an impact) and make up for my silences while in the field. Yet as I sat in front of my computer screen re-reading my interview transcripts and fieldnotes about life inside a Greek detention centre, I had the nagging suspicion that the research and articles which would be produced drawing on it would be irrelevant to the women I interviewed. They would never read them, nor would the questions we asked and sought to answer be likely to affect their living conditions drastically. In line with other researchers who describe the transcription process as emotionally draining (Darlington and Scott, 2002; Warr, 2004), I too had strong emotional responses to some of the interviews that were challenging. Therefore, I often did my transcription in public places, mostly cafes, to avoid being alone. Compounding matters, most of my interviews had to be translated in English for my supervisors. This means that I had to listen and reread them numerous times. The research may have been over, yet, it marked a turning point in my life. Working as a research assistant allowed me to save money, introduced me to a new field of confinement Two articles have been produced drawing on this piece of research (see Bosworth et al., 2016; 2018). However, data from this has been used in my later work too (see Fili, 2018a; 2018b). 25 53