Given that the project dealt with respondents from contexts characterized by high repression and extensive surveillance, security was a core concern throughout the research process. The protection of communications and research data was prioritized in order to minimize potential risks of harm.15 Great care was taken to create a safe space for conversation and limit intrusions into the professional and personal life of respondents. To obtain informed consent, respondents were notified from the outset about the purpose and background of the project. In doing so, a written consent form was not used to avoid creating an additional paper trail linking respondents by name to the project and scaring off potential respondents reluctant to sign an “official” document. All interviews were conducted in-person or via a Voice over Internet Protocol (VoIP) channel. Respondents were allowed to choose the application for online interviews, providing a first glimpse into their security behavior (e.g., the decision to use of applications considered to be more safe than others, such as Wire vs. Skype). Interviews were recorded only after the express consent of respondents was obtained. Transcriptions and notes resulting from interviews were anonymized and stored in encrypted documents. When using interview material and quotes in the written findings, the researcher removed all identifying information, such as names of locations or organizations, while seeking to preserve contextual depth. All interviews are referenced anonymously with an archival number and the month in which the interview was conducted. The selection of respondents has three potential drawbacks. First, due to the use of referral sampling, the researcher was only able to approach those respondents who were still active in their respective field. Any former activists, who may have quit as a result of regime pressure, were excluded. Second, with only two exceptions, none of the respondents had the intention to travel or return to their home country in the near future. This inclination meant they were potentially more immune to threats from the home regime. Third, some respondents may have had an interest to exaggerate threats to emphasize the relevance of their activities; others may have downplayed certain risks in accordance with their own “mental models” of security. The information gained through this research approach is therefore somewhat subjective and risks distorting the scale, pervasiveness, and impact of the repressive practices targeting diaspora communities. Nonetheless, the sample size of respondents was large enough to reach a level of saturation and identify recurring patterns across all three country cases. Furthermore, as a qualitative study, the research did not aim at a statistical solid representation but rather an in-depth description of the phenomenon of transnational repression and its repercussions. Finally, initial findings and draft versions of this report were shared and discussed with a number of respondents to correct misperceptions on the part of the researcher and give them a voice in the representation of their experiences. 15 In engaging with respondents from high-risk and activist contexts, the research for this report relied, in part, on the following sources: Glasius, M. et al. (2017). Research, ethics and risk in the authoritarian field. Palgrave Macmillan; Kazansky, B., Torres, G., Van der Velden, L., Wissenbach, K., & Milan, S. (2019). Data for the Social Good: Toward a Data-Activist Research Agenda. In: Daly, A., Devitt, K., & Mann, M. (eds). Good Data. Institute for Network Cultures, Amsterdam. http://networkcultures.org/wp-content/ uploads/2019/01/Good_Data.pdf. 9