The first author transcribed all recorded interviews in Dutch using F4Transkript and trained and monitored research assistants on Curaçao and in the Netherlands in transcribing and translating recordings in Papiamentu. The first and second author organized, coded, and analyzed the interview data in ATLAS.ti 8 and identified initial codes both deductively and inductively (Braun & Clarke, 2006). More specifically, we developed a coding scheme based on, on the one hand, the interview questions from the topic list. An example of a deducted main analytical category was “Bond between respondent and absent father,” coming from the verbatim interview question, “What is the contact/bond between you and your father like?” In addition, the coding scheme consisted of inducted codes based on responses of respondents. An example of an inducted code was “Respondent knows little about absent father,” which was identified based on responses showing that respondents hardly knew any details or characteristics about their father. Based on the first five interviews, the first author proposed a preliminary coding system to the second author. After some adjustments based on feedback from the second author, the first author trained the second author and a third coder in using this pre-agreed coding system. Subsequently, all three coders independently coded the three same interviews and met on several occasions to discuss their codes and to resolve discrepancies. After all three coders coded the third interview, we measured intercoder agreement for all our semantic domains (Krippendorff’s α = .803; Friese, 2020). Since codes overlapped in the vast majority of cases (80%), the first author continued to code individually. When the first author had coded all transcripts, the second author reread several of them to check whether they agreed to the coding. In the few cases where the second author disagreed, the first author carefully reexamined the coding and made a final decision. We ordered all statements of respondents belonging to a specific code corresponding to the frequency of occurrence and systematically compared, weighted, interpreted, contrasted, and eventually synthesized them (Silverman, 2006). Based on the constant comparison method (Corbin & Strauss, 2014), we synthesized these codes into themes, which we describe as follows.