2.2. Data Preparation and Analysis

Descriptive statistics were used to summarize sample characteristics. After each interview was conducted, the audio recordings were transcribed and verified for accuracy. Once initial participants were interviewed, content analysis was used with the assistance of NVivo 11 software (QSR International, Burlington, MA, USA)). Two members of the research team reviewed and coded the first 10 transcripts and compared notes about preliminary response patterns. This initial set of transcripts was coded both deductively and inductively based on sex communication literature and the emergent data. The lead author coded the transcripts in their entirety and the senior study member supervised and conducted spot checks to ensure quality and thoroughness. Initial response patterns guided the restructuring of the interview guide to more fully explore developing categories such as the nature of first exposure to SEM and the function of sex communication for sons. Twenty subsequent interviews were conducted until the study team agreed that saturation – the point when no new codes were being brought up in the interviews – was achieved. We identified four central themes on sex talk with parents from the perspective of gay, bisexual, and queer males: (1) prompts and triggers of sex communication, (2) parental approaches to communication, (3) sons’ reactions, and (4) the functions assigned by participants to the sex communication process. We relate our sub-themes to the buffer and brake hypotheses in light of participants’ reports of ubiquitous SEM consumption. Other dimensions of parent-child sex communication, such as the frequency of sex talks, parents’ knowledge of LGBT issues, and parental ratings as sex educators have been reported elsewhere [32,33].