Typed notes were entered into a database using NVivo 9 software, which supports qualitative data management and analysis. A code book was developed based on the discussion protocol and was refined throughout the coding process. We used the software to apply codes to the transcripts. At least two transcripts in each sample were coded independently by three members of the research team. Team members then met to reconcile differences in coding through discussion and consensus. After achieving a shared understanding of the coding scheme, research analysts coded the remaining discussion notes with review and approval by a second coder.
Analysis included the following six steps: (1) review of all data extracted by code; (2) estimation of relative frequency of topics mentioned within thematic codes; (3) assessment of primary patterns and trends within the themes; (4) identification of illustrative quotations; (5) summation of primary themes; and (6) discussion of similarities and differences in themes between parent groups. One team member summarized data for each code and generated themes. The research team discussed themes to clarify, confirm, refine, or elaborate them and consider implications. The resulting themes were used as the basis for this report. Although strict frequencies are not possible in qualitative data collection when discussions are semi-structured and not all respondents are asked the same questions, we apply the following general scheme when defining the relative rate of mention of topics: For the group of nine parents, "few" refers to mentions by at least three parents; "several" refers to mentions by at least 4-6 parents; "almost all" refers to mentions by at least 7-8 parents. For the group of six parent-leaders, "several" refers to mentions by at least 3-4 parents, whereas "almost all" refers to mentions by at least five leaders.