2.3. Data Management and Analysis

An experienced transcriptionist transcribed verbatim all the interviews from the audio files that used the language of the participants to best represent the dynamic nature of the living conversation. Transcripts were translated into English and reviewed by the researcher to ensure their accuracy and that no meaning was lost between the transcription and the translation. Thematic analysis was used to analyse data [63]. First, transcripts were read by the researcher to familiarize and immerse herself with the data to the extent of being familiar with the depth and breadth of the content. Second, the initial codes were generated from the data using manual coding of a few transcripts. Then followed the development of the codebook, where the themes were reviewed, refined, and named. The themes were given definitions that determined the essence of what each theme was about and determined what aspect of the data each theme captured. Once the codebook had been developed, consensus about the themes was reached between the researcher and the supervisor. Then, the transcripts were imported into NVivo QSR version 11 (QSR International, Melbourne, Australia), for storage and organization of files such as interview transcripts, field notes, and interview summaries. The findings were presented in themes and quotations that reflected mothers’ views regarding the growth and nutrition of school-age children. To ensure rigor, data was analysed by the author and a qualitative expert to reduce the effect of researcher bias, and to ensure that the interpretations were free from bias and the conclusions were credible. The study employed investigator triangulation, whereby the FGDs were facilitated by the researcher, the supervisor, and the moderator—who is an experienced qualitative researcher. Further strategies were used to ensure that the results were credible, and the researcher maintained a reflective attitude throughout the data collection and analysis process [64].