Quantitative validation studies alone may not be able to distinguish between instances when participants did not accurately report an event vs when participants did not understand a question. We used an explanatory qualitative study design to acquire an in-depth understanding of why some mothers in rural Nepal overestimate birth size of their newborn and their length of pregnancy.
We conducted two focus group discussions (FGDs) with study staff who administered a quantitative questionnaire and 12 in-depth interviews (IDIs) with mothers who had participated in a quantitative validation study. Transcripts were coded and analyzed for themes in patterns of meaning within and across FGDs and IDIs. Using this thematic map, we synthesized our data into common and divergent responses from participants to facilitate our interpretation of the quantitative findings.
We identified five themes specific to this analysis: difficulties with the length of pregnancy question, challenges in administering the birth size question, the perceived effect of time since birth on mothers’ ability to remember information, the language and style differences specific to this setting, and the study context shaping the relationship between study staff and mothers who participated and how this may have influenced mothers’ responses. Visual aids may help to scale the question about birth size within a cultural frame of reference for maternal reports to be more interpretable. Among both study staff and mothers, a longer period of time since the birth of a child was thought to be associated with diminished accuracy of maternal reports, a perception not supported by our previously published quantitative findings.
Poor validity of low birth weight (LBW) and preterm birth indicators based on maternal reports may be partly attributed to challenges in maternal understanding of questions assessing birth size and length of pregnancy. Additional research is needed to confirm these findings regarding maternal comprehension and to further evaluate the utility of visual aids developed for this study.