Self-report vs. direct measures for assessing corn soy blend porridge preparation and feeding behavior in a moderate acute malnutrition treatment program in southern Malawi

This analysis assessed whether caregivers' reports about the amount of oil added to corn soy blend (CSB) porridge were consistent with lab analysis and whether reported sharing of CSB porridge was consistent with direct observation. This was a secondary analysis of a feasibility study assessing 2 programmatic changes in a supplementary feeding program for treatment of moderate acute malnutrition (MAM) in southern Malawi. Intervention groups received standard monthly rations of CSB with increased oil along with social behavior change communication (SBCC) to increase the amount of oil added to CSB porridge and reduce sharing. A control group received the standard CSB and oil ration. Self-reported data collected through structured interviews with caregivers were compared with laboratory analysis of CSB porridge samples and in-home observation over a 5-day period. On average, participants over-reported the amount of oil used in prepared CSB porridge; self-report tended to be closer than the lab-assessed values to the amount recommended in the SBCC. Self-reported and observed sharing appeared consistent across groups. Overall, the self-reported and direct measures showed the same relationships among the groups. Self-report and objective measures were inconsistent but conveyed the same overall message.