An epidemiologist is planning a study to investigate the association between maternal folic acid supplementation and neural tube defects (NTD) in a low-resource setting in rural India. She has identified 300 women who delivered infants with NTD and 300 women who delivered unaffected infants, and will collect data on folic acid use during pregnancy. However, recall bias is a concern because mothers of affected infants may over-report or under-report supplementation differently than mothers of unaffected infants. What is the most appropriate next step to minimize this bias?