Health Reporting and Social Experience: Maternal Reports of Child Sickness in the Developing World

Valerie Lewis, Harvard University

Many social science researchers make use of survey data measuring acute child health symptoms, particularly child fever, child cough, and child diarrhea; often these mothers' reports are used as a direct measure of individual sickness and population level prevalence. However, there is considerable debate over the validity and reliability of maternal reports of child health in the developing world. I review the uncollected literature on self-reported health in the developing world across epidemiology, demography, public health, medicine, and economics. I use data from 35 recent Demographic and Health Surveys to show how the probability of reporting child illness is positively correlated with maternal health knowledge, indicating that maternal knowledge plays some role in reporting illness. The effect varies considerably across countries, however, shedding light on the seemingly contradictory findings from previous single country studies. Caution in the use and interpretation of these kinds of survey findings is discussed, as are implications for theories of perceived health.