Development Engineering: Behavioral Sensing Seminar Series

In developing countries, quantitative social science research relies heavily on infrequent surveys that capture the self-reported behavior of individuals and their families. These data can be biased and unreliable and are extremely costly to collect. Repeated surveying also has been shown to influence households’ behavior and reporting, which limits the utility of high-frequency data.

The Behavioral Sensing workgroup is charged with designing a more reliable data collection platform for development economics. The platform will combine rugged, wireless sensor networks for long term, non-intrusive monitoring of behavior (i.e. cookstove usage) with mobile-phone based tools for collecting and transmitting data. It will enable public health, behavioral, and social scientists to collect data that study participants would ordinarily share through self-reports, but with greater accuracy, reliability and frequency than is currently feasible.

The seminar series features the work of researchers involved in this initiative, which combine expertise in social science and engineering in order to generate comprehensive, demand-driven solutions to development challenges. Organized in partnership with the Institute of International Studies, the series supports the Development Impact Lab‘s (DIL) efforts to support the scaling up of pro-poor innovations with global impact.