Estimating Adult Mortality Using Sampled Social Network Data: Evidence From Brazil

Dennis Feehan , University of California, Berkeley
Matthew J. Salganik, Princeton University

Measuring adult mortality is fundamental to science and policy, yet hundreds of millions of people are affected by the scandal of invisibility: they live in places where death rates cannot be directly measured because most deaths are never formally recorded. Developing methods to estimate death rates has been challenging in part because it is rarely possible to collect data needed to estimate death rates in an environment where a gold standard is available for comparison. To help address this challenge, we conducted household surveys in 27 Brazilian cities (n ˜ 25, 000) to empirically validate several leading approaches to estimating adult death rates, including the new network survival method, the sibling survival method, and models based on those two methods. Our study contributes to understanding both how data about deaths should be collected and also how models can be used to improve the accuracy and precision of estimated death rates.

See extended abstract

 Presented in Session 219. Statistical Advances in High Resolution Modeling for Health and Mortality Outcomes