Stephen J Mooney1,2, Michael D Garber3. 1. Department of Epidemiology, University of Washington, Seattle, WA. 2. Harborview Injury Prevention & Research Center, University of Washington, Seattle, WA. 3. Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, GA.
Abstract
PURPOSE OF REVIEW: The 'big data' revolution affords the opportunity to reuse administrative datasets for public health research. While such datasets offer dramatically increased statistical power compared with conventional primary data collection, typically at much lower cost, their use also raises substantial inferential challenges. In particular, it can be difficult to make population inferences because the sampling frames for many administrative datasets are undefined. We reviewed options for accounting for sampling in big data epidemiology. RECENT FINDINGS: We identified three common strategies for accounting for sampling when the data available were not collected from a deliberately constructed sample: 1) explicitly reconstruct the sampling frame, 2) test the potential impacts of sampling using sensitivity analyses, and 3) limit inference to sample. SUMMARY: Inference from big data can be challenging because the impacts of sampling are unclear. Attention to sampling frames can minimize risks of bias.
PURPOSE OF REVIEW: The 'big data' revolution affords the opportunity to reuse administrative datasets for public health research. While such datasets offer dramatically increased statistical power compared with conventional primary data collection, typically at much lower cost, their use also raises substantial inferential challenges. In particular, it can be difficult to make population inferences because the sampling frames for many administrative datasets are undefined. We reviewed options for accounting for sampling in big data epidemiology. RECENT FINDINGS: We identified three common strategies for accounting for sampling when the data available were not collected from a deliberately constructed sample: 1) explicitly reconstruct the sampling frame, 2) test the potential impacts of sampling using sensitivity analyses, and 3) limit inference to sample. SUMMARY: Inference from big data can be challenging because the impacts of sampling are unclear. Attention to sampling frames can minimize risks of bias.
Entities:
Keywords:
Big Data; Research Methods; Sampling; Sampling Frames; Secondary Data
Authors: Miguel A Hernán; Mara McAdams; Nuala McGrath; Emilie Lanoy; Dominique Costagliola Journal: Stat Methods Med Res Date: 2008-11-26 Impact factor: 3.021
Authors: P S Mead; L Slutsker; V Dietz; L F McCaig; J S Bresee; C Shapiro; P M Griffin; R V Tauxe Journal: Emerg Infect Dis Date: 1999 Sep-Oct Impact factor: 6.883
Authors: Dustin T Duncan; Mona Sharifi; Steven J Melly; Richard Marshall; Thomas D Sequist; Sheryl L Rifas-Shiman; Elsie M Taveras Journal: Environ Health Perspect Date: 2014-09-23 Impact factor: 9.031
Authors: Stephen J Mooney; Lin Song; Adam Drewnowski; James Buskiewicz; Sean D Mooney; Brian E Saelens; David E Arterburn Journal: Obesity (Silver Spring) Date: 2021-10-04 Impact factor: 5.002
Authors: Stephen J Mooney; Jennifer F Bobb; Philip M Hurvitz; Jane Anau; Mary Kay Theis; Adam Drewnowski; Anju Aggarwal; Shilpi Gupta; Dori E Rosenberg; Andrea J Cook; Xiao Shi; Paula Lozano; Anne Vernez Moudon; David Arterburn Journal: JMIR Res Protoc Date: 2020-05-19