|Friday, February 22|
|PS2 Poster Session 2 (with refreshments)||
Fri, Feb 22, 4:45 PM - 6:15 PM
Facilitating Survey Design Decisions by Incorporating an R PackageView Presentation *Lauren Brooke Hund, Harvard School of Public Health
Marcello Pagano, Harvard School of Public Health
Keywords: survey design, communication, R package
Constructing user-friendly survey designs for resource-scarce settings requires thinking critically about how to compromise between accuracy and feasibility. We discuss our efforts at collaborating with WHO officials to revise the lot quality assurance sampling survey design for national neonatal tetanus elimination surveys. To aid in the discussion about selecting key survey design parameters, we engaged collaborators by providing a simple R package with basic plotting and summary functions that illustrated concepts and deepened understanding of the design decisions. In the context of neonatal tetanus elimination, we illustrate functions for sample size calculations and graphs showing risks of misclassification of elimination status for various user-specified design parameters. While our collaborators did not have statistical backgrounds, with some guidance, no one had trouble using R or the package. Effectively communicating with collaborators spread around the world is pivotal to ensure that the survey answers all relevant questions about neonatal tetanus elimination, and, by providing a simple package, we ensured that the concept of balancing accuracy and cost of the survey were clear.