Implementing novel, flexible, and powerful survey designs in R Shiny.

Abstract

Survey research is ubiquitous in the social sciences as a cost-effective and time-efficient means of collecting data. However, the available software for implementing and disseminating such surveys lacks flexibility, stifling researcher creativity and severely limiting the scope of questions that survey research can address. In this paper I introduce the use of R Shiny, an open source web application and scripting language, for implementing powerful, innovative, and fully customizable surveys. Through six applications rooted in important questions in political science, I show that R Shiny allows for (1) randomized question selection, (2) programmatic treatments, (3) programmatic survey flow, (4) adaptive question batteries, (5) sequentially block-randomized design, and (6) randomized intracoder reliability tests, expanding the scope, ease, and cost effectiveness of online survey research. I make all replication code available online.