Macartan Humphreys receives Statistical Software Award

July 01, 2019

The Statistical Software Award Committee of the Society for Political Methodology is pleased to announce DeclareDesign by Graeme Blair (UCLA), Jasper Cooper (Princeton; Columbia Ph.D. 2019), Alexander Coppock (Yale; Columbia Ph.D. 2018), Macartan Humphreys (Columbia and WZB Berlin), Clara Bicalho (WZB Berlin), Neal Fultz (Comake), Lily Medina (WZB Berlin), Aaron Rudkin (UCLA), and Luke Sonnet (UCLA) as the winner of the 2019 Society for Political Methodology Statistical Software Award.

The DeclareDesign suite helps people think through a research design prior to collecting data, with a suite of tools for fast analysis of data that previously was inconvenient in the R language. This coupling of explicit research design facilitation with analytical capability is relatively uncommon in the statistical software community. DeclareDesign is especially innovative because it rethinks research design as an integral component of statistical computing, and a process that can be addressed analytically via design structures and simulated data. Incorporating these procedures into a software package not only serves the function of structuring and streamlining the research design within a given project, it also aids in the goal of increasing transparency and reproducibility in social science research. This software, in addition, can make the process of rigorous social science research design more accessible and approachable to practitioners outside of academia. Finally, the DeclareDesign suite has the potential to facilitate discussion across divides in the discipline, with features allowing for both quantitative and qualitative research designs.

Documentation, vignettes, and a help/message board for the DeclareDesign suite are available on the project website<> and the code for the software is hosted on Github . Additionally, the logic of the approach is spelled out more fully in the forthcoming American Political Science Review article, “Declaring and Diagnosing Research Designs” by Blair, Cooper, Coppock, and Humphreys.