PRIISM researcher and SCSS doctoral student George Perrett, Dr. Jennifer Hill, Dr. Marc Scott, and collaborator Anugya Srivastava coauthored a paper in the The American Statistician focused on the advantages of user-friendly software implementations. They argue that powerful statistical tools often fail in practice because they are too difficult to use correctly. On the other hand, scaffolded software, such as thinkCausal can increase the usability of causal inference tools and reduce existing implementation gaps compared to traditional causal inference software. In a randomized study, participants were asked to estimate a causal effect in a dataset using one of three options for the approach. The findings were clear: participants who used thinkCausal were substantially more likely to obtain accurate estimates with less uncertainty.
