Nov 8, 2015  |  London

Research Reproducibility: Hands-on Experience with Replication

Through this session at MozFest 2015, we provide an opportunity to reproduce the results of a prominent academic publication. The goal of our workshop would be for attendees to be able to find a suitable publication, locate and download the original data from the Internet, reproduce the original results, and extend the analysis to probe the robustness of the published findings. This would require that the participants have some statistical knowledge, as well as programming ability (R preferably, since it is available free, but also possibly Stata, SPSS, SAS, etc. if students have that software themselves). However, by using a relatively simple randomized trial (such as “Are Emily and Greg More Employable than Lakisha and Jamal?” by Marianne Bertrand and Sendhil Mullainathan, 2004 just to name one possible example) on an important and easily understood topic (racial discrimination in employment), and by having attendees work in small groups, even beginners should be able to benefit.