Context: Pooling results of large replication projects yields a replication rate of about 50%. Low rates can be due to small sample sizes, low statistical power, and researcher degrees of freedom in analysis. Objectives: Push forward boundaries of scientific reproducibility and significantly expand number of replicated studies in economics, finance and related fields, with focus on non-experimental studies. Definition: Reproducibility is the ability to duplicate results using same materials and procedures. Replicability involves robustness replication (same data, different procedures) and direct replication (recoding using raw/intermediate data). Procedure for reproducing: Team assesses reproducibility, standardizes file structure and code, verifies data provenance, runs programs, completes report. Procedure for replicating: Studies are replicated by teams of replicators, who conduct robustness or direct replication and complete reports. Organize Replication Games where researchers team up to replicate studies. Project runs May 2023 to April 2027.
