John Cochrane soltou um post bacana sobre replicação em economia. Vale a pena conferir.
Science demands transparency. Yet much research in economics and finance uses secret data. The journals publish results and conclusions, but the data and sometimes even the programs are not available for review or inspection. Replication, even just checking what the author(s) did given their data, is getting harder.
I have seen many examples of these problems, in papers published in top journals. Many facts that you think are facts are not facts. Yet as more and more papers use secret data, it’s getting harder and harder to know.
The solution is pretty obvious: to be considered peer-reviewed “scientific” research, authors should post their programs and data. If the world cannot see your lab methods, you have an anecdote, an undocumented claim, you don’t have research. An empirical paper without data and programs is like a theoretical paper without proofs.
Neale Ahmed El-Dash, do Polling Data (que já mencionamos aqui algumas vezes, como no modelo de impeachment), acabou de divulgar dados de pesquisas eleitorais brasileiras publicadas entre 1989 a 2015. Você pode acessar os dados clicando em “Acervo/Past Elections”.
Quem usa o RStudio sabe há quanto tempo aguardamos isso! Provavelmente era uma das sugestões mais pedidas. E agora está disponível no preview release.
Uma série de palestras interessantes do Sackler Big Data Colloquium:
Hal Varian: Causal Inference, Econometrics, and Big Data
Leo Bottou: Causal Reasoning and Learning Systems
David Madigan: Honest Inference From Observational Database Studies
Susan Athey: Estimating Heterogeneous Treatment Effects Using Machine Learning in Observational Studies