I stumbled upon this writeup exploring the inability of economists to replicate the results of a famous Reinhart-Rogoff study (note, I did not know about this study before, I am no economist, and hence have no professional opinion or stakes in this). The problem seems to be in the insufficient access to the dat used by peers, and the inability to check the entire code by anyone interested. The methodology documented was not enough to reproduce the results.

Let’s hope that eResearch approaches such as our AURIN could help to at least partially improve the status quo in the social sciences…

Advertisements