Psychology uses 'registered replication reports' to improve reliability

The Guardian:

On Tuesday a group of scientists published a protocol, a set of instructions for running an experiment, which represents the culmination of a year’s discussion on the problems inherent in academic publishing. The protocol is the first in the Registered Replication Reports project. The scheme, led by Alex Holcombe, Bobbie Spellman and Daniel Simons, is far reaching, and requires putting aside egos and working together across universities and research groups to raise the standard of research.

I and others have written about this before, but the gist of the problem is that the academic publishing system is currently set up to encourage novel exciting findings, which is somewhat contrary to the way scientific research works. While novel findings are interesting, it’s cumulative converging evidence from a variety of different sources that provides us with scientific findings we can more readily say are evidence of true effects.

Read the whole story: The Guardian

See Barbara Spellman, at the