Nov 24, 2015 | By

Scientific Reproducibility – Self-Reflection is a Good Thing

Brian Nosek is a professor of psychology and Executive Director of the Open Science Project dedicated to improving the transparency and credibility of science. Earlier this year he and his team reported the results of a project seeking to reproduce 100 psychology studies published in top journals. The results, reported in Science, Nature and this Econtalk podcast, indicated that less than half the findings could be reproduced. According to Nosek, the point is not to critique individual papers but to gauge just how much bias drives publication in psychology. He believes that other scientific fields have the same problem – one analysis found that only 6 of 53 high-profile papers in cancer biology could be reproduced – and new reproducibility efforts are getting underway. Hopefully, the critical self-reflection represented by this effort will spark debate on science research and publication, and greater humility among scientists. For additional posts dealing with the problems of bias, see:

One Response to “Scientific Reproducibility – Self-Reflection is a Good Thing”

  1. George Gantz says:

    Another major initiative in the effort to increase attention to the reproducibility of research, this time in medicine, was highlighted in the February 6th issue of The Economist in “Let;s Just try that again”‘ (page 74). In 2012, Amgen published a paper in Nature reporting on its attempts to reproduce results in 53 high-profile cancer research papers – only 6 passed the test. Subsequently, a publishing house in London, Faculty of 1000, launched the Preclinical and Robustness Channel to provide an outlet for replication studies that have never seen the light of day. Progress is being made!

Join the Discussion

Why ask?