Fixing statistics is more than a technical issue

  • Andrea Saltelli, Philip Stark
  • Nature, January 2018, Springer Science + Business Media
  • DOI: 10.1038/d41586-018-00647-9

Fixing statistics is more than a technical issue

What is it about?

The title "Fixing statistics is more than a technical issue" means that the execution of statistical work requires a combination of statistical skills and of ethical virtue. Escaping the 'publish or perish' imperative requires a dose of abnegation in the execution of statistical work which runs counter the present system of incentives. There is a debate on whether better methods or better incentives are needed to fix the reproducibility crisis. In our opinion the existing methods are not the problem - though the way they are thought and applied is. Foremost we see a major difficulty to escape a culture which promotes publication at all cost. Being the coal face of science, statistics care can make the difference between publishing and not publishing, and this in turn can make a difference between having and not having a carrier.

Why is it important?

The crisis of science calls in question Enlightenment ideals as well as huge amount of resources. This is perhaps the most important crisis of modernity and links dangerously to the post-truth debate.


Professor Andrea Saltelli
Universitetet i Bergen

The crisis is unfolding at such a speed that academic publishing on the subject can hardly keep pace. A debate on these issues is urgent but made difficult by a persistent scientism, meaning by this the conception of science as a something which needs to be defended from the obscure forces of superstition, ignorance and religion. This vision of science it itself a religion and is systematically mobilized to fight science's enemies, e.g. president Donald Trump and his policies. A science 'at war' does not tolerate doubt and delays a solution of the present crisis.

Read Publication

The following have contributed to this page: Professor Andrea Saltelli