[R] OT: The topic of reproducibility in the media

Rolf Turner r.turner at auckland.ac.nz
Sat Oct 26 10:23:30 CEST 2013

Reproducibility is indeed important.  From the point of view of those in 
the statistics
community and in particular in the R community, the key issue is that 
the data on which
a publication is based should be readily accessible so that others can 
replicate and
possibly extend the analysis, and propose and present alternative analyses.

But to say that "most published scientific research is probably false" 
is drivelous
nonsense.  That sort of assertion is made by right-wing ideologues who 
are afflicted
with envy of the academic community and slang off at it to alleviate 
their own sense
of inadequacy.


     Rolf Turner

On 10/26/13 21:06, Liviu Andronic wrote:
> Dear all,
> I know that reproducibility is a big concern for the R community, so
> it may be interesting to some of the readers on this list that The
> Economist recently ran a series of articles denouncing the alarming
> number of shoddy and non-reproducible published papers:
> http://www.economist.com/news/leaders/21588069-scientific-research-has-changed-world-now-it-needs-change-itself-how-science-goes-wrong
> http://www.economist.com/news/briefing/21588057-scientists-think-science-self-correcting-alarming-degree-it-not-trouble
> They even went as far as stating that "most published scientific
> research is probably false":
> http://www.economist.com/blogs/graphicdetail/2013/10/daily-chart-2
> Anyways, food for thought for the weekend.

More information about the R-help mailing list