[Esip-preserve] Article on research replication

[Tilmes, Curt (GSFC-6190)]

Fantastic paper illustrating the importance of sufficient provenance information for replication of research results:

Learning from mistakes in climate research
Rasmus E. Benestad , Dana Nuccitelli, Stephan Lewandowsky, Katharine Hayhoe,Hans Olav Hygen, Rob van Dorland, John Cook
http://dx.doi.org/10.1007/s00704-015-1597-5

Be sure to check out the supplementary material that includes software with data/links to data, and actual code that can be used to replicate various research results (and detailed descriptions of various methodology flaws).

Not just one use case, but a collection of many great real world use cases for much of our work.

Stuff like this clearly illustrate why we need better persistent, resolvable identifiers for data:

Note. the 10.7 cm flux used to be fetched from URL ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/SOLAR_RADIO/FLUX/DAILYPLT.OBS which no longer seems to work.

Great quotable commentary too:

By copying the numbers in published tables, and providing these together with the source code, we hopefully should be able to prove the inappropriateness in a way that even deniers find it hard to deny. This is how science should work - solid piece of work will mean that the results should be reproduced over and over again...

Curt

--
Curt Tilmes, Ph.D.
NASA Goddard Space Flight Center
+1 301-614-6218 (office)
+1 443-987-6228 (mobile)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.deltaforce.net/pipermail/esip-preserve/attachments/20150827/c5a166ab/attachment.html>