The entry requirements for pooh-poohing a finding with which one disagrees
have been raised.  Now one must not only read the paper, but also dive into
highly technical supporting materials.

While other journals have encouraged the release of "supplemental data", the
scofflaw rate has been high, and the level of detail low.  PLoS is suddenly
being far more aggressive:

http://blogs.plos.org/everyone/2014/02/24/plos-new-data-policy-public-access
-data/
or
http://tinyurl.com/qcf3z3v

"...authors must make all data publicly available, without restriction,
immediately upon publication of the article."  The data must be available
within the article itself, in the supplementary information, or within a
stable, public repository.

The Obama administration did something similar for all federal work:
http://www.whitehouse.gov/the-press-office/2013/05/09/executive-order-making
-open-and-machine-readable-new-default-government-
or
http://tinyurl.com/cv69z5e

It is true that detailed data and notes are often misplaced or poorly
managed, and that several surveys have shown that only a tiny number of
papers can be supported with confidence, due to a lack of sufficient
archived background data.  The basic problem is that, in a time of reduced
R&D funding, there is never funding for archiving the notes and "raw work
product" of science.  At Bell Labs, we had a battalion of library science
majors who did the best they could to maintain complete archives, but few
organizations focus such resources on preservation work.

Even with sufficient data and notes, replication is very difficult, perhaps
troublingly impossible:
http://www.economist.com/news/briefing/21588057-scientists-think-science-sel
f-correcting-alarming-degree-it-not-trouble
or
http://tinyurl.com/ovwctme

But the downside is that demanding full accessibility on "all data" will
make many researchers reticent to publish in a journal that demands strict
full disclosure, because it means giving away the data they were planning on
using in subsequent papers.  Given how academia rewards researchers, the
term "Least Publishable Unit" has been coined.  There is an incentive to
publish as many papers as possible per project.  

There's also the problem of physics, where a single experiment can easily
produce over a petabyte of data.  Does anyone want to really slog through
millions of voltage traces at nanosecond intervals?  Genetics likely has a
similar problem in terms of the bulk of "raw data" generated.  How does one
"provide" so much data to the journal, and how would anyone be able to
download it all to be able to sift through it?  More importantly, why would
anyone?

But any move towards archiving more information has got to have a positive
outcome in general.  



  

 

             ***********************************************
The BEE-L mailing list is powered by L-Soft's renowned
LISTSERV(R) list management software.  For more information, go to:
http://www.lsoft.com/LISTSERV-powered.html