Thanks for the compliment!
This is a large and complicated issue. Some of us could swap personal
impressions and experiences at various levels, but there is an immediate
consequence for all who care about proper recognition of software
writing. Cite software articles and programs that were important for
your project in whatever form you can.
Nick
[email protected]
Lachenbruch, Peter
Certainly SJ articles would count as they are refereed. My concern was
for people like Kit and Nick and Roy who contribute much more than SJ
articles and don't get the academic rewards due them.
And I think Roy was particularly interested in mapping contributions of
user-written programs to academic brownie points. If they are published
in SJ or somewhere else, the issue is handled.
In many of my articles, I include Stata code used for simulations, but
the bulk is the results not the code.
Christopher Baum
Tony says
Roy brings up an interesting point. How are contributions to SSC or
user-written programs in general evaluated in academe? In my
experience, these would need to be refereed by one's peers to count
academically. On the other hand there are many extremely valuable
contributions from Roy and others that are easily worth many articles
that I see in the journals I read. Perhaps it is time to consider this
in various academic contexts. I think outreg2, fracpoly, mim, etc. are
easily worth publication points, but I'll bet they haven't counted for
much unless the author or department chair has been able to convince
some higher powers (e.g. 2^n) of their worth. Please let me be wrong
on this.
It should be possible to get 'publication points' by describing one's
work in a peer-reviewed article in the Stata Journal, all the more so
that it is now listed in both Sci-Math and SSCI indices of Thomson. Of
course, such an article is not just an expanded help file. I once
wrote a seasonal unit root routine with Richard Sperling. Decent code,
help file, examples, and set it aside. More recently Domenico Depalo
wrote a very nice, academically respectable article about the theory
of seasonal unit root testing, described the program he had written to
perform those tests (which does what our hegy program does and more),
and published it in SJ 9:3
(http://ideas.repec.org/a/tsj/stataj/v9y2009i3p422-438.html
). It seems to me that such an article should receive as many
'publication points' in most universities' evaluation processes as
many papers that merely take existing software and exercise it to test
a model -- particularly with SJ now in SSCI, which is a checklist item
for many European institutions.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
