[R] Rating R Helpers

John Sorkin jsorkin at grecc.umaryland.edu
Mon Dec 3 04:51:24 CET 2007


I believe we need to know the following about packages:
(1) Does the package do what it purports to do, i.e. are the results valid?
(2) Have the results generated by the package been validate against some other statistical package, or hand-worked example?
(3) Are the methods used in the soundly based?
(4) Does the package documentation refer to referred papers or textbooks?
(5) In addition to the principle result, does the package return ancillary values that allow for proper interpretation of the main result, (e.g. lm gives estimates of the betas and their SEs, but also generates residuals)?.
(6) Is the package easy to use, i.e. do the parameters used when invoking the package chosen so as to allow the package to be flexible?
(7) Are the error messages produced by the package helpful?
(8) Does the package conform to standards of R coding and good programming principles in general?
(9) Does the package interact will with the larger R environment, e.g. does it have a plot method etc.?
(10) Is the package well documented internally, i.e. is the code easy to follow, are the comments in the code adequate?
(11) Is the package well documented externally, i.e. through man pages and perhaps other documentation (e.g. MASS and its associated textbook)?

In addition to package evaluation and reviews, we also need some plan for the future of R. Who will maintain, modify, and extend packages after the principle author, or authors, retire? Software is never "done". Errors need to be corrected, programs need to be modified to accommodate changes in software and hardware. I have reasonable certainty that commercial software (e.g. SAS) will be available in 10-years (and that PROC MIXED will still be a part of SAS). I am far less sanguine about any number of R packages.
John 



John Sorkin M.D., Ph.D.
Chief, Biostatistics and Informatics
University of Maryland School of Medicine Division of Gerontology
Baltimore VA Medical Center
10 North Greene Street
GRECC (BT/18/GR)
Baltimore, MD 21201-1524
(Phone) 410-605-7119
(Fax) 410-605-7913 (Please call phone number above prior to faxing)

>>> <Bill.Venables at csiro.au> 12/1/2007 2:21 AM >>>
This seems a little impractical to me.  People respond so much at random
and most only tackle questions with which they feel comfortable.  As
it's not a competition in any sense, it's going to be hard to rank
people in any effective way.  But suppose you succeed in doing so, then
what?

To me a much more urgent initiative is some kind of user online review
system for packages, even something as simple as that used by Amazon.com
has for customer review of books.

I think the need for this is rather urgent, in fact.  Most packages are
very good, but I regret to say some are pretty inefficient and others
downright dangerous.  You don't want to discourage people from
submitting their work to CRAN, but at the same time you do want some
mechanism that allows users to relate their experience with it, good or
bad.   


Bill Venables
CSIRO Laboratories
PO Box 120, Cleveland, 4163
AUSTRALIA
Office Phone (email preferred): +61 7 3826 7251
Fax (if absolutely necessary):  +61 7 3826 7304
Mobile:                         +61 4 8819 4402
Home Phone:                     +61 7 3286 7700
mailto:Bill.Venables at csiro.au 
http://www.cmis.csiro.au/bill.venables/ 

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] 
On Behalf Of Doran, Harold
Sent: Saturday, 1 December 2007 6:13 AM
To: R Help
Subject: [R] Rating R Helpers

Since R is open source and help may come from varied levels of
experience on R-Help, I wonder if it might be helpful to construct a
method that can be used to "rate" those who provide help on this list.

This is something that is done on other comp lists, like
http://www.experts-exchange.com/.

I think some of the reasons for this are pretty transparent, but I
suppose one reason is that one could decide to implement the advise of
those with "superior" or "expert" levels. In other words, you can trust
the advice of someone who is more experienced more than someone who is
not. Currently, there is no way to discern who on this list is really an
R expert and who is not. Of course, there is R core, but most people
don't actually know who these people are (at least I surmise that to be
true).

If this is potentially useful, maybe one way to begin the development of
such ratings is to allow the original poster to "rate" the level of help
from those who responded. Maybe something like a very simple
questionnaire on a likert-like scale that the original poster would
respond to upon receiving help which would lead to the accumulation of
points for the responders. Higher points would result in higher levels
of expertise (e.g., novice, ..., wizaRd).

Just a random thought. What do others think?

Harold




	[[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help 
PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html 
and provide commented, minimal, self-contained, reproducible code.

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help 
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html 
and provide commented, minimal, self-contained, reproducible code.

Confidentiality Statement:
This email message, including any attachments, is for th...{{dropped:6}}



More information about the R-help mailing list