[R] Re: Protocol for answering basic questions

Tony Plate tplate at acm.org
Sat Dec 4 19:07:35 CET 2004


Perhaps something like the following paragraph should be added to the start 
of the "Posting Guide" (as a new paragraph right after the existing first 
paragraph):

Note that R-help is *not* intended for questions that are easily answered 
by consulting one of the FAQs or other introductory material (see "Do your 
homework before posting" below).    Such questions are actively discouraged 
and are likely to evoke a brusque response.  Questions about seemingly 
simple matters that are mentioned in the FAQs or other introductory 
material *are welcomed* on R-help *when the questioner obviously has done 
their homework and the question is accompanied by an explanation* like "FAQ 
7.2.1 seems to be relevant to this but I couldn't understand/apply the 
answer because ...".

Something like this would make it very clear up front what type of 
questions are not appropriate.  (I'm not trying at all to dictate the 
policy, but as far as I can tell, the above summaries the attitude of the 
majority of very knowledgeable helpers that respond to questions on R-help.)

Also, I think that John Maindonald's idea of a "I am new to R, where do I 
start?" page, with a link from the posting guide, is an excellent idea.

I'm aware that some feel that the posting guide is already too long, but my 
feeling is that if users don't read a very easily accessible posting guide 
AND post inappropriate questions AND become offended by brusque responses, 
then they are beyond where they can easily be helped.  The most important 
thing is to make it very clear what types of questions are and are not 
considered appropriate, so that beginning users know what they are getting 
into.

And the following might merit inclusion in the FAQ:

Why is R-help not for hand-holding beginner questions?
R-help is a high traffic list and the general sentiment is that too many 
very simple questions will overwhelm everyone and most importantly result 
in the knowledgeable helpers ceasing to participate.  The reason that there 
is no "R-help-me-quickly-I-dont-want-to-read-the-documentation" list is 
that no-one has felt that it would work well -- it is unlikely that many 
knowledgeable users of R would be willing to participate.  Without such 
users participating, it is likely that sometimes bad advice would be 
offered and stand uncorrected, because R is a complex language with many 
ways of doing things, some markedly inferior to others.  For these reasons, 
some feel it would be a very bad idea to create such a list.  (However, 
anyone who believes otherwise and wishes to start and maintain such a list 
or other similar service is free to do so.)  One reason for this overall 
state of affairs is that R is free software and consequently there is no 
revenue stream to support a hand-holding support service with paid 
employees.  So although the actual software is free, some investment in 
terms of time spent reading documentation is required in order to use 
it.  Furthermore, many of the frequent helpers on R-help have written 
introductory documents intended to help beginners with many aspects of 
learning and using R (e.g., "An Introduction to R", and the various 
FAQs).  Consequently they sometimes get fed up getting asked again and 
again the same question they have already written a document to 
explain.  Nonetheless, the general sentiment on R-help is very helpful -- a 
quote summarizes it well: "It's OK if you need some spoonfeeding (I need 
that quite often myself), but at least show how you have tried to use the 
spoon yourself, instead of just showing us your open mouth."  [Attribution 
to Andy Liaw, or remain anonymous?]


As some feel that sufficient time and bandwidth has already been spent on 
this issue, if anyone has any comments on this particular matter of an 
addition to the posting guide (or FAQ), feel free to choose to respond to 
me privately, and I will summarize as appropriate.

-- Tony Plate




More information about the R-help mailing list