[R] Protocol for answering basic questions

[John Maindonald]

Given the huge range of demands and interests of the people who
are subscribed, the list may well do about as much as can
reasonably be expected to address the needs of those starting out
with R.  A fair level of discipline is necessary, else the list will
become unmanageable.  Already some of my colleagues do not
subscribe because of the sheer volume.  There is a fair level of
tolerance for basic questions that are clearly and succinctly posed.
Submissions of the type "I am a total beginner - how do I start"
really do require rather prompt redirection to the FAQ and to the
official and contributed documentation.

The posting guide is already rather full.  But maybe a link could
be added to a page of the ilk: "I am new to R, where do I start?"
John Fox has I think occasionally posted responses that could
be used as a basis for such a page.  It should probably be in two
parts:
a) I am new to R and have limited statistical knowledge.
b) I am new to R, but with (I think) some reasonable level of
statistical knowledge.
If this is thought a useful idea  and there is not already such a
document, I am happy to help with it.

The discussion has ranged over a large number of issues, which
I think need to be separated:
1) There is a "getting started in R issue".
2) There are issues that relate to gaining the statistical knowledge
that will allow effective use of R.
3) There are issues of good statistical practice -- how can the
standard of use of statistical methods in application areas be
improved?
4) There are quirky points that cannot easily (or at all) be gleaned
from the documentation, and where a hint from others will be a
huge help both the questioner and almost certainly to others on
the list.  ("I am glad that you asked that.")
5) There are issues that are slowly being addressed, as part of
the ongoing development of R -- in the improvement of
documentation and in mechanisms that may make R easier and
slicker for everyone, novices and experts, to use.

Items 4 and 5 are well handled by this list, or by r-devel.  It can
and should make some contribution to 3.  There is a nether-nether
land between 1-2 and 4-5 where it can be useful.  Apart from
these contributions at the margin, 1-3 are really statistical and
R training issues, that are not well handled by this list, and
probably not by any list.

The most useful response to issues 1-2 (and, often 3) is to
direct inquirers to suitable training resources.  What is "suitable"
will however depend on personal circumstances and
geographical location.  So what do we say?  That it is good that
that this question has been asked, and that good answers are
sure to emerge slowly over the course of time?

John Maindonald.

On 3 Dec 2004, at 11:19 PM, r-help-request at stat.math.ethz.ch wrote:

> From: "Robert Brown FM CEFAS" <r.g.brown at cefas.co.uk>
> Date: 3 December 2004 7:07:51 PM
> To: <r-help at stat.math.ethz.ch>
> Subject: RE: [R] Protocol for answering basic questions
>
> . . . .
>
> I joined the community because I want technical assistance and I don't 
> question the technical skills of many of the respondants, but the 
> wider appreciation of needs of users i.e conceptual naivety.  It's 
> clear from this string that many beginners are leaving the forum and 
> this should be a cause for concern if we are seriously concerned with 
> propogating knowledge.  Unfortuantely this string does seem to 
> indicate this is not a major concern and so be it.
>
John Maindonald             email: john.maindonald at anu.edu.au
phone : +61 2 (6125)3473    fax  : +61 2(6125)5549
Centre for Bioinformation Science, Room 1194,
John Dedman Mathematical Sciences Building (Building 27)
Australian National University, Canberra ACT 0200.