[R] Protocol for answering basic questions

Baskin, Robert RBaskin at ahrq.gov
Wed Dec 1 20:48:20 CET 2004

> number.of.years.using.R * runif(1)
[1] 1.064863
I waited this many hours before responding:)

First, let me say thank you very much to the R team for ...the software
...the help-list ...other intangibles.  I am a relatively new R user and I
am struggling in my own way to learn R.  I follow the list regularly for
several reasons.  No matter what kind of answers I see, I think the benefits
of R far outweigh any irritation with any responder - this environment is an
incredible gift.  Bytheway - responses to my questions on this list have
been completely helpful and it has been clear that the responders had taken
time to respond thoughtfully and fully.

Second, I know that the list(s) maintainers have a difficult job and really
deserve a vote of thanks, and that the developers are really spending a huge
amount of time already, but is there some possibility that an "environment"
could be created so that a person such as myself could ask a discussion
question without generating unnecessary email traffic for the people who
should be spending their time working on developing the software?  Splitting
off a beginner list has been suggested before but I don't think this is the
answer - of course I don't have a good answer.  Also, not having 'expert
review' of answers is totally unacceptable.  But there is getting to be a
problem with bandwidth here (which I just incremented by 1).

Third, it appears to me - and I do say here that I am not necessarily
qualified to judge - that there might be some inconsistency in the way
responses are handled.  I know this is not well phrased but there has been
somewhat of an example the last few days that I am thinking about.  There
have been several questions asking how to construct variables named v1-v10
(for example).  The very useful (I don't think rude but what do I know)
replies have typically been along the lines of 'this is FAQ whatever *but
this isn't a good way to do things in R*'.  The first time I saw this reply
it was extremely helpful to ME because of the *don't do it this way in R*
part - no matter how rude it might have been perceived by the person who
wrote the question.  The *don't do it this way in R* part made me think
about some code I was working on.  Sandwiched in the middle of these
excruciatingly similar questions was a question about 'how do I write a for
loop to take a crosstab over v1-v10'.  There was a perfectly good response
with a for loop but no response about options to avoid a for loop - maybe it
wasn't appropriate in the situation but what do I know - so I wrote some
fake data and tried xtabs(cbind(v1,v2) ~ predictor) and it worked:)  Perhaps
I should have sent a question to the list about this - a year ago I would
not have done so because of thin-skinnedness - but I didn't do it now
because I am getting seriously concerned about bandwidth.

**I would like to respectfully say that it is important to ME and to other
beginners that the gurus provide the 'don't do it this way in R' pointers
(please beat the SAS out of me:)

After a lot of rambling, my main concerns are:
* learning good R programming techniques
* access to the R environment (including r-help)
* bandwidth on this list
I regret that I cannot be as articulate as the responses from the R team
members that I have seen on this list.

Thanks for R

-----Original Message-----
From: Robert Brown FM CEFAS [mailto:r.g.brown at cefas.co.uk] 
Sent: Wednesday, December 01, 2004 11:46 AM
To: r-help at stat.math.ethz.ch
Subject: [R] Protocol for answering basic questions

I have been following the discussions on 'Reasons not to answer very basic
questions in a straightforward way' with interest as someone who is also new
to R and has had similar experiences.  As such it with sadness that I note
that most seem to agree with the present approach to the responses to basic
questions.  I must thank those respondants to my own questions who have been
helpful, but there are some whose replies are in my opinion not only
unhelpful but actually rude.  Indeed I've now started using Splus instead of
R so as to have access to a 'proper' support service.  Indeed, the main
thing I've learned from R is a new respect for the values of commercial
software and a scepticism regarding free software. In the end my experience
of r help is that you get what you pay for. Many of the so called socratic
responses (in this list and the wider academic community) can be seen as
simply way to avoid additional work of a complete reply. 

Experienced R users don't seem to understand how difficult the program can
be to new users. Responding that the questioner should read the
'Introduction to R' or a similar document is like  answering a question for
directions to one's house with 'Buy a map'.  Most likely both such
questioners have already tried that and are asking because that approach
failed.  R is a language and like all languages it is simple to those that
understand it and complex to those who do not. Every schoolboy in Spain
speaks Spanish, but I know from experience that for most English people it
is very difficult to learn Spanish and take years of study.  If I'm asked a
question from a novice of a language (be it Spanish or R) do I reply
'consult the dictionary'. I would hope not!  I can tell repondants that
whilst many of my basic questions may seem simple it is not for lack of
studying the very sources they refer to.  If only learning was so simple.  I
suspect that the same is true of most question!  ers.

I speak as someone with a PhD and many years as a researcher in my
speciality as well as someone close to completing a masters in statistics
with distinction. As such I am not a total novice and would suggest that if
I'm having problems so are many; and it is not a result of lack of study as
so many responses seem to suggest.  Indeed it is revealing that several
responses suggest that they want to discourage questions so they don't
overwhelm r-help.  Understandable but not a recipe to encourage the use of R
by other than experts. The R community needs to decide of they really only
want expert statisticians users and make this clear if it is the case.
Alternatively if they are to encourage novices the present approach is not
the way to do it.

I can appreciate that many of the respondants are busy, but if that is the
case it would be better if they didn't reply at all. I was taught many years
ago that if you can't say anything nice/useful then to say nothing at all.
Something similar could well be applied to this list.  I must say that some
respondants are very helpful; and I thank them.  Leave these simple
questions to such people.  Indeed it seems surprising that some exteremely
experienced R users choose to reply to these basic messages at all; and it
seem it is mostly these people who are rude.  I would have thought it might
be better for them to concentrate on complex problems more suited to their
skills and interests and leave the simple questions to more sympathetic

Perhaps there is a case for two r help lists catering to basic and advanced
questions? Certainly if the R community is serious about appealling to users
outside advanced statisticians there is a need for a change of approach in r
help and elsewhere.  Russ Ackoff identified much of the failure of
management science as due to those who were 'mathematically sophisticated
but conceptually naive' and much the same could be said for many in the R

Finally, let me once again thank those who have responded helpful to my
queries in the past and ask them to continue in that vein; their assistance
and effort is greatly appreciated.


This email and any attachments are intended for the named re...{{dropped}}

More information about the R-help mailing list