[R] # of users of R

[j.logsdon@lancaster.ac.uk]

OK - my 0.01UKP's worth ...

As statisticians we should above all be able to make a good estimate (with
confidence intervals) of the R use.  I see three methods for sizing:

1 Taking a sample from the list, which will itself be biased (probably a
self-selected and therefore biased further) and asking these people to
estimate/count how many users there are,

2 In a future version, storing automatically in /usr/lib/R or wherever a
list of the users the first time they use R (ie when .R is set up).  This
number of unique entries on this list can then be requested via an email
to the installers email address (requested at download time).

3 A snowball sample, starting with the present 911 list members, where
people indicate (a) their applications area, (b) platform, (c...y) other
information and (z) nominate other users by email address, returning this
to a dedicated list address. The information is processed automatically,
checked against already known addresses and any new addresses emailed with
the questionnaire.

Method (1) seems to be under discussion at the moment, method (2) would
only asymptote as people downloaded the new versions but is simplest and
method (3) could provide useful further information re expert users etc
since it would be from the actual users rather than possibly the sysadmin
person.

This is not only of academic interest but could be used if necessary when
looking for commercial sponsors, equipment, grants etc.  The email
addresses should of course be kept private - we wouldn't want them
escaping into Outlook Express or a commercial list.

John

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._