Hey Chato
nice to hear from you, your suggestion is very timely and I hope you don't mind if I
forward it to RCom-l.
(everybody: Chato is a researcher based in Barcelona who recently completed a study of
gender differences in Wikipedia, with Mayo as a coauthor)
Chato: we discussed extensively a similar proposal during the last RCom meeting [1] (which
finished just minutes ago). The majority of RCom seemed to support the idea of a platform
on which individual editors could decide to participate in research in general, in what
study or type of research in particular, and with what frequency and to revoke their
permission to be contacted for these types of research at any point. This solution would
allow us to avoid the problem of gauging community consensus on every single subject
recruitment request that we get as well as the problem of finding an appropriate
recruitment method to suit everyone. We also discussed what role RCom could have in
reviewing and flagging recruitment requests before they get posted to this platform. The
notes of the meeting are here [2]
Melanie, Aaron and myself volunteered to start a proposal on Meta, I'd be great if we
could get your input once we have a first draft.
In the meantime, enjoy your holidays!
Best,
Dario
[1]
On Dec 22, 2011, at 1:16 PM, ChaTo (Carlos Alberto Alejandro CASTILLO Ocaranza) wrote:
Hi Dario,
Sorry if this is way too late, I was traveling these last days, now I am in Chile ;-) But
anyways, this is what I promise to try to write up after our meeting last week:
============================
Proposal: Wikipedia editor panel
Several research topics require some sort of survey/interview to be applied to a sample
of Wikipedia editors. Currently, this is done in most cases by directly contacting editors
via their user talk pages, which is considered a bad practice by the Wikipedia Research
Committee WRC.
It is proposed that the WRC maintains a large editor panel that can be partially assigned
to different research groups.
Editors would be invited to be part of this panel by a number of channels to be defined,
including the semi-annual survey. Editors would indicate the maximum number of different
surveys they would like to participate in per year (e.g. 1-4, 5-10, 11-50, 50+), and
fill-in a demographic form including age, gender, etc.
Researchers would apply to conduct surveys to subsets of this panel via the WRC,
indicating: the target number of editors requested, and some constraints (based on a
schema of the properties available for editors, provided by WRC).
The WRC would review the request, and on approval, and forward a URL provided by the
researchers to a sub-set of the panel matching the constraints requested by editors. (This
matching should balance load, there are a number of algorithms for this including
http://research.yahoo.com/pub/3312). After this, the survey would be handled directly by
the researchers, who would send a post-survey report to the WRC indicating e.g. response
rate received.
Why the alternatives are bad?
- Handling each research request on a case-by-case basis, aside from requiring more
effort by the WRC, would generate a number of different messages to editors, which can
create confusion among them.
- Allowing researchers to add questions to the semi-annual survey has a number of
problems: it may blow-up the time required to answer the survey, it may affect the
responses received given that users already have answered a long questionnaire, etc.
============================
Thank you,
--
ChaTo (Carlos Castillo) Let's connect! · LinkedIn · Facebook · Twitter @ChaToX