[RCom-l] Scope of the Research Committee

[Yaroslav M. Blanter]

Dear All,

this is just to start the discussion. Below I copy the ideas some of us
shared at http://eiximenis.wikimedia.org/ResearchCommitteeBios , and also
the original functions compiled by Erik.

Functions:

The committee's core function includes:

    * developing policy around researcher permissions for non-public data
    * supporting the development of subject recruitment processes
    * reviewing research projects when conflicts-of-interest arise
    * articulating and channeling requests for data and technical
resources
    * helping to formulate the key strategic research objectives of the
Wikimedia movement (see strategy.wikimedia.org)
    * helping to formulate small tactical experiments related to
Wikimedia's strategic goals
    * developing an open access policy as a requirement for significant
support from the Wikimedia Foundation
    * helping create a "starter kit" for researchers to avoid duplication
of effort

Opinions on the scope:

What possible subcommittees/workgroups to form?

    * Strategy

    * Data

    * Ethical review

    * Subject recruitment

    * Code

    * Tests

    * Funding

    * Publishing

    * Interface Research


[John Riedl]: My bias is that we should focus on creating workgroups in
areas that are exploring research that is particularly hard to carry out in
the current Wikipedia.  One way to identify these areas is to look for
types of Wikipedia research that seems like it would be very valuable but
that is seldom done.  Two examples of this type of research are: controlled
experiments with hundreds of users, and experiments in new Wikipedia
interfaces.

[Daniel Mietchen] I wonder how many subcommittees make sense given the
size of the committee. Prioritizing is necessary, though, and I think
substructures will naturally flow from there.

{Yaroslav Blanter] I guess we should understand 1) how meny members we
are; 2) define the scope. Then we can easily decide on the number of
committees and their scopes

[Milos Rancic] From the Wikimedia perspective, I would add that we need at
least three general tasks, which could be handled by listed committees or
by new one(s): (1) How to keep community healthy? (2) How to improve
existing projects? and (3) Which new projects are needed in the corpus of
free knowledge?

denny: Milos, you talking about Wiki(m/p)edia in general, or about
research projects in particular? I also agree, we can only do so much with
the size of the committee. Also, our task is obviously not to do the
research ourselve, but provide the interface between Wikimedia and the
wider research community.

[Milos Rancic] Denny, I am talking about research goals. It looks too
general, but as a Wikimedia bureaucrat, I would like to get scientific
answers to as many as possible questions related to those three general
questions. Moved at the level of particular researches, research which
would state that "edit" button should stay at some position because it
would improve participation -- would be a perfect example of research which
aim to cover the part of the second question (How to improve existing
projects?).

[Yaroslav Blanter] Sorry, I was a bit out of business for the weekend. I
in general agree with Milosh and with denny at the same time - we need to
understand what we can actually do with 8 committee members. Should we
first discuss what RESEARCH means? I guess there are several activities in
and around WMF projects, which are covered by this term and are in some or
other way in the scope of our committee, in no particular order: (1)
original research performed by WMF projects (there are only two where OR is
not prohibited - Wikibooks and Wikiversity), shoud there be some
guidelines? (2) Scientific research on WMF projects (probably mainly
Wikipedia) performed by either WMF-related or independent researchers; (3)
Non-WMF-related research performed by research institutions: Here our
business is only interaction between WMF projects and researchers at these
institutions. Do we want to handle all three? Is there is smth else which I
overlooked?

P.S. We started a serious discussion and I am not sure this is the best
interface. I believe Erik promised to open a mailing list, may be all this
can be moved into the list. 

[denny] I agree with Yaroslav on maybe waiting for the list. These
wiki-like interfaces are known to be not the best for discussions like
these <reference needed>. ;)

[WereSpielChequers] I'm with Milos on the need for research to inform
decision making on existing projects, there have been a number of decisions
where it would have been really helpful to have a researcher look at the
stats and prepare a report to inform the discussion. Also I'm with Daniel
on thinking that it is a bit early to be forming subcommittes. I'm hoping
tha part of our role will be in advising researchers where there are
alternatives that are acceptable to both researchers and the community and
part will be explaining to other researchers why some things are not
acceptable to the community. For example in the last year we have had one
research project where someone was vandalising articles in order to see how
effective our anti-vandalism efforts were, and another was adding fake spam
links to see how many people would click through to a spam site; The first
could either have been done as an audit of existing articles to see how
many were vandalised, or by analysing vandalsim reversions to see how long
it took to spot the vandalism.  I'm not convinced that there is an
alternative way to do the spam test, but as Wikipedia is unlikely to be
taking advertising in the foreseeable future and as the willingness to
click on a link is at least partially dependent on the trust one places in
the website where you see that link, then I don't see the value in testing
how well spam works in Wikipedia.


Cheers
Yaroslav