Gerard, I think that the work on Commons and WikiData is freaking awesome.
If I could clone myself I'd be digging into it immediately. Right now,
I'm working on measurement Wikipedias and large cross-wiki analyses. FWIW,
I think that the wikidata games are some of the most exciting things to
happen in Wikimedia wikis in a long time.
Rui, re. the survival graphs. Those are proportions. Multiply by 100 to
get percentages. i.e. the line starts at about ~24% and declines to ~7%.
I'd really like to revisit this work since we've standardized some of the
measures I was using and the new, standard definitions will result in some
differences. See
for the
updated definition. I'll try to schedule some time to get an updated
figure for ptwiki that goes back before 2006.
-Aaron
On Fri, May 30, 2014 at 5:30 AM, Rui Correia <correia.rui(a)gmail.com> wrote:
Hi Aaron
This is really a treasure trove of information. I am looking forward to
savouring it in detail. Many thanks.
One question for now on Point 5: the 3rd graph with values <1 - are those
percentages? Is the decimal notation correct?
Regards,
Rui
2014-05-30 1:52 GMT+02:00 Aaron Halfaker <ahalfaker(a)wikimedia.org>rg>:
Hi Rui,
You raised a lot of questions that I think I might be able to help
address.
I'm a research scientist working for the
WMF. My research focuses on
the
nature of newcomer participation, editor
motivation and value production
in
Wikipedia. See [1] and [2] (if you have the
time) for my most seminal
work
on the subject.
As you'll see in the study I referenced, my work directly addresses a
substantial portion of the questions you've raised. See also my team's
work with standardizing metrics[3] including survival measures[4] and my
work exploring retention trends in ptwiki[5]. See [6] for an example of
a
recent, cross-language study of newcomer article
creation patterns.
Also,
you might be interested in [7] since it confirms
your general concerns
about the speed of speedy deletions.
A lot of the work of /really understanding Wikipedia/ is only half-way
done
since it takes a long time build understanding
about previously
undocumented phenomena. The academic community, other researchers at the
WMF and myself are in the middle of developing a whole field around how
open collaboration systems like Wikipedia work, common problems they have
and how they can be best supported.
While we're developing this general knowledge about engagement,
production
and retention in our communities, we (the
research & data team) are also
working directly with product teams at the WMF to measure their impact on
key metrics (e.g. participation) with scientific rigor and to
challenge/develop/refine theory on which product strategies lead us
toward
our goals and which ones do not. See [8] and [9]
for examples of such
studies.
I welcome anyone who'd like to continue the conversation about what we do
and don't know about Wikipedia(s) to raise discussions at
wiki-research-l[10]. There are a lot more researchers on that list than
wikimedia-l. FWIW, I tend to follow that list more closely.
1. Summary:
http://www-users.cs.umn.edu/~halfak/publications/The_Rise_and_Decline/
2. Full paper:
http://www-users.cs.umn.edu/~halfak/publications/The_Rise_and_Decline/halfa…
3.
https://www.mediawiki.org/wiki/Analytics/Editor_Engagement_Vital_Signs
https://meta.wikimedia.org/wiki/Research:Ideas/Is_ptwiki_declining_like_enw…
https://meta.wikimedia.org/wiki/Research:The_Speed_of_Speedy_Deletions
8.
https://meta.wikimedia.org/wiki/Research:Onboarding_new_Wikipedians/Rollout
9.
https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_effect_on_newly_r…
10.
https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
-Aaron
>
> > From: Rui Correia <correia.rui(a)gmail.com>
> > Subject: Re: [Wikimedia-l] The first three weeks.
> > Date: May 29, 2014 at 5:07:45 AM PDT
> > To: Wikimedia Mailing List <wikimedia-l(a)lists.wikimedia.org>
> > Reply-To: Wikimedia Mailing List <wikimedia-l(a)lists.wikimedia.org>
> >
> > Hi James
> >
> > Do we have any figures on retention of new editors? How long does the
> > average new editor stay? What percentage of new editors stays on for
6
>
months; one year; two years? Do we have these figures for all
languages?
>
> New editors should be allowed space to grow. Wikipedia is so rich in
> developing all kinds of scripts, templates etc, that it would be easy
to
> create something to inform others that
someone is a new editor. Pages
by
> new editors should be left alone for a day
or two. There is nothing
more
> > disheartening than getting all excited about contributing only to
find
> that
> > someone comes along and either deletes your first attempt or
nominates
it
> > for deletion. I've have seen this happen WITHIN MINUTES of the
seminal
version being posted, followed up by
'warnings' on the editor's talk
page.
> I've seen edits reverted because the formatting of the source was
wrong.
It
> should be a basic pillar that before reverting, we see if we can
improve/
> fix the problem. Undoing a newcomer's
work and leaving something like
> WP:MOS as an edit summary is not helpful - if you are going to cite a
WP
> > policy, then do so by pointing directly to the specific page where
the
> new
> > editor can read about it. I know it is time-consuming to fill in edit
> > summaries, especially if one is doing a series of identical edits to
a
> > whole lot of pages. But we can use
technology to speed this up - on a
> blank
> > edit summary, a prompt will suggest earlier text and you can select
an
>
applicable one. On an edit summary with a reference to the section of
the
> page this does not work - so we need to find
a way around this, like
> splitting the field.
>
> No amount of ink about how welcoming WP is to new editors, IT IS NOT.
For
> > reference, this section has some interesting facts,
> >
https://en.wikipedia.org/wiki/Wikipedia#Contributors.
> >
> > We are also losing established editors, mostly because of edit
warring.
> > There are blocks coalescing around all
kinds of themes and issues and
> these
> > defend their turf.
> >
> > Pages that contain controversial details should display a specific
> notice -
> > not difficult to do, given the array of templates already in use.
Some
> > pages are the result of a compromise
reached after acrimonious
debate.
An
> > editor - old or new - who was not involved in discussions will not
know
> this
and might make an edit that detonates the powder keg and starts
the
> > war all over again. It would be so easy to display a notice on the
EDIT
> > PAGE saying something like "Hi, if
you were planning to edit .....[ x
> > detail] ... please read (link) the discussion and resolution on
this. I
am
> pretty convinced it would work far better than having thousands of
pages
> > locked ([semi-]protected). Some pages just require a simple message
on
the
> EDIT PAGE such as (example) "In the English Wikipedia we use the
spelling
> > *Braganza* and not *Bragança* when referring to the House of
Braganza.
Please do not change this.". There are 1,300
pages where Braganza is
mentioned, imagine how many headaches we could spare ourselves.
Some editors seem to derive pleasure from the constant reverting/
protecting - you soon get to know who the 'group' is and can read on
their
> talk pages comments and jokes about a "here we go again" scenario. It
is
as
> if they actually lie in wait for the next unwary editor to come along
and
> > make a change.
> >
> > At the same time, there are hundreds of thousands of pages that do
not
meet
> 20% of the quality criteria and nobody does anything to remedy them.
Yet,
> > do something like move the page, change the infobox and immediately
the
'owners' come out of the woodwork to revert.
Someone cited Ukranian in this thread and I would like to pick up on
that.
There is a tendency at the higher levels to
equate Wikipedia with the
English Wikipedia and all else are something else. This includes the
level
> of involvement by the Foundation etc in the non-English Wikipedias,
often
> with the justification (excuse?) that each
is independent. And of
course
> > each language WP will use this independence to its advantage when
> > convenient, as a reason why this or that is being done differently.
In
the
same breath, content that is specifically marked
as referring to the
En-WP
is then regurgitated as if it reflects the whole
WP, as here, in the
Portuguese WP:
https://pt.wikipedia.org/wiki/Confiabilidade_da_Wikip%C3%A9dia#Avalia.C3.A7…
> >
> > Independence is well and good, but not when for example the
Portuguese
WP
votes/
debates/ discusses/ relaxing sourcing policies. If WP is to be
judged on its reliability then on a number of key elements it must be
held
> to one standard with criteria that apply across the board. We can't
have
> different standards on reliability of
sources, notibality, etc.
>
> To shrug it off as an issue of the Portuguese WP is to bury our heads
in
> > the sand, to shirk responsibility, because such issues are
symptomatic
of
the
problems facing the WP as a whole and contributing to the reasons
that
make editors pack up and go.
Also from Portuguese WP, it is embarassing that since 2009 there have
been
> all kinds of processes to arrive at a solution for what to call pages
on
> > animals and plants - eg: cattle/ bull/ ox/ cow/ bos ... By the looks
of
it,
[[Cattle]] in the English WP has been locked for
years for the same
reason.
> This kind of thing snowballs and then other aspects come into play,
> overflow and contaminate other areas of the WP as if by contagion.
>
> James, from the link you provided, I see a reference to bias. We all
have
> our 'usual beats' but we all also
edit anywhere where we might happen
to
> > find something wrong. In doing that, you soon find out that just
about
> each
> > page has 'owners', usually 3 or 4 and these work as a team to
preserve
> > their way of seeing it. Very worrying
is that a lot of this happens
on
> > pages on big corporations, which raises
the spectre of the
possibility
(already proven) of 'editors' working for
money. Equally nefarious, I
have
noted a group of editos (5 or 6, plus socks [some
exposed, others
suspected] and countless IP accounts) who are active on a few hundred
pages
> deleting/ sanitising negative references to CIA/ US (and 'allies')
> involvement in right-wing coups all over the world and generally
anything
> > unsavoury about the US in all pages on conflicts in which the US has
> taken
> > part.
> >
> > In my experience, resolution mechanims for situations such as any
that
fit
> any of the cases above tend to favour the status quo. I have
investigates
> > some of these cases and it is quite apparent that in many cases the
> 'admin'
> > taking a decision is also part of group that is trying to defend a
> certain
> > point of view.
> >
> > Finally, I think it is time to think seriously and hard about
anonymous
> > (IP) editing. We can all be anonymous,
so with a username you are not
> less
> > so. I do believe that IPs who make a few edits here and there, often
> > unconstructive, would stop if they were not serious and do not want
to
>
bother registering. Conversely, one you register, it is as if you
become
> officially a member. It is unlikely that one
would bother registering
and
then
engage in vandalism and unconstructive editing.
Best regards,
Rui
_______________________________________________
Wikimedia-l mailing list, guidelines at:
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
Wikimedia-l(a)lists.wikimedia.org
Unsubscribe:
https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
<mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>
--
_________________________
Rui Correia
Advocacy, Human Rights, Media and Language Work Consultant
Bridge to Angola - Angola Liaison Consultant
Mobile Number in South Africa +27 74 425 4186
Número de Telemóvel na África do Sul +27 74 425 4186
_______________
_______________________________________________
Wikimedia-l mailing list, guidelines at:
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
Wikimedia-l(a)lists.wikimedia.org
Unsubscribe:
https://lists.wikimedia.org/mailman/listinfo/wikimedia-l,
<mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>