Re: [Wikimedia-l] Page views of male/female biographies?

[David Cuenca Tudela]

Of course, there is more than one way to skin a potato, but it doesn't mean
that those ways are useful, desirable, or informative. You say that readers
are more likely to access people who are named, than people who are
notable, but isn't that relevant? If notable people are not named, then we
can point at the issue, bring the information to the light and ask for
measures to be taken. Because as it is now, it seems as if we are at the
mercy of what others decide that is relevant, however I believe that the
community also can have a say in identifying media blind spots and
reporting them to the public.

I am not asking for "ongoing research", I am asking for data to back our
claims that wikipedia reflects the bias of the media. OTOH, for research
purposes it would be interesting to:
- evaluate the distribution of sources by gender and area of expertise
- correlation between page views and sources

"It should be noted that there is also an inherent bias in that there are
far fewer biographical articles about women in most categories, as compared
to men."

That is not so injurious. If we have 80% articles about men and 20%
articles about women, however 50% of pageviews go to men and 50% go to
women, suddenly the gender gap would be narrower, as it would show that
women, even with a reduced number of articles, have more public exposure.
Still there would be areas of expertise where men would attract more
pageviews than women, and vice versa, but that would be ok according to my
understanding.

It should be noted however, that in the depths of the gendergap rabbit hole
there is the core of societal (and individual) values, and how individuals
are rewarded in exposure (and money) according to those values. If women
are naturally more skilled than men in certain areas, why is not that
expertise recognized and valued? If our platforms are predominately male,
does it mean that our mission has an inherent gender bias?

@Strainu: thanks for the links!

Micru



On Thu, Dec 6, 2018 at 3:45 AM Risker  wrote:

> Hmm.  I think the subject of what you call "audience bias" is far more
> general than the tiny targeted area you're talking about.  I'm pretty sure
> that readers from Poland are thousands of times more likely to access the
> Wikipedia article about [name any town in Poland] than readers in Indonesia
> are.  I'm pretty sure that readers from all over the world are far more
> likely to access articles about people who are named in other publications,
> particularly the news media, than they are about notable but comparatively
> obscure article subjects who haven't recently been the subject of public
> interest.  I do not think you have made a good case for considering the
> viewing of articles of male subjects vs. female subjects to be directly
> linked to "audience bias".  We only need to look at the top100 articles
> viewed on any project to see that what drives page views is usually some
> event external to the Wikipedia projects.
>
> Page view data is pretty readily available - it is available for every
> single page on every single Wikipedia (and probably for a lot of other
> projects too, I've just never checked).  It would require some technical
> knowledge to write a script targeting page view information for articles in
> selected categories - such as page views of articles about women scientists
> - provided there is correct and appropriate categorization of the article.
> I'm the first to admit I'm incapable of writing such a script, but there
> are lots of Wikimedians who have such skills.
>
> It certainly looks like you are asking for ongoing research to be carried
> out on a topic that interests you (and, I am certain, a lot of other
> Wikimedians). I am unclear what this kind of metric would tell us about
> "audience bias" (or any other kind of bias, for that matter), but there may
> be value in better understanding the frequency of viewing of articles in
> certain categories and comparing them to related categories; for example,
> comparing the frequency of viewing of the average article about a female
> architect as compared to a male architect.  It should be noted that there
> is also an inherent bias in that there are far fewer biographical articles
> about women in most categories, as compared to men.
>
> Risker/Anne
>
> On Wed, 5 Dec 2018 at 18:20, David Cuenca Tudela 
> wrote:
>
> > Hi Tilman,
> >
> > I disagree with your appraisal that there are better venues for my
> > question. The gendergap mailing list is technically dead, before your
> > message the last one was from April. The other mailing list is related to
> > research, not to stats that should be readily available.
> >
> > From your answer (and the lack of more information) I understand that
> there
> > is a poor (inexistent?) tracking of audience bias. In my opinion these
> data
> > would be very useful to monitor how visitors evolve with more
> availability
> > of women's biographies. I have requested it to be added to the 

[Wikimedia-l] New Wikimedia password policy and requirements

[Chris Koerner]

The Wikimedia Foundation security team is implementing a new password
policy and requirements. [0] You can learn more about the project on
MediaWiki.org. [1]

These new requirements will apply to new accounts and privileged
accounts. New accounts will be required to create a password with a
minimum length of 8 characters. Privileged accounts will be prompted
to update their password to one that is at least 10 characters in
length.

These changes are planned to be in effect on December 13th. If you
think your work or tools will be affected by this change, please let
us know on the talk page. [2]

[0] https://meta.wikimedia.org/wiki/Password_policy
[1] 
https://www.mediawiki.org/wiki/Wikimedia_Security_Team/Password_strengthening_2019
[2] 
https://www.mediawiki.org/wiki/Talk:Wikimedia_Security_Team/Password_strengthening_2019

Yours,
Chris Koerner
Community Relations Specialist
Wikimedia Foundation

___
Wikimedia-l mailing list, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and 
https://meta.wikimedia.org/wiki/Wikimedia-l
New messages to: Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, 

Re: [Wikimedia-l] Mobile fundraising ads

[Samuel Klein]

I love the focus on mobile and smaller format interfaces, quite generally;
it's increasingly how I use the projects too!

A)  This banner-text-series is clearly impactful, gave me a bit of a jump
scare, and got me to read it to find out why. I'm still not sure how I feel
about it.
~ Visual effect: Messages that flow smoothly in and out of the reading
experience are even nicer.
~ Message: Is there an estimate of the total impact on all readers, as well
as total effective fundraising?  If there is a very effective
compact/delightful banner, and an even more effective large/ambivalent
one, is there some internal calculus about the overal impact of running the
former for longer vs. the latter for a short period?
I'd like to think the best possible messages inspire and delight and
draw on positive emotions while raising funds, including for those who
don't donate, even if they do not yield the most donations per view.

B)  The tracking of whether I've donated, when choosing to show or not show
me banners, is definitely lacking.  Part of this is that we have taken an
overly-paranoid approach to gathering and anonymizing user data.  It is
entirely possible to cluster users for the purposes of
not-continuing-to-show-banners (maintain a dictionary of
user-fingerprint-hashes-already-seen, check to see if the current user is
in there, don't show banners if they are) without being able to see what
pages a given user is viewing.

I wrote more about this here:
https://blogs.harvard.edu/sj/2018/07/25/anonymizing-data-on-the-users-of-wikipedia/
 Please consider doing this; it is really hurting the user-experience of
the wiki projects (not only in this instance -- in so many other basic
instances of usage stats + testing over time!), for no benefit to anyone.
___
Wikimedia-l mailing list, guidelines at: 
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and 
https://meta.wikimedia.org/wiki/Wikimedia-l
New messages to: Wikimedia-l@lists.wikimedia.org
Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, 

Re: [Wikimedia-l] Croatian Wikipedia: persisting far-right bias?

[Yaroslav Blanter]

Hi Tomasz,

whereas you are right in theory, a practical application of this method
requires (i) availability and acceptance of all these sources in the
community (for example, if one side published in Croatian and another one
published in English, Croatian Wikipedia is likely to use only sources
produced by one side whereas the English Wikipedia is likely to use sources
produced by the other side); (ii) healthy community which is aware of the
notions of systemic bias, neutrality, and is willing to apply these notions
in their editing (for which it must be big and diverse enough so that all
notable topics get sufficiently represented). For the specific situation
with the Croatian Wikipedia, I highly doubt that we have (ii) and I am
pretty sure we do not have (i),

Cheers
Yaroslav

On Thu, Dec 6, 2018 at 1:42 PM Tomasz Ganicz  wrote:

> Vast majority of sources in controversial topics are usually biased. There
> are topics where there is in fact no any non-biased sources. And - coming
> back to my previous example, having knowledge how automatic method o bias
> measurement works it is very easy to bully it:
>
> "According to unfaithful bastard X [source X1][source X2][source X3] the
> true is A. But, according to honorable and widely recognized expert Y
> [source Y] A it is not true, but the true is B."
>
> This sentence is quite obviously biased towards B POV, but  automatic
> measurement of sources will tell you that there is bias towards A POV.  And
> this is very simple, primitive example of bias. People usually tend to do
> it in much more subtle way. Sometimes one short, completely unsourced
> sentence at the end of very long article with hundreds of citations can
> completely ruin NPOV...
>
> Or imagine that you write article about a bishop - quite naturally most
> sources will be religious POV - which does not necessarily mean that the
> article is biased as it might contain only basic facts of that person
> retrieved from official church sources. Then - following this example  - in
> Polish Wikipedia - we have probably articles about all living bishops from
> major christian denomination. But if you would want to "prove" that Polish
> Wikipedia has pro-roman-catholic POV you can easily show that we have 162
> articles about roman-catholic Polish bishops and only 12 about orthodox
> bishops. And the numbers of citations is more or less probably of the same
> proportion. Why? Simply because we have in Poland 162 catholic bishops and
> 12 orthodox. Wikipedia cannot change it obviously ;-)
>
>
>
>
>
> czw., 6 gru 2018 o 02:19 Dennis During  napisał(a):
>
> > Yes the method can miss bias. But if the references* used are* biased, it
> > would provide clear, objective (though not irrefutable) evidence of a
> > general bias.  The more factual the discussion, the more likely it will
> be
> > that any conclusions of the process will be accepted, if not by all at
> > Croatia WP, then perhaps by some there and by most other observers.
> >
> > On Wed, Dec 5, 2018 at 5:45 PM Tomasz Ganicz 
> wrote:
> >
> > > I don't clearly understand Gerard what is your idea. Do you want to
> > measure
> > > NPOV by calculating how often the sources are used after somehow
> marking
> > > them to belong to one or another group of political, religous or other
> > type
> > > of  POV? And when you find that one group of them are more often cited
> > than
> > > the others, this is a symptom of systematic bias of given Wikimedia
> > > project? Well that might be quite misleading because the issue is the
> > > honesty and context of using sources.
> > >
> > > For example: One can write an article about any controversial topic
> using
> > > equal number of  sources supporting opposite POVs, but the text can
> still
> > > be quite biased:
> > >
> > > "According to unfaithful bastard X [source X] the true is A. But,
> > according
> > > to honourable and widely recognized expert Y [source Y] A it is not
> true,
> > > but the true is B."
> > >
> > > I don't believe in any kind of automated method of measuring NPOV. NPOV
> > is
> > > very complex issue needed human judgment. You can't avoid it.
> > >
> > >
> > > śr., 28 lis 2018 o 12:43 Gerard Meijssen 
> > > napisał(a):
> > >
> > > > Hoi,
> > > > I take offence calling it a faith-based process. We have a database
> > with
> > > > the citations of all Wikipedias. We have overriding principles that
> > > include
> > > > the NPOV and what the role of functionaries is in Wikimedia projects.
> > > When
> > > > they are a faith, they are our faith.
> > > >
> > > > My question to you is, why are you reluctant to start a process that
> > will
> > > > bring down many hobby horses including yours and the ones in your
> > > favourite
> > > > project. Why not start where we face an urgency? An urgency that
> > > undermines
> > > > Wikipedia as NPOV!
> > > > Thanks,
> > > >   GerardM
> > > >
> > > > On Wed, 28 Nov 2018 at 00:31, Dennis During 
> > wrote:
> > > >
> > > > > Why not test-run the 

Re: [Wikimedia-l] Croatian Wikipedia: persisting far-right bias?

[Tomasz Ganicz]

Vast majority of sources in controversial topics are usually biased. There
are topics where there is in fact no any non-biased sources. And - coming
back to my previous example, having knowledge how automatic method o bias
measurement works it is very easy to bully it:

"According to unfaithful bastard X [source X1][source X2][source X3] the
true is A. But, according to honorable and widely recognized expert Y
[source Y] A it is not true, but the true is B."

This sentence is quite obviously biased towards B POV, but  automatic
measurement of sources will tell you that there is bias towards A POV.  And
this is very simple, primitive example of bias. People usually tend to do
it in much more subtle way. Sometimes one short, completely unsourced
sentence at the end of very long article with hundreds of citations can
completely ruin NPOV...

Or imagine that you write article about a bishop - quite naturally most
sources will be religious POV - which does not necessarily mean that the
article is biased as it might contain only basic facts of that person
retrieved from official church sources. Then - following this example  - in
Polish Wikipedia - we have probably articles about all living bishops from
major christian denomination. But if you would want to "prove" that Polish
Wikipedia has pro-roman-catholic POV you can easily show that we have 162
articles about roman-catholic Polish bishops and only 12 about orthodox
bishops. And the numbers of citations is more or less probably of the same
proportion. Why? Simply because we have in Poland 162 catholic bishops and
12 orthodox. Wikipedia cannot change it obviously ;-)





czw., 6 gru 2018 o 02:19 Dennis During  napisał(a):

> Yes the method can miss bias. But if the references* used are* biased, it
> would provide clear, objective (though not irrefutable) evidence of a
> general bias.  The more factual the discussion, the more likely it will be
> that any conclusions of the process will be accepted, if not by all at
> Croatia WP, then perhaps by some there and by most other observers.
>
> On Wed, Dec 5, 2018 at 5:45 PM Tomasz Ganicz  wrote:
>
> > I don't clearly understand Gerard what is your idea. Do you want to
> measure
> > NPOV by calculating how often the sources are used after somehow marking
> > them to belong to one or another group of political, religous or other
> type
> > of  POV? And when you find that one group of them are more often cited
> than
> > the others, this is a symptom of systematic bias of given Wikimedia
> > project? Well that might be quite misleading because the issue is the
> > honesty and context of using sources.
> >
> > For example: One can write an article about any controversial topic using
> > equal number of  sources supporting opposite POVs, but the text can still
> > be quite biased:
> >
> > "According to unfaithful bastard X [source X] the true is A. But,
> according
> > to honourable and widely recognized expert Y [source Y] A it is not true,
> > but the true is B."
> >
> > I don't believe in any kind of automated method of measuring NPOV. NPOV
> is
> > very complex issue needed human judgment. You can't avoid it.
> >
> >
> > śr., 28 lis 2018 o 12:43 Gerard Meijssen 
> > napisał(a):
> >
> > > Hoi,
> > > I take offence calling it a faith-based process. We have a database
> with
> > > the citations of all Wikipedias. We have overriding principles that
> > include
> > > the NPOV and what the role of functionaries is in Wikimedia projects.
> > When
> > > they are a faith, they are our faith.
> > >
> > > My question to you is, why are you reluctant to start a process that
> will
> > > bring down many hobby horses including yours and the ones in your
> > favourite
> > > project. Why not start where we face an urgency? An urgency that
> > undermines
> > > Wikipedia as NPOV!
> > > Thanks,
> > >   GerardM
> > >
> > > On Wed, 28 Nov 2018 at 00:31, Dennis During 
> wrote:
> > >
> > > > Why not test-run the process on my favorite project - or yours?  We
> > > should
> > > > get started.
> > > >
> > > > I am skeptical of the quality of judgment without a foundation of
> > facts.
> > > > At Wiktionary we have two main definition evaluation processes, one
> > > > dependent on citations to which interpretative judgment is applies.
> IMO
> > > > this process works very well.  The other depends on opinion, votes,
> > > > supported by whatever facts or authority or bluster (my specialty)
> > > > advocates bring to bear.  That process, though adequate, is not as
> > > > satisfactory.
> > > >
> > > > Gerard Meijssen has suggested a faith-based process. If it is almost
> > > ready
> > > > to go, let it be validated and put to use.
> > > >
> > > > On Tue, Nov 27, 2018, 16:45 Benjamin Lees  wrote:
> > > >
> > > > > On Mon, Nov 26, 2018 at 8:06 AM Dennis During 
> > > > wrote:
> > > > > >
> > > > > > Who is the judge? Are we going to join Facebook, Google, Twitter,
> > et
> > > al
> > > > > as
> > > > > > the new press barons?
> > > > >
>