Discussion:
[Wikimetrics] Gender identification of a cohort?
Jami Mathewson
2014-07-07 21:38:43 UTC
Permalink
Hi, Wikimetrics folks!

I mentioned to Dan a while back that I would find it super useful if gender
(for those who self-identify on their user accounts) were incorporated into
the cohort information on Wikimetrics. Since this isn't currently
available, Dan mentioned asking others if they would find it useful.

If you would find it useful to be able to pull the gender of your cohorts,
please let us know! I personally work with students and would like to see
how much content our women are adding to Wikipedia, if there are any
differences in editing behavior from a gender standpoint, which academic
disciplines draw more women onto Wikipedia, etc.

Thanks!
Jami
--
Jami Mathewson
Program Manager
Wiki Education Foundation
jami at wikiedu.org <jami at wikiedfoundation.org>
User:Jami (Wiki Ed) <https://en.wikipedia.org/wiki/User:Jami_(Wiki_Ed)>
@WikiEducation <https://twitter.com/WikiEducation>
wikiedu.org

*Our organization supports the Wikipedia Education Program in the United
States and Canada.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.wikimedia.org/pipermail/wikimetrics/attachments/20140707/ef0e7030/attachment.html>
Tighe Flanagan
2014-07-07 21:51:53 UTC
Permalink
Hi Jaime, and everyone,

+1!

That would definitely be useful for me as well. Since the education
programs I work with are relatively small, we can usually suss out some
data informally or through surveys — but it would be great to be able to do
more than say X number of our participants are women. Saying X amount of
content is from female editors, and being able to do more data analysis
that Jaime mentioned would be awesome and could help inform decisions we
make about Wikipedia Education Programs in the future :)

Tighe

--
Tighe Flanagan
Manager, Wikipedia Education Program
Wikimedia Foundation
+1.415.839.6885 x6880
tflanagan at wikimedia.org
education.wikimedia.org
Post by Jami Mathewson
Hi, Wikimetrics folks!
I mentioned to Dan a while back that I would find it super useful if
gender (for those who self-identify on their user accounts) were
incorporated into the cohort information on Wikimetrics. Since this isn't
currently available, Dan mentioned asking others if they would find it
useful.
If you would find it useful to be able to pull the gender of your cohorts,
please let us know! I personally work with students and would like to see
how much content our women are adding to Wikipedia, if there are any
differences in editing behavior from a gender standpoint, which academic
disciplines draw more women onto Wikipedia, etc.
Thanks!
Jami
--
Jami Mathewson
Program Manager
Wiki Education Foundation
jami at wikiedu.org <jami at wikiedfoundation.org>
User:Jami (Wiki Ed) <https://en.wikipedia.org/wiki/User:Jami_(Wiki_Ed)>
@WikiEducation <https://twitter.com/WikiEducation>
wikiedu.org
*Our organization supports the Wikipedia Education Program in the United
States and Canada.*
_______________________________________________
Wikimetrics mailing list
Wikimetrics at lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikimetrics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.wikimedia.org/pipermail/wikimetrics/attachments/20140707/870fd55d/attachment.html>
Steven Walling
2014-07-07 22:38:21 UTC
Permalink
Post by Jami Mathewson
I mentioned to Dan a while back that I would find it super useful if
gender (for those who self-identify on their user accounts) were
incorporated into the cohort information on Wikimetrics
Hi Jami,

Do you mean the gender specified in a user's preferences?

Especially with the current design,[1] I would encourage us not to rely on
this preference as an accurate representation of the gender distribution of
Wikimedia accounts.

The purpose of this preference is for software localization, not developing
a profile of users. Consider that users have to wade through Preferences in
order to find this, and there is little incentive for them to fuss with it
if their language doesn't have a grammatical gender. It's not likely to be
reliable data on the topic. Tighe's approach is probably much more
accurate, especially when dealing with new editors.

1. Screenshot Loading Image...
--
Steven Walling,
Product Manager
https://wikimediafoundation.org/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.wikimedia.org/pipermail/wikimetrics/attachments/20140707/ef281894/attachment.html>
Jami Mathewson
2014-07-07 22:44:45 UTC
Permalink
But if they don't find it, they just wouldn't select it, right? Or was your
image suggesting it defaults to "he" unless they change it?

I'm not actually looking to look at the balance for all user accounts but
am looking to identify the women in a particular cohort to see what content
they are adding. So even having some who don't identify would be ok, so
long as I could get a better data set from those who do actually make the
choice.

I understand if that can't be made a priority, though, because, globally,
there are languages that don't have a grammatical gender. Is there any
other way to pull that data publicly?

(My cohorts are quite large, and surveys rarely get enough responses to
give me very good data).




On Mon, Jul 7, 2014 at 3:38 PM, Steven Walling <swalling at wikimedia.org>
Post by Steven Walling
Post by Jami Mathewson
I mentioned to Dan a while back that I would find it super useful if
gender (for those who self-identify on their user accounts) were
incorporated into the cohort information on Wikimetrics
Hi Jami,
Do you mean the gender specified in a user's preferences?
Especially with the current design,[1] I would encourage us not to rely on
this preference as an accurate representation of the gender distribution of
Wikimedia accounts.
The purpose of this preference is for software localization, not
developing a profile of users. Consider that users have to wade through
Preferences in order to find this, and there is little incentive for them
to fuss with it if their language doesn't have a grammatical gender. It's
not likely to be reliable data on the topic. Tighe's approach is probably
much more accurate, especially when dealing with new editors.
1. Screenshot https://i.imgur.com/NsoZJeL.png
--
Steven Walling,
Product Manager
https://wikimediafoundation.org/
_______________________________________________
Wikimetrics mailing list
Wikimetrics at lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikimetrics
--
Jami Mathewson
Program Manager
Wiki Education Foundation
jami at wikiedu.org <jami at wikiedfoundation.org>
User:Jami (Wiki Ed) <https://en.wikipedia.org/wiki/User:Jami_(Wiki_Ed)>
@WikiEducation <https://twitter.com/WikiEducation>
wikiedu.org

*Our organization supports the Wikipedia Education Program in the United
States and Canada.*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.wikimedia.org/pipermail/wikimetrics/attachments/20140707/448e3f43/attachment.html>
Nuria Ruiz
2014-07-08 13:56:06 UTC
Permalink
So even having some who don't identify would be ok, so long as I could get
a better data set from those who do actually make the choice.
I think Steven's point was that not enough people make the choice for the
gender data to be significant.

Before actually adding gender we should look at whether that data is
populated in a wide enough array of users.
But if they don't find it, they just wouldn't select it, right? Or was
your image suggesting it defaults to "he" unless they change it?
I'm not actually looking to look at the balance for all user accounts but
am looking to identify the women in a particular cohort to see what content
they are adding. So even having some who don't identify would be ok, so
long as I could get a better data set from those who do actually make the
choice.
I understand if that can't be made a priority, though, because, globally,
there are languages that don't have a grammatical gender. Is there any
other way to pull that data publicly?
(My cohorts are quite large, and surveys rarely get enough responses to
give me very good data).
On Mon, Jul 7, 2014 at 3:38 PM, Steven Walling <swalling at wikimedia.org>
Post by Steven Walling
Post by Jami Mathewson
I mentioned to Dan a while back that I would find it super useful if
gender (for those who self-identify on their user accounts) were
incorporated into the cohort information on Wikimetrics
Hi Jami,
Do you mean the gender specified in a user's preferences?
Especially with the current design,[1] I would encourage us not to rely
on this preference as an accurate representation of the gender distribution
of Wikimedia accounts.
The purpose of this preference is for software localization, not
developing a profile of users. Consider that users have to wade through
Preferences in order to find this, and there is little incentive for them
to fuss with it if their language doesn't have a grammatical gender. It's
not likely to be reliable data on the topic. Tighe's approach is probably
much more accurate, especially when dealing with new editors.
1. Screenshot https://i.imgur.com/NsoZJeL.png
--
Steven Walling,
Product Manager
https://wikimediafoundation.org/
_______________________________________________
Wikimetrics mailing list
Wikimetrics at lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikimetrics
--
Jami Mathewson
Program Manager
Wiki Education Foundation
jami at wikiedu.org <jami at wikiedfoundation.org>
User:Jami (Wiki Ed) <https://en.wikipedia.org/wiki/User:Jami_(Wiki_Ed)>
@WikiEducation <https://twitter.com/WikiEducation>
wikiedu.org
*Our organization supports the Wikipedia Education Program in the United
States and Canada.*
_______________________________________________
Wikimetrics mailing list
Wikimetrics at lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikimetrics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.wikimedia.org/pipermail/wikimetrics/attachments/20140708/0b95b314/attachment.html>
Loading...