Thanks everyone who’s responded with such thoughtful comments. I will look further at the "The Pipeline of Online Participation Inequalities: The Case of Wikipedia Editing” paper that Isaac referenced, which seems interesting.
I think the plans that Rebecca Maung described for the 2021 Community Insights survey are bang on. If we want to help answer the question of whether particular races or ethnicities are under-represented, it makes sense to gather the data in a form that can be compared to the general population in a country. E.g. I’d like to have data that say, “X percent of U.S.-based Wikipedia editors identify as Black, whereas according to the latest census, X percent of the U.S. general population identifies as Black.” This is more interesting than data that say, “X percent of all Wikipedia editors in the world identify as Black”.
Another thought is that diversity within each Wikimedia project, not just diversity across all projects, is something people value - especially in the big projects like the English Wikipedia. So it would be helpful to see racial/ethnic identity, country of residence, and gender percentages for each project.
And another thought is that while some people will be offended if you ask them about their race, other people will be offended if you put out a survey that’s intended to measure diversity and doesn’t ask anyone about their race.
Cheers, Su-Laine
On Sep 21, 2020, at 1:39 PM, L.Gelauff lgelauff@gmail.com wrote:
Ah, you're assuming some automated country-detection, rather than self-identify. I see.
Lodewijk
On Mon, Sep 21, 2020 at 12:59 PM Stuart A. Yeates syeates@gmail.com wrote:
Everyone from China and Saudi Arabia (two countries which systematically block wikipedia) are likely to be taking technical measures to disguise their country.
That's a lot of people, but I'm not sure how many editors that is.
cheers stuart
-- ...let us be heard from red core to black sky
On Tue, 22 Sep 2020 at 07:01, L.Gelauff lgelauff@gmail.com wrote:
Just thinking out loud.. are we looking for actual race/ethnicity/etc
data,
or is it rather that we're looking for whether someone belongs to an
under
represented group in their specific situation? If it is the latter, there may be ways to phrase the question without asking for actual
demographics.
Stuart; do you have any indication for how large a portion that group
is? I
am aware of public pages being potentially disguised as such, but wasn't familiar with stories about this happening in a survey context (although
it
does not sound implausible).
Best,
Lodewijk
On Mon, Sep 21, 2020 at 11:39 AM Stuart A. Yeates syeates@gmail.com
wrote:
Another point not touched on by other commenters is that even if ideal race / ethnicity question(s were developed for every country in the world, users from some countries commonly disguise their country due to censorship in that country, so we there would be a whole class of systematic errors where we asked users the wrong country's question(s).
cheers stuart -- ...let us be heard from red core to black sky
On Tue, 22 Sep 2020 at 05:00, Isaac Johnson isaac@wikimedia.org
wrote:
Adding another point from Rebecca Maung who helps run the annual
Community
Insights surveys <https://meta.wikimedia.org/wiki/Community_Insights
but
isn't currently on this listserv so couldn't respond directly:
This year's Community Insights survey (reporting scheduled for early
is the first that will ask Wikimedia contributors about race and ethnicity-- but only in certain geographies. Due to all the excellent points made in this thread, we have never asked a race or ethnicity question, but this year we decided to start asking locally relevant questions where we could. This year only editors in the US and
Britain
will
see a question about race or ethnicity, tailored to their local
contexts.
In the coming years, we will expand the countries and geographies
that
see
a question like this, prioritizing places where there is a larger
editor
presence and local laws and norms allow such questions. We have not
yet
discussed asking about religion in the Community Insights survey.
On Mon, Sep 21, 2020 at 9:20 AM Isaac Johnson isaac@wikimedia.org
wrote:
As pointed out by others, the highly contextualized nature of
religion,
race, and ethnicity between countries makes it very difficult to
impossible
to craft questions that are not overly reductive but still somewhat universal. Despite this challenge, understanding diversity in a way
that
captures these aspects is obviously quite important as they often
figure
very strongly into power and representation within history, media,
etc.
In general, if you're looking for large-scale surveys of editors,
the
Meta
category (Category:Editor surveys https://meta.wikimedia.org/wiki/Category:Editor_surveys) is
actually
quite complete (same for readers https://meta.wikimedia.org/wiki/Category:Reader_surveys). In particular, I wrote what little I could find about these topics
into
this
section of our recently published knowledge gaps taxonomy: https://arxiv.org/pdf/2008.12314.pdf#subsubsection.3.1.7
The April 2011 editor survey took the approach of just asking
people
how
they felt they were different from others in the community -- this
specific
question is not one that I would advocate today (asking people to
identify
all the ways in which they may be "outsiders" is not particularly welcoming) but this is also probably the style of approach (asking
people
how well they feel represented within Wikipedia content or editor community) that you'd have to take to get information on ethnicity
/
race /
religion without writing country-specific questions:
https://upload.wikimedia.org/wikipedia/commons/7/76/Editor_Survey_Report_-_A...
On Mon, Sep 21, 2020 at 6:12 AM Stuart A. Yeates <
syeates@gmail.com>
wrote:
> The ethnicity / race question is an incredibly hard question to > compose in an internationalised way. > > Pretty much every country in the world uses different terms and
there
> are some very confusing cases where the same term is used in
different
> countries to mean very different things (e,g, "Asian" in UK
English vs
> New Zealand English). This is derived from varying legal
definitions
> (for example blood quantum vs one-drop laws); the history of > colonisation and waves of immigration to the country; along with > cultural differences. > > cheers > stuart > -- > ...let us be heard from red core to black sky > > On Mon, 21 Sep 2020 at 21:55, Federico Leva (Nemo) <
nemowiki@gmail.com>
> wrote: >> >> Su-Laine Brodsky, 21/09/20 08:19: >>> I’m wondering if any large-scale surveys have been done that
ask
> Wikipedia editors about their race, ethnicity, or religion? >> >> What international standards exist to phrase such questions? >> Denominations commonly used in surveys in one country may be
considered
>> horrific or even illegal in others. >> >> I see OECD considers it a difficult problem too: >> >> ---- >> >> 76. Current NSOs collection practices cluster around three
broad
>> categories: 1) all OECD countries collect information on some
diversity
>> proxies such as country of birth (36 OECD members); 2) a small
majority,
>> mostly Eastern European countries, the United Kingdom and
Ireland,
>> gather additional information on race and ethnicity (16 OECD
members);
>> and 3) only a handful of countries in the Americas and Oceania
collect
>> data on indigenous identity (6 OECD members). Diversity
statistics
are
>> collected from the perspective of either enumerating the size
of the
>> relevant populations (typically in the census) or of comparing >> well-being outcomes across different population groups. >> >> 77. While privacy and human rights legislation sometimes
prevents
or
>> discourages the routine collection of diversity data, the need
to
>> improve data availability and quality is being recognised in
most
>> countries. Many countries are piloting the addition of new
ethnic
>> response options to more accurately reflect the make-up of their >> societies (e.g. Ireland, the United States), while Belgium is >> considering allowing collection of race and ethnicity data
within
the
>> restrictions imposed by the national legal framework. Within the >> European Statistical System, the inclusion of more detailed
migration
>> information is also being considered: The Framework Regulation
for
>> Production of European Statistics on Persons and Households
European
>> foresees the incorporation of questions on the country of birth
of
the
>> respondent’s parents in the Labour Force Surveys (from 2020),
the
>> European Health Interview Survey, the European Union Statistics
on
>> Income and Living Conditions, the Household Budget Surveys and
the
>> Community surveys on ICT usage in households and by
individuals. The
>> European Union Agency for Fundamental Rights is pursuing its
Roma
and
>> Travellers Survey to collect comparable data in six selected
Member
>> States in 2018 (FRA, 2018[77]). >> >> ---- >> >> >
https://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=SDD/DO...
>> >> Federico >> >> _______________________________________________ >> Wiki-research-l mailing list >> Wiki-research-l@lists.wikimedia.org >> https://lists.wikimedia.org/mailman/listinfo/wiki-research-l > > _______________________________________________ > Wiki-research-l mailing list > Wiki-research-l@lists.wikimedia.org > https://lists.wikimedia.org/mailman/listinfo/wiki-research-l >
-- Isaac Johnson (he/him/his) -- Research Scientist -- Wikimedia
Foundation
-- Isaac Johnson (he/him/his) -- Research Scientist -- Wikimedia
Foundation
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l