Wikipedia:Wikipedia Signpost/2015-06-03/News and notes

News and notes

Three new community-elected trustees announced, incumbents out

The electorate

The election saw a sharp increase in the number of voters, nearly tripling from just 1809 in the previous community election two years ago to 5167 this time. The Foundation's James Alexander has posted two interesting statistical tables. One is on the voter turnout over the course of the two-week election, which shows signs of increased voter interest at a few points in time. The other is on turnout by wiki as a proportion of total voters and in relation to the total eligible voters on each wiki.

The English Wikipedia was home to the largest proportion of voters (31.6%), followed by the German (12.0%), French (6.9%), Italian (5.8%), Russian (5.7%), and Spanish (4.8%) versions. Together, these six sites accounted for almost two-thirds of the total votes cast. Aside from some of the smaller sites, the proportion of eligible voters who actually voted was highest in the Ukrainian Wikipedia (25.2%), followed by the Arabic (17.7%), Italian (16.1%), Farsi (15.1%), and Polish (12.5%) versions. Of those eligible on the English Wikipedia, 8.3% voted; other large Wikipedias managed better: German (11.0%), French (10.8%), Italian (16.1%), Russian (10.8%), and Spanish (11.2%). Retiring trustee Phoebe did point out to the Wikimedia mailing list that some editors are active on more than one site, which may affect the fine resolution of these data.

The results in detail

The results of the Board election mean that there has been a clean sweep of positions by white males in all three WMF elections: FDC, FDC ombudsperson, and the Board. After the announcement of the three trustees, this was the cause of heated discussion on Facebook, among thanks and compliments to the three incumbents: "So now 2/10 Board members will be women, and only one from outside Europe or North America?" Phoebe Ayers replied: "Yes, the new appointees are great but I was proud of us for having a gender-balanced board, which is so rare in both nonprofits and corporations. The current trustees have already discussed making this a priority for future appointed seats."

It was Andrew Lih (Fuzheado) who pointed out that "the two female candidates had the 1st and 3rd most votes in this election, but the oppose votes countered this. ... I have to say this year's elections were a bit odd in that the voting method wasn't well publicized or easily discoverable until the ballot box opened. Previous elections used the Schulze method (amended: though last year was also S/S+O)." Dariusz Jemielniak wrote: "Gender diversity took a major hit. ... opposing votes are highly controversial, also because different cultures may be more or less averse to them". Current Board chair Jan-Bart de Vreede (Jan-Bart) wrote that "the Q&A is heavily slanted towards the English speaking community and a few were able to dominate (also issues we have to fix)." He continued: "we really should look at changing the election system so that it will go towards solving [the diversity problem]".

For years, the Signpost's coverage has emphasised the support votes rather than the full data generated by the ternary support–neutral–oppose system (apparently imported from the English Wikipedia's Arbitration Committee elections in 2013). Among other issues, the S/(S+O) formula greatly inflates the appearance of electoral "percentage" support for the candidates. Thus we have set out the numbers of support votes in the table below, with the percentage of all voters who supported each candidate, and the orders of voting strength both in terms of support votes alone and the formula that counts towards electoral success or otherwise. Red shows candidates whose ranking was reduced by the formula, and blue shows those whose ranking was increased by the formula. This appears to be the second election in which the S/(S+O) system has made a substantive difference to the outcome; two candidates' positions on the success–failure boundary were inverted in the 2013 English Wikipedia Arbitration Committee election.

Candidate	Ranking based on "support"	Adjusted ranking: "support–oppose" formula	No. of support votes	Percentage of voters supporting
María Sefidari	1	↓4	2184	42.3%
Dariusz Jemielniak	2	↑1	2028	39.2%
Phoebe Ayers	3	↓5	1955	37.8%
James Heilman	4	↑2	1857	35.9%
Denny Vrandečić	5	↑3	1628	31.5%
Tim Davenport	6	↓9	1571	30.4%
Mike Nicolaije	7	↑6	1524	29.5%
Peter Gallert	8	↑7	1467	28.4%
Cristian Consonni	9	↑8	1381	26.7%
Samuel Klein	10	10	1330	25.7%
David Conway	11	11	1192	23.1%
Ali Haidar Khan (Tonmoy)	12	↓13	1134	21.9%
Mohamed Ouda	13	↓15	1112	21.5%
Edward Saperia	14	14	1109	21.5%
Josh Lim	15	↑12	1969	20.7%
Sailesh Patnaik	16	16	1010	19.5%
Syed Muzammiluddin	17	17	816	15.8%
Nisar Ahmed Syed	18	18	735	14.2%
Houcemeddine Turki	19	19	590	11.4%
Francis Kaswahili Kaguna	20	↓21	386	7.5%
Pete Forsyth (withdrew)	21	↑20	108	2.1%

Gregory Varnent, of the election committee, has linked people to the post-mortem page for ideas and discussion.

The new trustees' views

In our coverage before voting began, we presented numerical displays and analysis of the candidates' views on five propositions and ten "priorities" we had put to them. A 1–5 Likert scale for the propositions ranged from "strongly agree" (1) to "strongly disagree" (5), with a neutral/opt-out "3". We received responses from all but Francis Kaguna, and Houcemeddine Turki got back to us after copy-deadline; we have now included Houcemeddine's data in the averages for candidates who were unsuccessful, and compare those averages with those of the three new trustees.

As expected, the individual trustees track differently from the averages. Dariusz and Denny are more favourable than the average towards merging the two affiliate-selected with the three community-elected Board seats in future elections. Given his background in computer science, Denny is relatively keen to appoint more tech experts as trustees, while Dariusz and James are yet to be convinced of this notion. All three new trustees favour using the $27M in Foundation reserves to seed-fund the new endowment, two of them strongly so. Dariusz and Denny are strongly against the idea of completely forbidding paid editing, whereas James is neutral on this, perhaps given his experience in discovering large amounts of plagiarism and paid editing both on- and off-wiki. (He has written about his experiences with paid editors and plagiarism in Signpost op-eds.)

Comparing the trustees' rankings from 1 to 10 of the 10 priorities we had put to them against average rankings by the other candidates revealed sometimes-stark differences between each of them, and between them and the others. Dariusz and James rate increasing global-south participation significantly lower (7th and 6th) than the average, while Denny rates it above the average (2nd). James and Denny score increasing editor retention at 2nd and 1st, above the average of nearly 4th, while Dariusz scores it only 6th. Investing in mobile tech attracts favourable rankings from Dariusz and James (3rd and 4th), but interestingly, Denny ranks it below average, at 6th. Investing more in collecting data is a significantly lower priority for Dariusz (9th)and Denny (8th) than the average for the other candidates and for James (around the 5th priority). All three trustees spurn the notion of funding more offline meetups, with straight 10s, against an average of a little higher than 7th. Implementing VisualEditor gains favour from Dariusz (4th, against the average of lower than 7th), but James rates this 9th and Denny is close to the average. Denny is strong on reducing the gender gap (3rd), but James is not (9th), and Dariusz tracks the average at 5th. Advocating internet freedom is 7th, 8th, and 9th among the new trustees, against an average of about 7th.

The other candidates rate allocating resources to the engineering challenge between 5th and 6th priority. Here the new trustees beg to differ in greater favour of the notion. Dariusz rates engineering to improve readers' experience as his very top priority; it is James' 3rd priority, and unexpectedly Denny's 5th, close to the average. Dariusz rates engineering to improve editors' experience a little lower than he did for readers' experience—2nd, while James and Denny are keener on this aspect (3rd becomes 1st, and 5th becomes 4th, respectively). This might make for interesting conversations with the WMF's executive director, Lila Tretikov.

Contacting your representatives

We asked the three community-elected trustees whether they are happy for their constituents to contact them as their representatives on the Board, and if so, what mode of communication they would prefer. Dariusz wrote: I think that on-wiki method of communication is best for most cases, and for delicate matters email may be preferred (I can be reached through "email this user" feature, and my email is also publicly available). Denny says "obviously" he is happy to communicate: "For now, my talk pages would be best—either on the Croatian, German, or English Wikipedias, or Wikidata or Meta." James nominated his talkpages on Meta or the English Wikipedia, or email function (his email address is also widely known). "Twitter is not as good. And I don't check Facebook."

Brief notes

FDC round 2 results announced: The Funds Dissemination Committee, the body in charge of distributing Foundation funds for groups affiliated with the movement, have released their results for funding requests for this year's second round of annual plan grant funding. For more information on the proposals refer to previous Signpost coverage.
Wikimedia UK organizes Wikipedia science conference: Wikimedia UK published a post to their blog announcing the organization of a conference jointly organized by Wikimedia UK and by the Wellcome Trust. The conference will be held in a Wellcome Trust conference center and attended to by researchers both in the UK and attached to the WMF, exploring "open access, wiki communities, and the scientific process"; an affiliated hack athon has been organized with financial support from CrossRef.
EventZoom Developer Fred Johansen published a guest post to the Wikimedia Germany blog highlighting the use of Wikidata for event visualization via the EventZoom website.
Translatewiki.net efforts: In a post to the WM Blog Wikimedia Sweden highlighted the results of their efforts organizing a translation rally on translatewiki.net; see also previous coverage for more information on translation efforts involving the site.
Change in AffCom procedures emplaced: Following the presentation and discussion of the previously mooted relaxing of Affiliations Committee procedures, a resolution putting those changes into effect has now been committed. Refer to previous Signpost coverage for more details on the nature of the changes.
Education Report: The May edition of the Education Report has been published.

← Previous "News and notes"

Next "News and notes" →

In this issue

3 June 2015

News and notes

Blog

Discussion report

Discuss this story

These comments are automatically transcluded from this article's talk page. To follow comments, add the page to your watchlist. If your comment has not appeared here, you can try purging the cache.

Blue? I don't see any blue names in the ranking chart, but I do see green ones. I've double checked my settings and looked from three different devices, and it looks green on all of them, not blue. Incidentally, red and green are the colours most difficult for colour-blind individuals to differentiate. Risker (talk) 18:59, 6 June 2015 (UTC)[reply]

Risker, thanks, and fixed. I completely forgot about red–green when bumbling about with the colours. Tony (talk) 08:19, 7 June 2015 (UTC)[reply]

Mooted? Please be careful using words that have very different or even nearly-opposite meanings in different parts of the English-speaking world. In some countries such as England, "moot" typically means "open for discussion/presented as an idea." In other countries, "moot" typically means "unimportant" or "reduced it unimportance." davidwr/(talk)/(contribs) 19:30, 6 June 2015 (UTC)[reply]

Good point. The use of 'moot' as a verb meaning 'to reduce or remove the practical significance of'[1] is an obscure use that I wasn't aware of until I looked it up in a dictionary, though the adjectival meaning of 'insignificant' is well known. (I guess it is OK to edit the article to use a better word: I don't think we should expect all Signpost authors to be familiar with obscure ambiguities around the world.) --Hroðulf (or Hrothulf) (Talk) 12:34, 8 June 2015 (UTC)[reply]

My suggestion for solving the "diversity problem," so-called, would be for Jan-Bart to take a wikibreak and resign his seat in favor of a female from outside of Europe and North America, preferably one older than 50 years old. Don't put "diversity" on the community-elected seats, put that on the board for its appointed 5 seats. The gender/age/ethnicity of those can be directly controlled. The last discretionary appointee was Guy Kawasaki, a North American male. Money where mouth is, and all that. —Tim Davenport /// Carrite (talk) 21:37, 6 June 2015 (UTC)[reply]
Both Jan-Bart and Stu are stepping down at the end of their term this December, as noted two years ago when they were reappointed. And in fact the last two appointees fit the demographic you mention. – SJ + 00:18, 7 June 2015 (UTC)[reply]

Concern trolling aside, I agree any solution to address this issue should apply to the entire board, not just the community seats. There's no reason the composition of the board can't be rearranged to have seats, either additional ones or re-purposed existing ones, designated for a specific continent, for example. Gamaliel (talk) 00:30, 7 June 2015 (UTC)[reply]

Whether or not Carrite's suggestion was genuine, it is an eminently sensible way of achieving greater diversity, is guaranteed to achieve its aim, and can be implemented now if it's thought to be that important, without needing to change voting procedures and wait 2 years for the next election. I also agree with Carrite's suggestion that age diversity is important and should be taken into account along with gender and ethnicity.--greenrd (talk) 10:56, 7 June 2015 (UTC)[reply]
I'm looking at the Facebook conversation and going "what the heck?" at some of the comments. "If you look at the voting counts, the election was, once again, dominated by major languages and projects. At this point, I seriously think the Board needs to do something in order to give more incentive to people outside of the U.S. and Europe to actually participate in our movement's politics." So the biggest projects should be disenfranchised? The English-language Wikipedia proportion of votes (30.64%) was actually less than the proportion of their eligible voters (35.38%). --NeilN ^{talk to me} 00:07, 7 June 2015 (UTC)[reply]
- I've tired of pointing out that if you don't read English, you're in the gutter on Meta election pages: relying on the availability of volunteer translators is not going to work. Is it a comfortable moral construct that "if x language community can't supply a volunteer translator, that's their fault"? WMF elections are surely a high priority for contracted translation into targeted languages. For designing a system, even on a trial basis, I'd like to see an editor survey of English-language reading ability among the non-English-language communities, particularly outside Europe.
  What proportion of Japanese Wikipedians can struggle through an English-language candidate statement, and the instructions? Only two-point-something percent of their elegible editors voted. Isn't the apparently insular Japanese community one of the elephants in our living room? If we can get the data <cough>, we'll know whether to prioritise—for example, Japanese, Arabic, Spanish, and Russian, over German and French, where a much larger proportion of eligible voters might have enough English. Don't we have stats from an editor survey some years back that show what proportion from each language group also read or edit en.WP? That would be a start, but new and better data are needed if this is to be a truly international body. I have to say that I'd rather spend money on good translation to make the movement more cohesive and more democratic through language accessibility than on some of the rather expensive offline activities I see being put to grantmaking that can make only a tenuous case for impact. Tony (talk) 03:22, 7 June 2015 (UTC)[reply]
If chapters / user groups (I am not very clear on the distinction) think that hiring translators is important, they can apply for funds / use their existing funds to do this. I don't see why the foundation has to be held responsible if relevant chapters / user groups exist and don't make an effort.--greenrd (talk) 10:59, 7 June 2015 (UTC)[reply]
I actually doubt that these chapters or user groups exist for every "relevant" language. And I think that if WMF wants diversity of candidates/voters/trustees it in fact is indeed their responsibility not to make that only hardly possible. → «« Man77 »» 11:31, 7 June 2015 (UTC)[reply]
Surely all of these chapters of ours could handle doing translations...? If the WMF asked. ResMar 16:57, 7 June 2015 (UTC)[reply]
Probably. I, however, doubt that it is their job to translate what WMF wants to have translated. But yes, most of them probably could handle that. This however fails to cover languages as unimportant as Japanese and Turkish. → «« Man77 »» 22:17, 7 June 2015 (UTC)[reply]
- Concerning the Facebook thread, I'm disappointed that this walled garden was used as a forum to discuss Wikipedia; trying to access the thread, I find I need a FB account. I don't have a FB account for the simple reason I have enough online time sinks to deal with. Something I hope folks remember in the future. -- llywrch (talk) 23:07, 7 June 2015 (UTC)[reply]
I'm very concerned by the results of the election. Denny and Dariusz don't want the the terms of use to forbid paid editing. James and Dariusz don't think that increasing the number of editors in the "global south" should be a priority, James doesn't think that it should be a priority to reduce gender gap. And neither thinks that Wikimedia should advocate for freedom of information. (Source) --NaBUru38 (talk) 23:32, 7 June 2015 (UTC)[reply]
- NaBUru, it's an order of priorities; it doesn't necessarily mean that a candidate disagrees with pursuing a priority just because others might crowd in further up the list. Tony (talk) 02:26, 8 June 2015 (UTC)[reply]

I, for one, am quite satisfied with the results of the election. ResMar 12:51, 8 June 2015 (UTC)[reply]

James and Dariusz said that their top priorities are providing more engineering resources to improve user experience. I think that the Foundation should have as first priority to develop the community. I'm very concerned that the elected members of the Board are more worried about the website design. --NaBUru38 (talk) 21:27, 8 June 2015 (UTC)[reply]

And that matches what I most want to see from the our community advocates. ResMar 02:29, 9 June 2015 (UTC)[reply]

Please flip your graphs so they're not completely counterintuitive. Also, now that we have multiple graph extensions, there's no reason for these graphs to be un-editable images. —Swpb^talk 13:30, 11 June 2015 (UTC)[reply]

I've fixed the first one; hopefully someone else can do the second. —Swpb^talk 14:55, 11 June 2015 (UTC)[reply]

Some improvements in the table, and the first graph I like, except it's now the opposite of the vertical arrangement of the second graph. I see no how-to guide for this "multiple graph extension". Tony (talk) 15:53, 11 June 2015 (UTC)[reply]

The syntax is here: mediawikiwiki:Extension:EasyTimeline/syntax. The second graph may be better arranged horizontally. There's also mediawikiwiki:Extension:Graph, but it's beta. —Swpb^talk 17:21, 11 June 2015 (UTC)[reply]

My impatience won out, and I did the second graph. —Swpb^talk 18:44, 11 June 2015 (UTC)[reply]

@Swpb: We're recruiting a data viz person at the Signpost right now. The problem with the Graphs extensions is that it's a full suite graphing utility that's hard to use directly and is very complicated, so we would love to have someone experienced with them or willing to learn them to help us out with our data visualizations. ResMar 20:47, 11 June 2015 (UTC)[reply]

@Resident Mario: Thanks for the offer. I really don't know EasyTimeline that well; my work above is very kludgey. I know you guys at Signpost are often shorthanded, and I'd love to help out when I can, but I can't responsibly commit to a regular role. —Swpb^talk 19:23, 12 June 2015 (UTC)[reply]

If you're interested in the graph extension, I recommend reading https://github.com/trifacta/vega/wiki/Tutorial . Bawolff (talk) 05:45, 12 June 2015 (UTC)[reply]

Thanks, Bawolff. I've read a little and bookmarked it. Strikes me that to get it out to editors more widely, a video tutorial, tightly produced with animation and voice-over (and subtitles into other languages), would be a good spending of donors' funds by WMF tech. The Youtube file currently available is a start, but at 52 minutes' duration is far too big a chunk for most users; I also wanted something that gets straight to process after a brief intro explaining, in 20 seconds, the advantages of interactivity and editability onwiki. Worth chopping into two to four shorter, digestible, specifically themed tutorials, don't you think, aimed at Wikimedians who know the basics of spreadsheets? Tony (talk) 08:27, 12 June 2015 (UTC)[reply]

@Tony1: Well you could say that about quite a few things related to mediawiki :). But yes, a video tutorial might be nice (I don't honestly expect anyone to make one, but it would be nice). I expect that the tutorial I linked to would be mostly useful for people who are programmers, and our target audience is not just people who know how to program (And even then, I know how to program, but the section on data transforms is still a pretty big mystery to me). Bawolff (talk) 22:20, 12 June 2015 (UTC)[reply]

@Swpb: You did that with EasyTimeline!?! I'd assumed you did it with the new Extension:Graph extension that has been enabled on-wiki. This extension enables an extensive visualization grammar called Vega, and I'm super enthusiastic about finding someone who can do Vega graphs for us, even if I don't have the time to learn to do it myself. You can see examples of the capacities of the grammar here. In the long-term this grammar will rarely be used naked: just today I ported over one Lua wrapper template, from the German Wikipedia in this case, which makes doing this much simpler. Yurik suggested on my talk page that a bot that centralizes all of the graph templates either to Mediawiki or to Meta would be the best long-term solution. I have some ideas about data you can start building visualizations of using just this template, but stuff like force graphs or tree maps would also be lovely to see. ResMar 23:01, 12 June 2015 (UTC)[reply]

llywrch You and I are in the same boat my friend. Tharthan (talk) 01:34, 16 June 2015 (UTC)[reply]

Explore Wikipedia history by browsing The Signpost archives.

Home

About