The Wikimedia Foundation's volunteer election committee has announced the election results for the three vacant seats on the Board of Trustees. Dariusz Jemielniak (Pundit), James Heilman (Doc James), and Denny Vrandečić (Denny) are set to take up their two-year terms on the Board. They will replace the three incumbents, all of whom stood this time unsuccessfully: Phoebe Ayers (phoebe), Samuel Klein (Sj), and María Sefidari (Raystorm).
Dariusz is a steward, and a bureaucrat and checkuser on the Polish Wikipedia, and has chaired the WMF's Funds Dissemination Committee, which recommends the allocation of annual operating grants for eligible affiliates, since its inception in 2012. He is a full professor of management at Kozminski University in Poland, and researches open collaboration projects such as Wikipedia and F/LOSS, narrativity, storytelling, knowledge-intensive organizations, virtual communities, and organizational archetypes, using interpretive and qualitative methods. He is a native speaker of Polish, and has near native-speaker fluency in English.
James has a significant track-record in advocating for the improvement of Wikipedia's health-related content. He is an active contributor to WikiProject Medicine and is the president of Wiki Project Med. Last October, the Signpost reported the publication of the first Wikipedia article as a peer-reviewed academic journal article, in Open Medicine ("Dengue fever: a Wikipedia clinical review"), for which James was first author. James is a Canadian hospital emergency physician, and is a clinical assistant professor at the Faculty of Medicine, University of British Columbia. He is a native English speaker.
Denny was the first administrator and bureaucrat on the Croatian Wikipedia. He studied at the University of Stuttgart in Germany, and gained his PhD in computer science and philosophy from the Karlsruhe Institute of Technology—both prestigious institutions. He joined Wikimedia Germany to launch the Wikidata project in 2012, and now works at Google. He counts himself as a double native speaker of Croatian and German, and speaks English at a professional standard.
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 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.
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.
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."
Discuss this story
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]