Talk:Wikimedia Foundation elections/2021/Post Analysis

Add topic
From Meta, a Wikimedia project coordination wiki
Latest comment: 2 years ago by Nosebagbear in topic Neutrality Banner


Does it make sense to merge this page into Wikimedia Foundation elections/2021/Post mortem? That's the page where the discussion of the election outcome seems to be occurring, and it is consistent with where the outcome of the election in previous years.

@JKoerner (WMF): You renamed the page - do you think it makes sense to undo that renaming, and centralize the discussions? TomDotGov (talk) 01:41, 9 September 2021 (UTC)Reply

Hey there, TomDotGov, Thanks for reaching out. The facilitation team decided to rename the page to Post-Analysis because "Post-Mortem" is rather morbid and doesn't translate well into some languages.
One reason to keep the pages separate is that one report will be from the facilitation team who supported the election (Post-Analysis) and the other can be exclusively for community feedback. Of course, the community feedback is informing the Post-Analysis. The Post-Analysis will include procedural pieces, like reflecting on the timeline, and planning activities sooner to meet the needs of communities and candidates. Best, JKoerner (WMF) (talk) 13:58, 9 September 2021 (UTC)Reply
Does that mean I should wipe my queries here. One is duped on the washup page, but I can remove the question-ones as well and add them to that? Nosebagbear (talk) 22:07, 9 September 2021 (UTC)Reply

Demographic profiling of commenters[edit]

@JKoerner (WMF): Please tell whoever wrote the second paragraph in this section, who is presumably also the person who wrote the area discussed here back in February, that they really have to stop doing this kind of thing. It is exceedingly unacceptable behaviour. --Yair rand (talk) 23:29, 21 October 2021 (UTC)Reply

Hey there, Yair rand, Thanks for the feedback. To learn from what we are doing it is important to document what happened. I am sorry you do not agree with the term used her. I do recall in that prior discussion inviting you to make a suggestion of a more comfortable term for you. That opportunity still exists. Best, JKoerner (WMF) (talk) 21:41, 1 November 2021 (UTC)Reply
The problem is not the terminology, as had been made clear. Attaching personal demographic information (especially private information) to comments is not okay. --Yair rand (talk) 21:52, 1 November 2021 (UTC)Reply
Thanks again, Yair rand, for your feedback! Best, JKoerner (WMF) (talk) 22:56, 1 November 2021 (UTC)Reply


There is a lot about this report that seems to reflect the opinions of the authors rather than being a neutral discussion of the election. A few examples:

  • "This concern was largely but not entirely voiced by white western men." Is there some reason why the concerns about "bias and privacy violation" are less valid when coming from particular genders and ethnicities? If not, why was this included?
  • "Seven of the 19 candidates answered those questions. Three of those seven were elected." "There is no strong correlation between answering all community questions and getting elected."
  • "Identifying and supporting women and non-binary candidates for the Board of Trustees." seems hard to justify when the result was as gender-equal as it's possible to get.
  • "Not all eligible voters had their votes counted in the election. The Elections Committee struck some valid votes. A quick review showed this was approximately seven of the 78 struck votes." This seems super important, and deserves a lot more discussion then it's being given here. At the same time, there doesn't seem to be any proof of it, and without evidence, we have a report that's sort of making the accusation of election tampering without supporting facts.

I haven't had a chance to go through the report in detail, but I think these sort of point to a non-neutral tone that isn't fitting for a report like this. I'd hope that these points can be addressed and the report revised. TomDotGov (talk) 23:30, 21 October 2021 (UTC)Reply

@DBarthel (WMF) These points haven't been addressed. Please do not unilaterally remove the neutrality banner until consensus has been reached that this main space page is neutral, as otherwise this report could misinform those who might rely on it for future elections. TomDotGov (talk) 15:01, 26 October 2021 (UTC)Reply
@TomDotGov Thank you for your input. The Movement Strategy & Governance team is sharing it's insights and point of view on the Board Elections in this report. Though every input is very welcome to clear up the text, this report is still the MSG team's point of view. These learnings and insights might differ from yours, and it is totally acceptable to criticize points on the report. We learn from any criticism, be assured. But as much as I am not supposed to edit your comments and statements directly, I would ask for the same respect for this text from the team. Due to this I kindly ask you to restore the page to the state at where you changed it. Thank you! DBarthel (WMF) (talk) 16:12, 26 October 2021 (UTC)Reply
@DBarthel (WMF) Thanks for responding here, but I believe you are mistaken. While this had been a discussion page in the past, [1] converted it to a main space page. Since this isn't a discussion page, it is subject to the same community improvement that any other page is. I respectfully ask that you address and correct these neutrality issues, and gain consensus that the report is written in an unbiased way. Alternatively, this could be moved to a subsection of the Post mortem, which is a page clearly governed by discussion rules that allow anyone to respond with corrections. TomDotGov (talk) 16:39, 26 October 2021 (UTC)Reply
Hi there again, TomDotGov. I can appreciate that your point-of-view might differ from what is reported in this report. That can feel uncomfortable. Unfortunately I cannot suggest we alter this report because your perspective differs from what is in this report. Your opinion on this report does not make the information any less valid or real. Some of the items you noted above were actually feedback from community members drawn from the discussion of the 2021 Board election. Again, I am sorry that you have concerns with the report, but unless there is broader criticism and documentation showing this report represents untrue or biased information, I am not sure why the banner should stay. Best, JKoerner (WMF) (talk) 01:11, 2 November 2021 (UTC)Reply
@JKoerner (WMF) This is a content page, and content is edited collaboratively. While it would be perfectly fine for me to edit this content page like any other content, I believe that a banner pointing to discussions like this is a better approach.
We have multiple people here that have pointed out problems with this content, and suggested ways to improve it. I believe my criticisms are reasonable, and so do other people, and so I ask that you address them, rather than making them hard to find. I also believe that the criticism from other editors are also reasonable and unaddressed, except by a promise by DBarthel to eventually address some.) TomDotGov (talk) 01:34, 2 November 2021 (UTC)Reply
Hi there, TomDotGov, Yes, DBarthel and I talked yesterday. We discussed the situation with the neutrality banner. We see three people commenting on the content of the report, and those comments cover different aspects of the report, and this is after it has been posted for over 10 days. We cannot consider this disagreement with the neutrality of the report widespread. This is why we removed the banner. Personally, I find it curious that the banner was placed without any discussion about the need for such banner.
Unfortunately, the Movement Strategy & Governance team cannot alter the information in the report because three people disagree with the content. I am truly sorry you find this to be an issue. To make the most of the response, I'll address your items point-by-point. I'm glad to link to the information if that would be helpful. Please let me know about this and I can make those links.
1. The argument about bias concerns was interesting coming from mostly white men from Western regions. This is interesting for several reasons. Pictures have been used previously in Board elections and to my knowledge there were not discussions about bias in prior election cycles regarding the use of pictures of candidates. It was also interesting that people from a majority social group (white men from Western regions) were expressing bias concerns. Discussion about pictures and videos should be considered before future election cycles. We also must consider what other community members from some regions have said about pictures and videos being important to them and their communities for connection and engagement purposes.
2. You quoted the report about candidates who answered all of the community questions and a portion of those candidates getting elected. Less than half of the candidates who answered community questions were elected (42% or 3:7). If you are suggesting the three candidates who answered community questions and were elected had a stronger advantage, I'm not sure how that could be proven given there are other factors that could have meant those candidates were elected. Largely, we heard community members did not read the candidates' answers to community questions (even the smaller group of 11 questions). Knowing this, we cannot see how answering all community questions had a strong impact on three candidates' election.
3. This is a statement developed out of community member comments on the Post-Mortem report.
4. This is a situation brought up by a community member and can be seen by reviewing the struck votes. I personally agree with you about finding out more. I'd like to personally know more about why votes are struck and if there is a process for appealing struck votes. I am sorry you feel the report is suggesting election tampering. I do not see any suggestion of election tampering in the report.
If you have any additional concerns beyond those I have addressed, please reach out. Best, JKoerner (WMF) (talk) 22:05, 2 November 2021 (UTC)Reply
@JKoerner (WMF) Thank you for replying on the content. I'll point out that this is a content page, and as part of metawiki, it's a page that anyone can edit. We don't have a WMF-only page protection level, and this page is not owned by the WMF or any subset of WMF employees or contractors. So even if the Movement Strategy & Governance team can't alter this page, all other editors can, and have the responsibility to. It may be that this should be a PDF on commons, rather than a wiki page.
I'd also say that while interest in this seems to have waned, three people pointing out flaws is still more than the two that have supported the report as it stands, especially when those two people have multiple conflicts of interest. (Being on the same team and being the subjects of the report are obvious conflicts of interest.) So far, there hasn't been anyone that isn't the subject of the report that believe's it's unbiased.
When we have such conflicts, we need to either make sure the report is unrebuttably neutral, which is the usual case, as most reports are written in a neutral manner. If we can't get to that level by revising the report, then the right thing to do is to use the fact that this is a content page that anyone can edit to indicate that the content is not neutral. What we can't do is leave it here, so that in a year people refer back to the report but not the concerns.
1. You're once again choosing a particular axis to characterize people over, in a way that's basically unacceptable. Those same people are almost certainly part of multiple minority groups, and so choosing one particular demographic groups to emphasize smacks of bias that shouldn't be in a neutral report.
2. Nosebagbear did the math, so I won't repeat it. But by your logic, the highest percentage that anyone could get is 57%, or 4:7. The last candidate to be eliminated also answered many questions, so if we had one more slot open it would have been 4/5 candidates than answered the questions. I believe this to be significant - but more importantly, you're inverting the burden here. The report asserts - "There is no strong correlation between answering all community questions and getting elected.", and that doesn't seem to be supported in any way, since there really does seem to be a correlation.
3. It appears about once, and the actual conversation is more balanced then this report makes it appears.
4. Do you have more information about this? I think I've been following this rather closely, and haven't seen it at all, but I may have missed it. I don't know what to call not counting valid votes other than election tampering, especially if it's done systematically.
TomDotGov (talk) 22:57, 2 November 2021 (UTC)Reply

Unconsidered areas, improper stress, failure to comply with unclear mechanism for ensuring vetting prior to Board review[edit]


@DBarthel (WMF): I have several concerns with this I'd like to flag up. There's quite a few points, so I've endeavoured to tag them with tl;dr titles at the front of each. Especially the first has a lot of detail and various sub-questions, so please let me know if it needs clarifying.

Questions, comms, and evidence Firstly, this is what this report says about the questions Each Board election the community has the opportunity to submit questions for candidates to answer. This usually lines up with the Call for Candidates. This year, the MS&G facilitators supported the Elections Committee. The recommendation to the Elections Committee was to select about ten questions. This would allow for candidates to answer questions and have translations provided. We believe this provides a more equitable opportunity for candidates. Some community members posted the entire list of 61 questions for candidates to answer. Seven of the 19 candidates answered those questions. Three of those seven were elected. Four candidates who were not elected answered the entire list of questions. They were eliminated in calculation rounds 6, 8, 11 and 14. There is no strong correlation between answering all community questions and getting elected.

So. Not a thing about the fact that this recommendation was offered without an open statement of such. Not a thing that it was stated that the questions would only be "collated", not actually selected at all. Not a thing that this exact issue arose in 2017. Not a thing about how those questions were selected. Not a thing about the months long delay involved in ever responding to concerns about non-communication. Not a thing that an WMF facilitator was apparently unable to ensure that a response, either from ElectCom, or the strategy team, could be acquired - to the tune that even she thought our unhappiness and frustration was warranted. Given that the reason (given, literally, months late) was that it wasn't purely ElectCom's fault, but on the strategy side too, I'd have hoped that the only group in a position to have all the answers didn't cover it is...disturbing. The most mentioned concern on the washup suggests that the facilitators should also include it, especially with the unique access to information on the issue they would have. I know the exact dates that Electcom were notified each time. Where are their internal notes saying they couldn't respond without discussion external to that committee? What are the dates of their messages to the relevant members? When did those members respond?

I held off from raising this as a topic in the most recent Board Talk (CAC) because I'd hoped we'd get more detail, and you stated that a share of the fault lay external to ElectCom due to a need for multi-party communication. But we have no details on that, and even if the entire issue were to be aggregated, the multiple problems (questions, communications, clarity of comms, responsibility external to Electcom, lack of oversight of Electcom) aren't covered at all!

Maths I'd also like the correlative relation "r number" given for the full questions/partial questions answering - a quick loading of SPSS suggests that actually there is a correlation - this feels like another maths argument with the WMF where the working needs to be shown.

Support clarification I, too, would like to ask about Identifying and supporting women and non-binary candidates for the Board of Trustees - can it be clarified how this was not met in the most recent election?

Meeting communications report expectations As a workshop member I firmly approved of the findings of the Movement communications insights/Report. But Staff across departments are considering regional diversity beyond those areas identified above. They will be presenting a more thorough proposal later in the fiscal year. I found it somewhat concerning that a proposal for the structure of the Board is intended to reach a "thorough proposal" level before being presented more broadly. Per this page, this type of thing should not be progressing so far before being examined. Any assistance regarding communication @ELappen (WMF): can provide to both/all parties would be fantastic.

Presentation to the Board I assume that this report is being presented to the Board (which is quite right!), but as it excludes so much of what is in the post-mortem, I'd like to ask how that is being fairly presented to get sufficient "eyeball time". As the community's own position, it is of course at a minimum, in need of equal attention by the entire board, rather than one or two interested souls. Is it being collated and presented alongside? Or if some of us write up a summary, will it be presented alongside? When is the timeline of being sent out to trustees or being presented?

In the event that this is already attempting to be a collation of that post-mortem (plus discussions with trustees, candidates etc), I'd be even more concerned by the level of excluded detail, but also why it didn't go through a review stage prior to this general publishing (as was done so well with the movement communications insights report) Nosebagbear (talk) 16:51, 26 October 2021 (UTC)Reply

@Nosebagbear: - Thank you very much for pinging - as you raise rather complex issues, I would like to bring these topics to the team and respond to you next week. I know that this is somewhat unsatisfying for now, but I guess this is more than I can answer right away. Please bear with me. Best, DBarthel (WMF) (talk) 11:54, 27 October 2021 (UTC)Reply

Hey there, Nosebagbear! Thanks for writing up such thorough feedback. Regarding not including such explicit detail about the community questions, that was a decision we made because it was already well-documented on Meta-wiki. The goal was to get more folks to read the report. If it's too long, it won't get read or translated. Would it be more favorable if I linked to some of those pages with the discussions about the community questions?
Regarding your calculations in SPSS, please share. I did some calculations and found no strong correlation. I'd like to see if I have miscalculated.
About supporting women and non-binary candidates, that was feedback from the community. You can view it in the Post-Mortem. Regional diversity was also a discussion point after the Board election results announcement. Community members were discussing on Telegram about regional diversity and their disappointment at the lack of regional diversity in the election results. This report is stating that this will be something to consider. "Some community members suggested designated regional seats. A brief (~4 week) feedback period before the next Board election can discuss ideas to improve regional diversity. Underrepresented regions are key stakeholders in this process," which can be seen in the report here.
I believe the last piece is the detail you see as missing from this report. I mentioned above that we had to balance between a length to include content and a length to make sure people read it and it was able to be translated. If there is anything in particular you think that is not addressed, perhaps it can be linked in the report as I suggested above?
Let me know if I have missed anything. Best, JKoerner (WMF) (talk) 01:34, 2 November 2021 (UTC)Reply
@JKoerner (WMF): I have to deny that it's well documented at all. We still have absolutely zip idea on why the issue of not stating that selection would happen was allowed to occur, let alone re-occur.
Quim did, in his selfless fashion, reply The task of producing these questions was distributed, with some bits done by the facilitation team and some by the EC and this is why producing the answers needs the collaboration between both teams. I'm sorry that this is taking so long, and I am happy to take the blame. We were very busy working on other aspects of the election and the EC alone wasn't in a good position to respond because, as said, my team has been involved as well [...] But we never received any further information on why this co-operation was so blisteringly inefficient. Especially since Quim is pretty quick about replying to my queries. 2 weeks, maybe, 11? Nope, not buying it. Where did those delays come from? Especially since much of that was before the actual during-vote responsibilities arose. The (jointish) statement didn't provide any answer to either of these aspects. Nor did it provide any answer to the poor level of oversight of Electcom.
That there are a lot of pages of myself and others asking about these issues doesn't mean it's well just highlights how lacking in answers we are. If this analysis won't provide those answers, then you need to clarify how on Earth we are supposed to get them, given that ElectCom won't tell me themselves. I'm looking for a tool less hammer-like than needing to call a meta-RfC vote of no-confidence to disband and reform ElectCom to make it more responsive to the community.
About supporting women and non-binary candidates, that was feedback from the community. You can view it in the Post-Mortem. really confused me...if this is the basis for inclusion then that completely contradicts the whole not including the bits about the questions.
I'm kind of confused by the second part of your third paragraph. It doesn't seem to directly address the "don't propose formed proposals prior to consultation" query. I'm struggling to either rebut or agree, since I not actually sure what it's saying as distinct from the current report.
While I can see an argument for some detail being excluded from the actual text, I still can't see answers as to how this post-analysis is being presented to the board, and how the post-mortem contents (the presence of which is being used as grounds to both include some facets and exclude others) is being made equally evident to the Board. If that detail need not be contained in the actual report to be translated, it could certainly be included here. I will transcribe to the post-mortem as appropriate. Nosebagbear (talk) 12:41, 2 November 2021 (UTC)Reply

Neutrality Banner[edit]

I've noticed multiple WMF-related users have been removing the neutrality banner from this page, while there are still multiple open concerns with the contents. I'll put the question to @JKoerner (WMF), @DBarthel (WMF), and anyone else that's been interested - what's the right way to indicate that there has never been consensus that the report is neutral and accurate? TomDotGov (talk) 22:07, 1 November 2021 (UTC)Reply

Hi there, TomDotGov! Thanks for reaching out about this. I removed the banner for similar reasons as DBarthel (WMF). You are the only person who has expressed such neutrality concerns and posted the banner. It seems as though this is not a wide-spread opinion. I have removed the banner and noted as such. Best, JKoerner (WMF) (talk) 22:55, 1 November 2021 (UTC)Reply
I think TomDotGov is correct on this point. Concerns have been raised about the content, and have not yet been addressed. Some indication of dispute seems reasonable. --Yair rand (talk) 23:48, 1 November 2021 (UTC)Reply
Hi again, @JKoerner (WMF). I realize it's difficult to have your work product criticized like this, as it has been in the past. At the same time, one of the movement strategy goals is to Manage Internal Knowledge, and that means being sure reports like these are neutral and accurate.
Seeing as how there's already a meeting that will address the numerous errors in the report, I don't think it makes sense to leave an inaccurate page up without any warnings. Once that's finish, we can discuss if the changes made resolve the issues. TomDotGov (talk) 23:53, 1 November 2021 (UTC)Reply
Hi there, TomDotGov, Please see my prior contributions to you above in response to this. Also, I am confused. You say, "Seeing as how there's already a meeting that will address the numerous errors in the report..." What meeting might that be? Best, JKoerner (WMF) (talk) 01:13, 2 November 2021 (UTC)Reply
@JKoerner (WMF), see DBarthel (WMF) above, "I would like to bring these topics to the team and respond to you next week." I was assuming that was going to be some sort of meeting, though it's possible there is some other method of it being brought to the team that I am not aware of. TomDotGov (talk) 01:37, 2 November 2021 (UTC)Reply
Hi again, TomDotGov, Yes, DBarthel (WMF) and I spoke about this yesterday. Best, JKoerner (WMF) (talk) 20:08, 2 November 2021 (UTC)Reply
  • Were we to compare this to an article, I'd currently say this report has significant "absence neutrality issues - which is mainly where absences are causing neutrality faults, even if unintentional". Beyond that, there are a few specific aspects that are odd, and I'm not sure how to classify them. Examples include that some areas get included because they're raised in the post-mortem, rather than actually being able to cite a clear objective evidentiary base, while others are excluded because they're raised in the post-mortem. Nosebagbear (talk) 19:40, 3 November 2021 (UTC)Reply