Wikipedia:Requests for comment/Central Notices

Recently the Foundation and Research Committee launched a banner using the Central Notice system in collaboration with Harvard University Berkman Center. This met with both positive and negative comment in various noticeboards: One of the criticisms was the lack of discussion/input the English Wikipedia community was granted - and the lack of notification prior to the launch. Feedback on the mailing lists seems to suggest that the Foundation and various committees are not aware of the communities preferences regarding Central Notices, partly because it has not been discussed before. This RFC is intended as a discussion to help develop a guideline or policy, to be posted at WP:Central Notices. Please add new sections for new ideas, where appropriate; I have added some starter sections below.
 * Administrators%27_noticeboard/Incidents
 * Village_pump_(technical)

Fundraising banners
Fundraising banners are largely accepted by the community (with some dissension over their size/shape). So are we happy to allow the Foundation relatively free reign over the fundraising CN's? --Errant (chat!) 10:47, 10 December 2011 (UTC)
 * Those god-awful fundraising banners a few years ago argue otherwise. MER-C 03:28, 11 December 2011 (UTC)
 * Second what MER-C said, the infamous "Wikipedia Forever" banners were objected to, quite strongly, by the community at the time. I think we might say that the community tolerates such banners as a necessary evil, so long as there is a certain standard of professionalism and good taste displayed.  Lankiveil (speak to me) 03:51, 11 December 2011 (UTC).
 * Some years ago I unilaterally disabled the fundraising banner on enwiki because of how bad it was. They haven't repeated the javascript abomination since then, but I still don't entirely trust them. --Carnildo (talk) 06:50, 11 December 2011 (UTC)
 * While I dislike the idea, these banners help supply the Foundation with the funds necessary to keep English Wikipedia up and running. For us to refuse these banners would be, in my opinion, "shooting oursleves in the foot". עוד מישהו Od Mishehu 08:20, 12 December 2011 (UTC)

Non-fundraising banners
What sort of non-fundraising banners would be appropriate/accepted? --Errant (chat!) 10:47, 10 December 2011 (UTC)
 * Banners which are notices related in the internal functioning of the English Wikipedia, or Wikipedia as a whole, may be permitted. Examples include ArbCom election notices, request for community input on discussions, or acute needs for editing assistance, such as in the Bell Pottinger incident. (Examples are notational and do not constitute a list of acceptable notices). Gerardw (talk)
 * Why would we use centralnotice for that and not Sitenotice? The Arbcom is English Wikipedia only.  All of the things that you suggest are, in fact.   Philippe Beaudette, Wikimedia Foundation (talk) 06:06, 11 December 2011 (UTC)


 * I agree on banners related to Wikipedia's internal functioning, such as board member elections or site-wide polls on Wikipedia's future development (eg. the image filter). I'd further accept banners relating to external research of Wikipedia.  Using the CentralNotice function to recruit for research unrelated to Wikipedia is unacceptable in my opinion, as is any other form of advertising.  In all cases, banners should be used sparingly: I use Wikipedia on over a dozen different computers, and playing whack-a-banner gets tiring. (Note: "Wikipedia" should be read as "Wikipedia and all other projects operated by the Foundation") --Carnildo (talk) 06:57, 11 December 2011 (UTC)
 * Ads is ads. Academia is no less competitive than Coke vs Pepsi. Whichever research group gets enough data to justify a thesis gets to publish. Come in second and you're not publishing, or publishing in a less prestigious journal. If WMF decides they need to run ads pay the electric bill, that's their decision. But I'd rather see a good honest Budweiser ad appearing next to hops then WMF running ads and pretending they're not ads, or they're not picking winners and losers in granting research groups access to Wikipedia users. Gerardw (talk) 20:23, 11 December 2011 (UTC)
 * I somewhat disagree: in general, Academia expands the boundaries of knowledge of humankind, and its objectives are close with what Wikipedia does. However, I think that one important requirement to any research advertised on Wikipedia could be added: all raw data should be published without delay under CC-BY-SA license, so that any scholar could use it to do their research. Pundit | utter  00:01, 16 December 2011 (UTC)
 * This is something that we've discussed at some length on RCom and I don't think we are a huge way from you. But two particular points where we differ from your position, Firstly raw data needs anonymisation before posting, even if no one question on its own would identify an editor sometimes a combination would. Secondly we do tend to support the idea that the researcher shouldn't have to publish their data before their research report, otherwise they face the risk of being overtaken.  Ϣere Spiel  Chequers  01:24, 18 December 2011 (UTC)
 * Anonymization is an obvious necessity, I agree. With the risk of being overtaken... well, access to Wikimedia banners is a great privilege. Many researchers don't have it and won't be able to get it. The ones who receive this valuable resource (and can, for instance, plan the questions asked, the research design, etc.), should not be additionally privileged by being the only ones who have access to the data, when they process it. After all, one could imagine a situation in which a researcher is just lazy and does not release their research report for a while. Also, the researcher has an advantage at the start as well, since they are the ones who anonymize the data, etc. The risk of being "overtaken" is a one that highly privileged researchers should face (btw, it would be valuable to additionally take care and make efforts to make sure that the researchers from less advantaged and less privileged research communities, mainly non-Western ones, are not "overtaken" in terms of research privilege access, as they may be now... I'm writing this as a definitely privileged scholar). Pundit | utter  03:03, 19 December 2011 (UTC)
 * Yes we should be careful about privileging some researchers over others. If we are going to allow third party researchers to ask questions of Wikimedians using our site then we need to create a level laying field, and I would hope we can all agree that we shouldn't let anyone sit on data indefinitely. But if someone comes up with the idea for a novel question then I can see the argument for giving them a temporary exclusivity on the data, and the sentiment on RCom was very much in favour of giving the researcher who thinks of a new angle of research the opportunity to publish first. One risk of doing otherwise is that people might cut corners to publish first. What would you think of this compromise?  Ϣere Spiel  Chequers  21:36, 22 December 2011 (UTC)
 * Apologies for my late reply, I was in the offline world for over a week. I think that 6 months is a fair compromise and gives adequate time to write up the research results. My minor concerns/questions:
 * researchers have to release their resulting research under an open access Gold basis - if read widely, this would require them to refrain from publishing in top journals (usually non-open access),
 * "non personal raw data" seem to be retained by the researcher and other researchers would not gain access to them (I understand that some authorization should be required, but still it would make sense to allow researchers to apply for raw data access instead of applying for a new research poll),
 * judging by the standards in the research which caused so much fuss recently, Wikipedic nicknames are not considered to be identifying. Yet, I can easily imagine that for many Wikiholics linking their nickname with the research results is more than identifying. Therefore I would suggest obligatory anonymization of nicknames in all research, unless the research project has a justifiable reason to keep the nicknames, and clears out with some sort of Wiki research ethics committee. Pundit | utter  18:45, 28 December 2011 (UTC)


 * I believe that banners for logged-in users promoting interests related to the general Wikimedia community are acceptable, like the POTY contest - one of the main Wikimedia-wide events or any large discussion, like a license change. theMONO 21:59, 28 December 2011 (UTC)

Link policy
No banner ad should contain a hyperlink to anywhere off Wikipedia, specifically English Wikipedia, meta, and commons. Gerardw (talk) 12:10, 10 December 2011 (UTC)
 * And wikimediafoundation.org, for one. Is this an actual policy? If so, where is it documented; if not, what would be the reason for instituting it? 67.6.163.68 (talk) 00:30, 11 December 2011 (UTC)
 * And strategy wiki, and outreach wiki, and... the list goes on and on. If we must do something like this, I would strongly prefer "a Wikimedia project site", but I think even that is tying our hands.  Philippe Beaudette, Wikimedia Foundation (talk) 06:07, 11 December 2011 (UTC)
 * Why would it tie your hands? My financial institution does the equivalent. User clicks on the banner and goes to a WMF hosted web page with the external link that says:
 * Where the link goes.
 * Purpose of the link -- why is WMF providing this.
 * Privacy, terms of usage.
 * Relationship between WMF and the external host.
 * ? Other stuff I'm not thinking of right now. Gerardw (talk) 20:19, 11 December 2011 (UTC)
 * This is a very good idea. --Errant (chat!) 23:09, 11 December 2011 (UTC)
 * I'm not necessarily opposed to the idea, but I have some doubts your bank really does the equivalent but obviously can't check myself without knowing the bank. Most likely they normally warn you when you are visiting an external site as do a number of sites, but in no way guarantee they will always warn you, and a free to make agreements with other organisations where they explicitly decide they don't warn you, or simply make mistakes. In fact, in the specific case of banks and similar, from my experience quite a lot of them embed Google Maps or perhaps Bing Maps or similar external sites and give no warning of this at all. Nil Einne (talk) 18:23, 18 December 2011 (UTC)
 * I've never observed it with USAA. Nobody Ent (Gerardw) 22:02, 18 December 2011 (UTC)
 * Linking out of the Project sites is okay with me, but it should be immediately obvious that a banner is doing so -- the Harvard/Science Po banner's javascript-based form submission is exactly the wrong thing to do. --Carnildo (talk) 07:04, 11 December 2011 (UTC)
 * Gerardw - a nice suggestion. Carnildo - how would you make it obvious on a banner itself? – SJ +  21:50, 11 December 2011 (UTC)
 * At a bare minimum, moving the mouse over a hotspot should show the target URL in the browser's status bar -- the way the Harvard/Science Po banner was constructed, it showed nothing. --Carnildo (talk) 22:34, 11 December 2011 (UTC)
 * Naturally, these are understandable concerns, even though the Harvard center clearly is not getting much additional publicity through Wikipedia, nor does it purposefully mislead the user about the address. Some rules can be established for the future, even if most of us agree that this particular research was totally ok to be advertised. Pundit | utter  00:04, 16 December 2011 (UTC)
 * Any Wikimedia site should be legitimate, along with some possible exceptions for strongly affiliated other things. Toolserver should probably be considered OK as well. theMONO

Community input
To what extent should the community be given input into banners that are shown on English Wikipedia? At the moment they are scheduled on Meta - should that calendar be mirrored here on-wiki? Should the community give explicit approval for banners via consensus gathering? --Errant (chat!) 10:47, 10 December 2011 (UTC)

The fundamental question is who owns these banners. Choyoołʼįįhí:Seb az86556 > haneʼ 11:08, 10 December 2011 (UTC)
 * The banners posted on CentralNotice are controlled by the Wikimedia Foundation, not here on the English Wikipedia. This allows for large-scale announcements, such as the fundraising messages, on every page across all Wikimedia wikis. No admin here on the English Wikipedia can edit or modify them unless they also have admin access on the central Meta-Wiki site in their own right. Zzyzx11 (talk) 17:22, 10 December 2011 (UTC)
 * No, they are controlled by meta admins. Also, we can turn them off if we like: Village_pump_(technical) --Kim Bruning (talk) 06:14, 11 December 2011 (UTC)


 * No objection whatsoever to fundraising banners. Absolutely nothing else should appear on the same basis as fundraising banners, as such notices would make the fundarising banners less prominent and thus less effective. —WFC— 11:42, 10 December 2011 (UTC)
 * The WMF controls their websites. They can do whatever they want; they can use whatever banners they please. While I dislike banners, it's not my place to argue how they should be used. The WMF has made it very clear that they can and will ultimately override community consensus (e.g., autoconfirmed-only article creation trial) so there is no point in wasting the community's time with another RfC that the WMF is free to ignore. If someone opposes banners, and they think the WMF is wrong, they can end their involvement on the WMF's sites. I no longer care what the WMF does, because I can't actually do anything about it other than complain. / ƒETCH COMMS  /  18:39, 10 December 2011 (UTC)
 * The WMF exists primarily to support the projects, which in practice means following community consensus. The proposal to changing article creation permissions, as I understand it, ran into disagreement between two community groups - the group of devs and the group of en:wp editors.  More devs than editors have contracted or worked for the WMF at some point, but their views are their own, and there was strong opposition to the idea among devs to counter strong support among editors.  I'm not sure of the right way to resolve that kind of split -- even the attempts to have a single conversation to reach new consensus failed at the level of choosing between en:wp and mwiki as the canonical wiki for discussion.
 * This sort of disagreement is a concern, and an important one to work out, but does not support the point you claim above. There is simply no mechanism for the WMF to decide how to handle disagreements between community subgroups. (This is one area where a community-elected Project Council - that perennial idea - might actually be of use.) – SJ +  21:50, 11 December 2011 (UTC)
 * Hmm, well. I've not really been paying attention to that dispute - my interests here don't often put me near the new articles issues. But from that description if it is dev vs. community I'd possibly suggest that the community probably knows what is best for a community process than the devs :) The dev group is smaller, and not as active on-wiki (some not at all), so it would seem reasonable to argue they are more out of touch with the issues the community wants to address. Just thinking aloud; I am not aware of what arguments the devs have put forward :) --Errant (chat!) 23:18, 11 December 2011 (UTC)
 * When I have asked for clarification in the past, the "ruling" from WMF was that banner contents (with the exception of fund raiser banners) was governed by community policy. If we fail to agree a policy based on our consensus, well that's our fault for being so disorganized. --Fæ (talk) 20:25, 10 December 2011 (UTC)
 * Sounds right to me. I'd say that is a recognition of how the system works and should work, not a "ruling". (The WMF has no mechanism for passing such rulings.) Within the bounds of existing policy, everyone assumes the right to be bold... we should be open to new things without endless bureauracy. The Research Committee is dedicated to collaboration with outside researchers -- who otherwise have a hard time figuring out how and where to communicate on the projects.  It seems that communication could have been better here!  I bet they would welcome suggestions (including the reasoning of anyone who feels en:wp should not collaborate with research groups). – SJ +  21:50, 11 December 2011 (UTC)
 * Wikipedia controls wikipedia. The WMF was founded by wikipedia contributors, not the other way around. If we don't like the WMF, we can fork if we have to. Everyone would prefer that doesn't happen, of course :-) . --Kim Bruning (talk) 21:14, 10 December 2011 (UTC)
 * Fae, that's what they say, but given that they are free to override community policy, I would not give much weight to their word. After all, the community deals with articles based on consensus (we don't ever see the WMF creating articles now, do we?) yet the WMF still overrides consensus we've reached regarding the creation of articles. Kim, I think we both agree that forking Wikipedia would be silly at this point, but the other alternatives are to either ignore the WMF and just write articles, or just leave altogether. But the WMF, although founded by Wikipedians, now governs them. We can't overthrow our owners. / ƒETCH COMMS  /  04:15, 11 December 2011 (UTC)
 * No they're not free to override community policy. Foundation employees need to follow policy and consensus, and explain themselves, just like everyone else, else they'll find themselves blocked. What are you on about? And no I do not agree with you that forking is impossible, what's this? The right to fork is a defining property of a free content project. Can't fork? It's not free content. Next you'll be telling me that wikipedia is not an encyclopedia. :-P --Kim Bruning (talk) 05:41, 11 December 2011 (UTC)
 * Kim is correct. The point of preserving the Right to Fork is to ensure that governance != ownership.  Also, the use of "us vs. them" language isn't helpful here -- most WMF staff are also Wikipedians. – SJ +  21:50, 11 December 2011 (UTC)
 * Us vs. them is very much how the site is run. The regular Wikipedians do not control the servers or money. We cannot tell the WMF to stop wasting funds on frivolous things like WikiLove. Governance does not equal ownership, but it does equal control of resources. I can choose to work with the WMF, inside the system; I can also choose to work without it (by forking, for example) but I'll be responsible for financially and technically supporting anything I do. / ƒETCH COMMS  /  03:03, 12 December 2011 (UTC)
 * Kim, given that the WMF is in financial and technical control of Wikipedia, I'd say they are free to override community consensus. As I mentioned earlier, they have already done so in the past, recently with the autoconfirmed-user-only article creation trial. We can surely block WMF accounts but we cannot block them from doing what they want with Wikipedia. If they wanted to turn off the power to Wikipedia servers, what would we do, break into the data center and turn them on again? Re what I said about forking, I don't mean it's not physically possible, but it would be silly&mdash;unless the WMF decided to make the default skin purple and pink or charge money or something, a fork would hardly be able to do anything but duplicate content and struggle to get funding. We can fork, but that would be a rather fruitless effort (remember the es.wp fork? neither do I). / ƒETCH COMMS  /  20:02, 11 December 2011 (UTC)
 * Kim, given that the WMF is in financial and technical control of Wikipedia, I'd say they are free to override community consensus. As I mentioned earlier, they have already done so in the past, recently with the autoconfirmed-user-only article creation trial. We can surely block WMF accounts but we cannot block them from doing what they want with Wikipedia. If they wanted to turn off the power to Wikipedia servers, what would we do, break into the data center and turn them on again? Re what I said about forking, I don't mean it's not physically possible, but it would be silly&mdash;unless the WMF decided to make the default skin purple and pink or charge money or something, a fork would hardly be able to do anything but duplicate content and struggle to get funding. We can fork, but that would be a rather fruitless effort (remember the es.wp fork? neither do I). / ƒETCH COMMS  /  20:02, 11 December 2011 (UTC)


 * I don't care whether it's approved ahead of time or not, but every banner should have a link to a discussion page so concerns can be addressed, rather than having people scouring VP and meta and everything looking for clues. This should be obvious, especially if we're going to start having third-party banners. —Designate (talk) 22:44, 10 December 2011 (UTC)
 * Agreed. This was actually a recommendation for this banner that -for some reason- wasn't implemented. --Kim Bruning (talk) 01:28, 11 December 2011 (UTC)


 * It is my understanding that the widespread distribution of the Berkmann banners was due to a software bug, and I don't think it's appropriate to respond to this RFC as if it were intentional. 67.6.163.68 (talk) 00:27, 11 December 2011 (UTC)


 * The Wikimedia Foundation is running the servers and we are doing the content. We don't want them to interfere with the content beyond what is necessary for legal reasons or to make sure Wikipedia remains an open encyclopedia. And we should not interfere with the operation of the servers. That said, communication is always good, and if it comes before any new features are introduced, then it prevents misunderstandings. Hans Adler 01:42, 11 December 2011 (UTC)
 * Obviously they can do whatever they want; thus far they seem to want the communities to be self managing. Coming to consensus on our concerns is part of that process. Gerardw (talk) 02:14, 11 December 2011 (UTC)
 * Eh? No way they can do whatever they want. You can't run an organization like that! --Kim Bruning (talk) 05:39, 11 December 2011 (UTC)


 * This discussion is a waste of time. Of course most people don't like anything that can be interpreted as promotional, and a banner which includes off-wiki links and logos is obviously going to create concern, and no doubt the WMF will be evaluating. The community should focus on hammering the WMF about important stuff like the current ANI discussion showing that a new user apparently can send any number of abusive emails to harass editors (ANI discussion, ). Johnuniq (talk) 03:23, 11 December 2011 (UTC)
 * Thank you, that's entirely correct. If the WMF is going to run banners, nothing we can do, so let's try and address the problems we can do something about. / ƒETCH COMMS  /  04:15, 11 December 2011 (UTC)
 * Eh? We can do plenty. What are you talking about? --Kim Bruning (talk) 05:37, 11 December 2011 (UTC)
 * The fact the banner is currently not running tends to give lie to that theory Fetchcomms :) --Errant (chat!) 11:26, 11 December 2011 (UTC)
 * I agree, sorry to sound a bit wet or naive, but polite persistence does get listened to by the WMF. Communications may not be speedy or easy, but the end result was that the banner was taken down reasonably quickly by a meta admin and stayed down. I think that speaks volumes about how the WMF is reluctant to override a positive community consensus if it is expressed credibly enough. I have been personally uncomfortable with how parts of our discussions have veered to the pointy, ranty and griefy, I believe we make better progress when we can avoid muddying the water with tangential arguments. --Fæ (talk) 21:56, 11 December 2011 (UTC)
 * Agreed. My intention here was to start communication from en.wiki outward, and hopefully this will then be returned by the foundation/rcom. Indeed; all the noises so far seem to be "well tell us what you want". OK, so if the decision is "no research banners" then there may be an impasse for us to solve, but at least there is dialogue. And I doubt we will reach such a scenario :) It's frustrating that the lines of communication are so poor. One of the reasons we are still stuck on the slow-as-mud "everything must have explicit consensus" situation is precisely because of the tendency for outside groups to start something without telling anyone. It's a defence mechanism to help the community keep its autonomy. but in my experience; once you get people actually talking, the kinks iron themselves out. If slowly. --Errant (chat!) 23:14, 11 December 2011 (UTC)
 * Fae, perhaps we could also find a way to begin the autoconfirmed-user-only article creation trial, too? There was plenty more consensus for that than there was for removing some banners. And people were even getting paid to click on the banners. This entire situation is illogical. / ƒETCH COMMS  /  03:03, 12 December 2011 (UTC)
 * Before this discussion gets stuck on what the WMF vs. the community can or can not do, the outcome of WP:ACTRIAL is documented  here.  Anyone familiar with  the WMF infrastructure and  active Wikipedians can draw their conclusions. Kudpung กุดผึ้ง (talk) 07:42, 12 December 2011 (UTC)


 * Fundraising banners only; any other would require a wide consensus through an appropriate discussion (RfC, or whatever). That should be the way things work - for anyone wanting to make a significant change that affects a great many pages. Unfortunately, WMF have failed to take the appropriate steps to obtain consensus; therefore, if we need to make this explicit, so be it. Anyone - WMF, or any other member of the community - is quite free to make a suggestion, such as a specific banner - rather like Jimbo is doing, right now, about SOPA. In terms of decisions like this, WMF-proposed-ideas should be subject to the same community consideration as ideas from other users. If I had added a new 'advert-like' banner (theorizing I could), without appropriate consensus, then I imagine I'd have been reprimanded (warned, blocked, pitchforks, etc).  Chzz  ► 07:01, 13 December 2011 (UTC)
 * Input is helpful in many cases to make the most common - and the least expected - concerns apparent. But it should not be controlling. If there's a genuine negative reaction that brings a lot of complaints, obviously it would be wise to listen and adjust appropriately. CarolMooreDC 06:30, 15 December 2011 (UTC)


 * Personally, I don't object to appropriate banners if they help to fund the Wikimedia Foundation or to achieve something free, open source, collaborative, and/or academic. However, I feel the banners we use at the moment are too large and should be reduced in height by about a third.— S Marshall  T/C 11:05, 17 December 2011 (UTC)

Rcom comment
See:Administrators'_noticeboard/Incidents --Kim Bruning (talk) 23:54, 11 December 2011 (UTC)
 * Now archived at Administrators'_noticeboard/IncidentArchive731. --Fæ (talk) 16:55, 13 December 2011 (UTC)

Berkman Sciences Po banner taken down
Moved section - the comments within this section, through to the one from ThurnerRupert at 13:57, 17 December 2011, were originally posted on the admin noticeboard, at [[Wikipedia:Administrators' noticeboard/Incidents# (this specific version). Because this RfC is open, I've removed the thread from ANI and pasted it below. [[User:Chzz| Chzz ]] ► 15:20, 17 December 2011 (UTC)

Hi folks! :) I come back to you as regards our Berkman Sciences Po research banner. As mentioned in the previous community discussion here on AN/I, the banner was taken down by Brion and Beria within hours after its launch on December 8th, so that we were unable to reach our target number of participants in this experiment. I have absolutely no doubt that this decision was taken in good faith, based on the assumption that the banner was malfunctioning and running at 100% on English Wikipedia. However, this was not the case, as we actually took this research project as an opportunity to work together with WMF towards developing a new banner code that can selectively display banners to specific subsets of editors based on their user metrics, which is a significant improvement in the flexibility of the banner tools now at the disposal of WMF and the community. Our hope was that this new feature would help reduce the general banner overload for this and future campaigns.

Now that the banner is down, we've been paying great attention to the concerns raised here by the community, notably as regards its ad-like design. I'd like to reiterate that the first banner proposals that we made to WMF (there were 13 of them in total) did not feature our logos. We included them in order to comply with one of WMF's request, with people (in my view somewhat legitimately) being concerned about the fact that if this banner was to redirect people to a third party website for the purpose of performing an experiment, it should be made as clear as possible from the very beginning that this project was not run by WMF. We were acting in good faith, and remained 100% committed to respecting the community throughout the 18 months process that eventually led to our research project being launched on enwiki.

We would like to resume this campaign for 24 hours maxi in order to collect the 650 additional responses that we need for our research. So we are now working to address the comments made here by the community before the study goes live again. First, we've been setting up a FAQ about this research on the research page of this project on meta. Second, we will remove our logos from the banner and include a mention that this study is run by the Berkman Center and Sciences Po in plain text at the bottom. This mention will be clickable and redirect to our research project page, so that participants know were to find information about this study. Third, we will introduce a intermediary step between the moment when a user clicks on the banner and the actual redirection to our website, in order to make it very clear, transparent and opt-in that some (otherwise publicly available informations) are passed to us along the way. Those informations are: (i) Wikipedia username (ii) edit count (iii) account registration date and (iv) user privileges. We use these informations to perform a participant eligibility check before login users in the study. If this sounds like a plan, we will post the new banner design here as soon as it's ready in order to get community feedback before the study goes live again. I hope that this will help address the community's concerns and allow us to move forward with the research. SalimJah (talk) 18:34, 16 December 2011 (UTC)
 * That's great - thanks for taking the concerns raised in good faith and actively coming up with a solution to them! Particularly the passing information across thing. :) I for one have no concerns, now, in supporting enabling the banner for a 24h period so you can finish the research. --Errant (chat!) 19:44, 16 December 2011 (UTC)
 * When do the results come in? I admit that I did the study "smartly" in some cases, but dumbly in others.  I want to see the results to see how normal or outlying I was. Hasteur (talk) 20:06, 16 December 2011 (UTC)


 * 1. Will the data collected be made available under the CC-BY-SA 3.0 License ?
 * 2. Is Wikipedia going to post a banner for every academic research group which wishes access to its user base?
 * In any event, it is inconsistent for WMF to market Wikipedia as free from advertising and then post advertisements. Nobody Ent (Gerardw) 21:33, 16 December 2011 (UTC)
 * 1 The anonymised data will be released under an open licence as per meta:Research_talk:Dynamics_of_Online_Interactions_and_Behavior
 * 2 Good question. There has been a certain amount of discussion on RCom and elsewhere as to whether we need to regulate the amount of research surveying of the community or merely regulate its quality and ensure that the anonymised data is openly licensed. I think we should limit the amount of surveying of the community, and have tabled a proposal for doing so at meta:Research:Omnibus_Survey.  Ϣere Spiel  Chequers  09:37, 17 December 2011 (UTC)


 * Personally, I would be happy to see the banner restored. bobrayner (talk) 21:40, 16 December 2011 (UTC)


 * There should be absolutely nothing in the banner that takes you directly to the survey and thus submits any data; the only acceptable banner is one that takes you to the explanation and the warning about data being submitted. No-one should (even accidentally) click and hand over their user-information without a warning given. Choyoołʼįįhí:Seb az86556 > haneʼ 22:23, 16 December 2011 (UTC)
 * Agreed; this does not seem to rise anywhere close to the level of informed consent. I can't imagine my IRB - or any others - approving a study that collects data from participants even when they explicitly exercise their right to not participate. Merely clicking on a banner ad with the text "Please help advance research with a quick interactive online experiment" does not seem sufficient to convey informed consent.  Further, is the data collected via that click discarded if participants opt out of the survey?  CAN (non-)participants explicitly opt out?  ElKevbo (talk) 01:38, 17 December 2011 (UTC)
 * Yes: the "intermediary step between the moment when a user clicks on the banner and the actual redirection to our website" that we propose to introduce will serve exactly the purpose that you describe. Upon clicking on it, the banner will expand slightly to inform you of the information that will be passed to us for performing the participant eligibility check, and ask whether you want to continue. If so, the information will indeed be passed and you will be redirected to the survey. Note that at this stage, you have not agreed to participate in the study already, which you will do upon reading the description of the study on the survey landing page. If you eventually decide not to take the survey, the data collected will be discarded. SalimJah (talk) 02:16, 17 December 2011 (UTC)
 * Thanks; that seems to address my concerns related to informed consent. ElKevbo (talk) 03:39, 17 December 2011 (UTC)
 * Shouldn't this be listed at Template:Centralized discussion? --Lambiam 22:33, 16 December 2011 (UTC)
 * It may not have any bearing on the issue that is currently on the table but I have been confused ever since I read that the earlier discussions decided that a banner is less intrusive than a message posted to user's Talk pages. That seems to be completely backwards to me.  (It also seems to be contrary to best practices in soliciting participation in a survey e.g. Dillman's Tailored Design Method.) ElKevbo (talk) 01:38, 17 December 2011 (UTC)
 * A banner will only be seen by those targetted by the survey, and only for the duration of the trial. A talkpage posting would generate an email to the 70% of us that have enabled Email telling us that our talkpage had been edited, and the talkpage notice would sit until archived. So if you were on wikibreak for the duration of the campaign you wouldn't see the banner invitation, but if it had been a talkpage invite you'd have that sitting on your talkpage when you're back. I suspect this means that a banner is more intrusive for a few hundred highly active users and a talkpage invite more intrusive for the thousands who are here a few hours a month. There's also the practical issue that a banner can be displayed until you have achieved the target number of responses, but with a talkpage invite you need to put a date cutoff on the invite - with an obvious temptation to set a long date because most people running surveys aren't worried at getting excess response providing they get their target.  Ϣere Spiel  Chequers  09:20, 17 December 2011 (UTC)
 * I think this version is acceptable and the banner should be restored. Mike Christie (talk - contribs - library) 02:02, 17 December 2011 (UTC)
 * Incidentally, I think it would be helpful to also post at WT:VPT and at the meta discussion. Mike Christie (talk - contribs - library) 04:10, 17 December 2011 (UTC)
 * The first link should be WP:VPT. Graham 87 09:20, 17 December 2011 (UTC)
 * Mike, could you explain what version of the banner you are referring to as being acceptable? I can only see a link to the original version rather than the one that has yet to be produced without logos. Fæ (talk) 09:44, 17 December 2011 (UTC)
 * I had no problem with the original version, so I am commenting here solely on the description of the changes given above -- the improved segregation of user data, for example. I didn't (and don't) think the logos were an issue; it was quickly evident that this was not (in my opinion) advertising, and that it was WMF-sanctioned, so I felt it was OK. Mike Christie (talk - contribs -  library) 14:02, 17 December 2011 (UTC)
 * Thanks for the clarification. It seems a moot point as SalimJah has stated above that a version without logos will be proposed, so luckily that's one less thing to negotiate a new consensus for. --Fæ (talk) 20:18, 17 December 2011 (UTC)


 * Who will be the single point of contact that will have authority for implementing any final central notice? What will be the transparent process for handing complaints from our community or the public that may relate to the central notice, the linked websites, the study itself and the published data? Thanks --Fæ (talk) 09:34, 17 December 2011 (UTC)
 * I am sorry; I appreciate your efforts to make changes, and I understand that the Foundation has indicated that this would be acceptable, however I personally strongly object to this use of a banner. I believe that the banners should only be used for the annual fundraiser, and critical announcements; I do not think that it is part of the mission of this project to help an external body perform research. I think that if it is permitted, it sets a dangerous precedent; hundreds of other research campaigns could make similar requests, and it would be difficult to justify rejecting one over another. I do not want this project to turn into a research experiment.  Chzz  ► 13:21, 17 December 2011 (UTC)
 * imo no other organization than wikimedia foundation or chapter should be able to conduct anything by using a banner. thats the promise to the editors and users, and thats the promise which is the spine of movements financing. in case wmf or a chapter conducts it (i guess there is a chapter in isreal and france ...), they are able to clearly define who will participate and why (it might of course be some university) and under which legislation the original data is collected by whom, name all data fields that are collected, how long this data is stored, what happens with the data, who else gets the data, either original or treated somehow. if there is a payment involved it should ideally be treated by a different party which only gets the payment information. --ThurnerRupert (talk) 13:57, 17 December 2011 (UTC)
 * Chzz & ThurnerRupert: The goal of Wikipedia and WMF is to promote the free diffusion of knowledge. As a research team, our goal is to create new knowledge and release it for free. I think that those objectives are eminently complementary, even more so that we collect data that will be directly useful to WMF and allow the community to tackle research issues about itself that are not addressed through the semi-annual editor survey. The local chapters have already started to recognize and leverage the complementarities that exist between this project and academic research. A few months ago, Wikimedia France has run a research banner in collaboration with researchers at Télécom Bretagne in order to collect 16,000 responses to a research survey aimed at better understanding how Wikipedia is used, both from the reader and the contributor's perspective [1]. We are not using Wikipedia and WMF's resources for our own narrow research purposes here. During the last two years, our research team at the Berkman Center has developed a deep and sustained research collaboration with the Wikimedia Foundation. We are deeply committed to performing this research in the least intrusive, most community compliant way possible and want to let the public benefit widely from the results and data we generate. When the community opposed our initial recruitment method proposal (that we presented on the Admin Board back in march 2010) and suggested the CentralNotice option as a less intrusive method instead, we took this as an opportunity to develop new banner features that would be useful to the community for purposes of tailoring future CentralNotice campaigns to the group of eligible users. It took us 18 months to get this highly efficient banner ready, and now that it has been taken down, we are working again towards developing another new banner feature that would fully address the (legitimate) privacy concerns raised above, with the hope of promoting data management transparency for this and future Wikimedia campaigns. Our objective is to advance the "big picture" and add to the sum of knowledge, while proposing to Wikipedians to undertake an interesting (and fun!) exercise that would make the use of the research money that we spend as beneficial to this project as possible. I agree wholeheartedly with the idea that this project should not turn into a research experiment. I don't really understand, however, the argument according to which now that a precedent is set, it will be impossible for the community to monitor or reject the other similar research requests that might ultimately accrue. I first have the feeling that the number of research teams out there that would be interested in building a long-lasting relationship with WMF and the community for purposes of performing research in the spirit of leveraging their own resources to advance the purpose of this project should be manageable. Second, there is no reason imo why the community couldn't define from its own perspective what kind of Wikimedia research is acceptable and what kind is not, eventually tweaking the study's methods in order not to interfere with the normal, smooth functioning of the community (as we are trying our best to do here). Our narrowly targeted banner has already run for about ten hours on enwiki and collected 830 complete responses. It would arguably not take much more time to collect the 650 additional ones that are required to complete this project. Aren't a few more hours of an enhanced recruitment banner - revised along the lines suggested by the community - worth the "trouble" in the face of what we are all trying to achieve here? Rather than considering it impossible or even harmful to try to build a successful and mutually beneficial research collaboration with any kind of committed external body of researchers, why don't we take this opportunity to set the foundation of a workable path for future researchers interested in investing their resources and expertise in a way that serves the purpose of this community? As a matter of process, I'd be thrilled to have this research demonstrate that such a successful collaboration can be achieved... SalimJah (talk) 20:07, 25 December 2011 (UTC)
 * Will you be publishing both the raw data your research under the CC-BY-SA 3.0 License and the GFDL? Will you be submitting to peer reviewed journals? Nobody Ent (Gerardw) 20:14, 25 December 2011 (UTC)

Could someone please provide any updated info on this? Thanks. -- Nazar (talk) 08:34, 20 January 2012 (UTC)
 * Answered here. SalimJah (talk) 15:05, 20 January 2012 (UTC)

Where to go for an announcement to be fully announced
I write as a mildly-committed Wikipedia member, not at all related to WMF. But I have been observing these Wikipedia community-WMF communication difficulties with some concern (although I hate this sort of adversarial community vs WMF attitude some have, I view the two as interlinked and interdependent and ultimately interwoven) and I think one source of confusion (and I may be wrong here) is that there is a poor sense of where the WMF should make an announcement so that the community has the opportunity for sufficient say and sufficient notice. There seems to be a number of places, IRC, Mailing Lists, Village Pump, RFC's, Centralized Discussion template, Signpost, etc. So I think there should be some discussion in this regard. It seems that WMF did announce this banner in some places, but not others, and so I think there should be some discussion to make clear what is the best place for the WMF announcements. I admit I'm not absolutely sure if this is a matter of confusion or not, but it seems that way for me, so just throwing it out there, discuss if you'd like Jztinfinity (talk) 10:21, 25 December 2011 (UTC)