Research talk:Omnibus Survey

Dear WSC,

Thank you for posting this proposal. I do subscribe to your worry that we might end up with a 'tragedy of the commons' if researchers use the Wikimedia community as a goldmine for quick and easy access to research data. However, I am not convinced that a Omnibus survey would solve this problem. I see three issues in particular:

• The envisioned benefits will only be worth pursuing if the only survey we allow researchers to conduct is to participate with the Omnibus survey. I am not sure if we either can get enough support for this or how we are going to enforce the use of such an Omnibus survey.
• The investments for such an Omnibus survey (software development) can be quite substantial while the efficiency gains are quite marginal: we can skip some demographics questions but I do not think that this will represent a significant return on investment.
• Even if we address the first two issues then I do not think that we will be able to free the necessary resources to develop an Omnibus survey. There are too many competing functionalities and other improvements on the wishlist of the Foundation.

An easier solution (to implement) would be that registered users give permission to share certain demographic data for research purposes and when such an editor participates with a survey then such demographic questions are automatically answered.

Let's sharpen our thoughts more so we can prevent a tragedy of the commons but come up with a solution that is easier to implement. — The preceding unsigned comment was added by Drdee (talk)

Hi Diederik,

Yes this will only work if we say this is the only way to get access to our community via watchlists, mailing lists or other mass communication. But if we don't say that then IMHO if the number of surveys I am asked to complete is typical of an EN wiki admin and if that continues or increases them the community will start to block this. As for getting support for this, are you talking about support from the research community or from Wikimedians? Obviously one group would prefer unfettered access to the wikimedia community, and the other is increasingly grumbling about spam. It might be that neither community would accept an Omnibus survey, in which case I suspect the community will simply block research spamming from anyone bar the foundation.

The investment does not need to be that great, we already have a survey and we could add more questions to it quite cheaply. Alternatively if money was the prime consideration we could simply tell the Academic community that we will permit one survey a year provided you all collaborate and produce it at no cost to the foundation other than the watchlist to the community. My preferred option is that we expand our editor survey and we invest in making it efficient for the people completing it, as I think that is the best way to maintain the privacy of our community members and be open to the whole of Academia.

I suspect that the Omnibus Survey would be easier to implement than a system where we stored a sheaf of demographic questions on those who opt in to that and release them to each researcher who surveys them. Aside from the technical complexities I doubt that we'd get a sufficient response for that to work. WereSpielChequers 11:37, 13 July 2011 (UTC)

This is really good. The only individuals who should be access individual data are trusted members of the Wikipedia communities (e.g. checkusers) and WMF employees. As far as the software development issue, as Wikimedia is open source WMF could determine the necesssary stubs and leave it to the research communities to develop the implementations. As the benefit from voluntary response i.e. not very good surveys is limited, Wikipedia is not losing much if the surveys are delayed for software development. 65.75.37.114 11:42, 17 December 2011 (UTC)

• May be a question for today's discussion: Who will be compiling / acting the surveys? Obviously, not the researchers themselves. Do you suggest to have one or several Rcom members to do it every year?--Ymblanter 17:22, 22 December 2011 (UTC)
  • I'm assuming that the foundation would resource that. If it came to decisions as to whether something should go in then RCOM or the community might get involved, but Foundation staff would do the data processing to make sure that only WMF employees had access to the raw data. WereSpielChequers 22:50, 2 January 2012 (UTC)