Thanks Erik for this clear and, as far as I can see, rather comprehensive report
There will always be mistakes done, both from us as individuals and as organizations.
Critical, though, is that we treat these mistakes with openness and tranparancies and that we learn from our mistakes
In my opinion you have now handled this issue in this correct way, and I hope we now can take in the necessary learnings in our routines
Anders
Erik Moeller skrev 2014-03-21 21:23:
Hi all,
I've just met with Lisa Gruwell and Sara Lasner about it to get more of a debrief of the situation. For the purpose of clarity, I'm looking into this on Sue's behalf while she's traveling; she should be able to look into it next week. As noted previously, this isn't a project I was previously familiar with, so bear with me if I'm getting any bits wrong.
Here are some initial high level observations:
- This project was initiated by Sue Gardner in response to a request
by Liz Allison from Stanton Foundation, who initially attempted to fund the project with a direct grant to Harvard. For administrative reasons, both Harvard and Stanton ended up preferring to have Wikimedia Foundation act as a fiscal sponsor for the position. (This included administrative oversight by Sara Lasner, and a minimal degree of programmatic oversight by Sara and Lisa, with a primary programmatic point of contact at the Belfer Center.) The project was overseen by Lisa Gruwell.
- WMF agreed to help recruit candidates for the position and to
provide three candidates to the Belfer Center for selection. Frank, Siko and Lisa participated in the first round of interviews. The first candidate we put forward was a former Harvard librarian and active WIkipedian, but she was rejected by the Belfer Center for a lack of knowledge in the field on International Security. Then, the Belfer Center posted the JD on a list-serve of top academic programs in International Security. WMF interviewed two candidates from this pool and Belfer selected Timothy Sandole for his strong academic background in International Security. He had just completed a master's program at Columbia University.
- The Stanton Foundation has a long-standing interest in promoting
awareness regarding issues of international security and nuclear security, which dates back to the founder of the Foundation, Frank Stanton (former president of CBS).
- The Stanton Foundation does not have a financial interest in these
topics. With that said, Liz Allison, who heads the Stanton Foundation, and Graham Allison, who heads the Belfer Center, are wife and husband, and the Stanton Foundation funds other programs related to international security.
- As noted previously, the Wikimedia Foundation communicated about the
program in a blog post: http://blog.wikimedia.org/2012/04/26/can-you-help-wikipedians-collaborate-wi... Timothy Sandole also disclosed his affiliation with the Belfer Center on his user page, but did not disclose the funding relationship with WMF or Stanton in the same manner.
- Timothy's residency included training programs, but it was heavily
weighted towards editing work.
- Sara Lasner acted as an administrative point of contact at Wikimedia
(handling payments, vacation requests, etc.). Not being steeped in Wikimedia's culture, Sara gave minimal guidance regarding policies and practices, but forwarded instructional materials and pointed out the above conflict-of-interest issues to Timothy. There was a communications contact at the Belfer Center, James Smith, who provided subject-matter guidance.
- Timothy himself compiled a weekly report to the Belfer Center and to
Sara, and a final report at the end of the project. With his permission, I've published the final report here:
https://commons.wikimedia.org/wiki/File:Timothy_Sandole_-_Belfer_Center_Repo...
I have a copy of the weekly memos as well, and we've asked for his permission to release them.
In addition, the Wikimedia Foundation compiled a report to Stanton at the end of the project largely identical to the report to Timothy. We've asked the Stanton Foundation for permission to release this report, as well, for the sake of full transparency.
Edits like the following are indeed problematic:
- potentially undue visibility for research conducted by the head of
the Belfer Center
- potentially undue visibility for the Stanton Nuclear Security
Fellowship, which was funded by the same Stanton Foundation which funded the program.
In September 2012, Sara Lasner had a call with Timothy Sandole specifically asking him to be conscious of not over-representing Harvard University in his research, and Lisa Gruwell sent an email to James Smith and Timothy Sandole regarding awareness of conflict-of-interest issues in general.
Timothy's edits weren't monitored in detail by the Wikimedia Foundation. We'll take a closer look now, and appreciate the community's help in ensuring that, in light of the above potential conflicts-of-interests, that they're consistent with policies and guidelines.
At the same time, it's important to note that Stanton Foundation did not stand to benefit financially from this project. The nature of potential bias here is more subtle (e.g. over-representation of certain academic perspectives). Again, it's important to note that in this respect, the relationship with the primary potential "beneficiary" (Belfer Center) was fully disclosed on Timothy's user page.
As far as I can tell, everyone involved acted in good faith - the Stanton Foundation tried to expand the availability of freely licensed high quality information in areas that it concerns itself with; the Wikimedia Foundation sought to try to support this request (based on a longstanding positive relationship with the funder); Timothy and the Belfer Center made good faith efforts to improve the areas of content he was tasked to improve. What appears to have been missing:
- a full and honest upfront conversation between WMF and Stanton early
on about any perceived or real conflicts-of-interest issues in the context of this work;
- strong follow-through in ensuring the highest standard of disclosure
regarding all funding relationships, beyond the initial blog post, and continued reporting, including the final report;
- a sufficient level of training and oversight for Timothy Sandole
beyond administrivia.
A lot of this can likely be traced to running this within fundraising -- it's totally fine that fundraising isn't positioned to oversee complex Wikipedia-related programs, so we underestimated the complexity of the work and situated it in the wrong place in the org.
Liam Wyatt and Pete Forsyth are right to point out that they noted the risks and issues early on, and they're also right to point out that the community-developed WiR program places emphasis on non-editing work for good reasons. Frank Schulenburg and LiAnna Davis also provided internal feedback and criticism early on, pointing out the COI issues and the risks regarding the project. (Thank you to both of them, and others who provided internal feedback; it's appreciated.)
We're now heading into the weekend, and beyond answering questions that are easy to answer without coordination, I'll refrain from commenting much further til next week. In response to Pine's point, I agree that we shouldn't rush to conclusions and give ourselves time to more fully evaluate things, so consider all the above to be preliminary and subject to revision as we more fully understand the situation.
We will update the wiki page at https://outreach.wikimedia.org/wiki/Wikipedian_in_Residence/Harvard_Universi... with more information and details. I encourage others to participate in this as a collaborative process.
The Wikimedia Foundation did and does not intend to undertake similar efforts again (programs that include paid editing), but these kinds of issues can extend to any program that includes active work on content. So my initial take is that we should aim too ensure that content-related programs are undertaken under a clear and simple set of public guidelines, and are situated in parts of the organization well-positioned to support them with subject-matter expertise. We'll discuss this more, and follow up on this as well.
Thanks,
Erik