Hello all,
Looking through Jake's great list, I'm most taken with 6, 7, and 8. In particular, 6 and 7 assumes the motivation of our partners is somehow profit based. As an academic economist, this seems appealing to me, but I wonder if it is true. I think most of our editors want to use TWL because it will improve the accuracy and usefulness of wikipedia and they want to be a part of that improvement. I wonder if our partners don't feel the same way. If they do, then it would be useful to show our partners that inexperienced editors who apply for access to TWL use that access to improve wikipedia (perhaps as much or more than experienced users). However, if that isn't true, if inexperienced editors aren't using TWL very much or aren't editing wikipedia after getting TWL accesses, then that may be a more pressing problem than expanding access. Sam, can that be something data analysts can look at?
Best,
Shane

On Tue, Aug 2, 2022 at 5:00 AM Sam Walton <swalton@wikimedia.org> wrote:
Hi all,

Thanks again for this discussion, I appreciate all your thoughts on this topic.

Steven, I agree that starting with some data analysis to understand how we might want to lower the activity requirements is a simple and achievable first step. I've begun summarising this topic at https://phabricator.wikimedia.org/T314357, including some of the questions we might want to ask of the data. Please feel free to comment there with further thoughts on that aspect.

In terms of comparisons with the board election criteria, one nuance we have here is that it only takes one or two users abusing the terms of use (e.g. scraping and mass-downloading content) for publishers to start pulling out. We've been conservative with the criteria for fear of a few users causing everyone to lose access.

Lastly, I want to highlight something Jake noted - while it's true there are paid folks at the Wikimedia Foundation who could be working on renegotiations and lowering these requirements, that work has to come at the expense of something else, primarily expanding the library with new content for the existing eligible users. Deciding to work on this will mean pausing some of that work.

Please share your thoughts on data questions we should answer to support this discussion at the Phabricator ticket above!

Best,
Sam

On Tue, 2 Aug 2022 at 08:17, Jake Orlowitz <jorlowitz@gmail.com> wrote:
Writing with my emeritus hat on, and no special influence...

1. When TWL started, the requirements were 1 year and 1000 edits.
2. The change to 6 month and 500 edits was made because we could get away with it at that point; otherwise it was just arbitrarily "better"
3. We want as many good editors as possible to have TWL access.
4. There has to be some minimum criteria or else bad actors will create and farm accounts to use solely for non-editing purposes.
5. There has to be some minimum criteria or else publisher partners will feel unsafe and revoke access.
6. There is tremendous benefit *to publishers* from being included in TWL, namely more citations and pageviews
7. There is benefit *to publishers* from having more editors have access to TWL, if they actually use it to edit
8. I see no practical reason why within 5-10 years the requirements shouldn't be a mere 1 months and 100 edits with 10 recent edits and no blocks.
9. Renegotiation is time-consuming but beneficial, and if done regularly, can keep partnerships fresh
10. I suggest the TWL team set a 3-year goal of lowering requirements to 3 months and 300 edits with 10 recents edits and no blocks.
11. This interim goal could be agreed to in advance and over time, so that partners would have time to adjust
12. Partners could be provided with data that this wouldn't lower safety or security and would increase citations
13. Partners could be provided estimates of citation increase based on current rates times the increased editor base
14. The effort to reduce requirements is no more important than adding new partners or adding partners to "bundle" access--these three should all be pursued by staff.

Again, just my opinion as a TWL old-timer.

Cheers,

Jake Orlowitz
Founder of The Wikipedia Library
Seeker of well people and sane societies
Lead at WikiBlueprint
  media: @jakeorlowitz
  blueprint: wikiblueprint.com


On Fri, Jul 29, 2022 at 12:33 AM Paul S. Wilson <paulscrawl@gmail.com> wrote:
Not to debate, but to ponder:

While the Wikimedia Foundation has many millions of dollars --  passionate advocates, volunteers, editors, donors, lawyers, and tax-exempt status -- such is not the case with over 80 for-profit publishers which have already been generous enough to lower their paywalls for the Wikipedia Library.

Each of these 80 plus publisher have hammered out a legally binding deal with Wikimedia, one that that cost them valuable time and money, lost revenue, and legal expenses.. 

Once.

Try again - or lose them? 

For what? For whom? To what benefit?



On Thu, Jul 28, 2022, 9:39 PM Steven Walling <steven.walling@gmail.com> wrote:


On Thu, Jul 28, 2022 at 7:25 PM Paul S. Wilson <paulscrawl@gmail.com> wrote:
Às Sam Walton wrote, "lowering the requirements would necessitate going to each publisher, having them agree, and then likely signing a new agreement".

This is a fruitless discussion.

Negotiating (or renegotiating) contractual agreements with partners is why we have staffed partnerships and legal teams at the Wikimedia Foundation. There are people we pay with donor funds to do nothing but this. 



On Thu, Jul 28, 2022, 5:55 PM Steven Walling <steven.walling@gmail.com> wrote:


On Thu, Jul 28, 2022 at 1:50 PM Shane Murphy <shane.music@gmail.com> wrote:
I'm a TWL coordinator in charge of approving applications for Project MUSE. I don't have much for feedback or a strong opinion, but here is my experience. Almost all applications I get qualify (and are approved) and when I reject an applicant I try to include a message encouraging reapplication when the applicant has more experience. I do not think that I have ever seen a reapplication in those cases, although I could be wrong. To me, that implies that many editors who do not get access either found that they did not need access or decided not to continue editing Wikipedia (I just skimmed the applications I've rejected ~25 over 5 years or so, it looks like most are not active editors anymore but a few are). I'm not sure if rejected applicants stopping editing is a great outcome, but the high rate of accepted applications and the low rate of long-term participation of rejected applications makes me think the process is working well.

What the current process doesn’t show is the hundreds of thousands of potential editors who don’t even apply. 

As another example here: voting in the Board of Trustees elections requires 300 total edits and 20 recent edits. Why is Wikipedia Library access more restrictive than what counts as an active volunteer we trust to elect our governing board? 


Shane Murphy
smmurphy

On Thu, Jul 28, 2022 at 4:24 PM Paul S. Wilson <paulscrawl@gmail.com> wrote:
I concur with Sam Walton's expert  observation (not conclusion): renegotiating current legal agreements with some 80 plus generous publishers is a much higher bar than the already nominal requirements for Wikipedia editors to access their resources, a low bar for some 57,000 editors.

Paul

On Thu, Jul 28, 2022, 11:48 AM Steven Walling <steven.walling@gmail.com> wrote:


On Thu, Jul 28, 2022 at 3:37 AM Sam Walton <swalton@wikimedia.org> wrote:
Hi,

I'm the Product Manager for The Wikipedia Library, thanks for starting this conversation! Let me start by providing some background context around how and why the criteria are set this way.

Way back when the library started, it was an uphill battle to get publishers to agree to participate. While some were happy to jump right in, others were very wary about making their paywalled materials free for a nebulous group of folks from around the world. The editing requirements were primarily designed to convince those publishers that the only people who would be getting access to their materials were active Wikipedia editors who would be using their access primarily to contribute to Wikipedia. The initial requirements were actually even higher than they are today, at 1000 edits and 12 months of activity. We lowered that to 500 edits and 6 months around 2015 because it was clearly excessive. Last year we added the 10+ edits/month and no active blocks criteria as we shifted ~half of the publishers to the instant-access model where no applications are required. I think those two criteria are pretty minor compared to 500/6 (if you don't meet the 10 edits criteria you can make them all in the space of an hour and get access again right away).

Unfortunately, lowering the requirements isn't quite as easy as simply making an internal/community decision. Almost all publishers who provide access via the library sign an agreement with us. We don't pay for access, so adding new publishers is entirely dependent on convincing them to join the program. That agreement currently includes the editor activity criteria. This means lowering the requirements would necessitate going to each publisher, having them agree, and then likely signing a new agreement. This was enough of a challenge when we made the 2015 change, but by now we have something like 80 partnerships and the criteria need to be the same for all publishers, so this would be quite the undertaking. That's not to say it isn't worth the effort, I just want to clarify what's required.

In terms of who qualifies under the current criteria, the figure is somewhere around 57,000 active editors in total today, obviously with more receiving the eligibility notification each day.

In terms of lowering the requirements further, I do agree in principle that it would be nice if newer good faith editors were able to access the library, but we need to balance this with publishers' willingness to continue providing this free access. When we say '500 edits and 6 months of editing' this generally puts folks we're pitching to at ease, and makes them confident the program isn't likely to be used by people who aren't actively editing Wikipedia and have simply registered a free account to use the library for personal reasons. I'm not sure what the boundary is there, i.e. whether we would be as successful if we said '250 edits and 3 months', for example. I'll also say these numbers aren't hugely data-informed. We didn't analyse at what edit count/tenure threshold we can be confident in saying 'this is an active and engaged content contributor' versus 'theres a chance this user has made trivial edits simply to get access', but that's certainly a research project I could see being useful here, rather than making an arbitrary change that we can't justify with data.

Because of the work involved in changing these criteria I'm very hesitant about doing so, but I'd love to hear what you all think and whether the above changes your perspective at all.

Best,
Sam

Thanks for all the background Sam. 

"makes them confident the program isn't likely to be used by people who aren't actively editing Wikipedia and have simply registered a free account to use the library for personal reasons" 

I would say that any reasonable person would probably agree that along with all the other reasonable criteria like no blocks and a minimum number of recent edits (which I agree seem useful), that a bar of at least 200 edits and 3 months of editing would also suffice to represent real commitment to editing Wikipedia. 

If we haven't even *asked* publishers about lowering the requirement for seven years, don't you think it's worth at least starting a conversation about expanding access, given that it's pretty impactful for our mission? A good start might be doing a simple data pull to see how many more editors a lower bar would include. 
  
 
On Wed, Jul 27, 2022 at 1:38 PM J West <jessamyn@gmail.com> wrote:
I can absolutely see lowering the number of edits required, but it
might also be worth looking at or adjusting what sort of edits we are
looking for or adjusting them per-resource though that might get
messy. For example, people who are creating articles need to have a
different, and one could possibly argue higher, level of Wikipedia
literacy and familiarity than someone doing automated edits. They're
both valuable! But may have different needs w/r/t the Library.

As someone who approves people for a few different WL resources, I see
basically two kinds of applicants

- People who just apply for access to ALL WL resources and just are
rolling the dice about whether they'll be accepted. What they apply
for doesn't match their areas of interest or expertise or even editing
areas, or their qualifying edits are all profile page edits
- People who apply for a narrowly-tailored set of resources that match
their editing expertise

I guess the larger question is whether WL resource access is seen as a
perq for longer time editors or if it's supposed to just be a tool to
help people edit Wikipedia.

Thanks for sharing that's valuable to hear. I would put money on the fact that a lot of helpful new editors who don't yet qualify wouldn't even bother to apply, for lack of awareness or just giving up after seeing the requirements on the main landing page. 

Your larger question is definitely the right one. I would say that if our mission is to give everyone on the planet free access to the sum of all knowledge, we need to treat it like a resource that's available to anyone who wants to make a good faith effort to create and edit articles. The bar for access should be the minimum necessary to ensure that you've shown real interest and ability to make use of the library. If there are limits on the volume of editors who can participate, there are other tools we can use like making access expire if you don't use it. (Do we do that already? Pardon my ignorance here.)
 

_____________
Jessamyn West
User:Jessamyn
box 345, randolph vt 05060

On Wed, Jul 27, 2022 at 4:26 PM Steven Walling <steven.walling@gmail.com> wrote:
>
> I wanted to start a discussion on lowering the threshold for access. Very very very few people qualify for the current requirements of 500+ edits, 6+ months editing, 10+ edits in the last month, and no active blocks. In fact basically this excludes any new editor no matter how good faith and helpful they have been. Even just lowering one of the account age or edit count thresholds would go a long way.
>
> I recently was pretty shocked to discover this high of a bar for access, after recommending the library as a resource to a new editor who has been doing a great job and (as a young student) could use access to academic source material in creating science-related content. I won't name them, but as an example this editor has over 300 edits and has created just over 50 articles, mainly for missing plant species.
>
> Do the participating institutions require this level of exclusionary criteria? How can we gather data to show them that there are good content contributors being excluded here?
>
> These requirements seem pretty absurd especially since many of the largest resources in the Library, like JSTOR, give any random person with a Google account access to 100 free articles per month. The risk profile of a Wikipedia who say has,100 edits and 1 month of experience has got to be less than that? We should pilot a threshold like that.
>
> Steven
> _______________________________________________
> Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
> To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org


--
Sam Walton
Senior Product Manager, Moderator Tools

_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org
_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org


--
Sam Walton
Senior Product Manager, Moderator Tools

_______________________________________________
Wikipedia-Library mailing list -- wikipedia-library@lists.wikimedia.org
To unsubscribe send an email to wikipedia-library-leave@lists.wikimedia.org