Grants talk:TPS/Wikimania scholars/Proposed 2015 Process

Re: Recruitment

On "these banners would now target logged in users with over 50 edits": that's consistent with CentralNotice/Usage guidelines, so I agree. However, the aim should be to have many many more applications than we currently do (see Wikimania/Scholarships/Vision 2015). In WMIT's experience, banners help little, while targeted emails are effective: by sending an email with Wikimania scholarship in the subject, one year we got some 150 interesting applications, as opposed to some 50 coming from other recruitment channels. Our newsletter recipients include current and past members, past donors and interested people; WMF can probably identify a suitable segment of recurring and "solid" donors to email. --Nemo 15:49, 3 December 2014 (UTC)

Definitely agree that targeted emails are very effective in increasing interest & applications, and both WMF Staff and the Wikimania Coordinator would be sending out such emails through multiple mailing lists. The banners would simply be used as a way to reach a broader audience. --Shouston (WMF) 18:46, 3 December 2014 (UTC)

Re: Perception that going to Wikimania is a “reward” for large contribution / activity

This is IMHO, the issue of the program. --Nemo 16:16, 3 December 2014 (UTC)

I think moving away from it simply being a "reward" for contributions is a great idea. The problem with having it as an reward is there are far to many deserving users for far too few scholarships, and unfortunately many users see not getting a scholarships as a declaration that they are not a worthy contributor – hence the common response of "I was demotivated and my contributions started to decline" by those who were unsuccessful. In the most extreme cases, abuse was directed at members of the Scholarship Committee over the result, which I hope can be avoided this year. CT Cooper · talk 16:47, 2 January 2015 (UTC)

Re: Relevant experience in the Wikimedia movement

I don't find this evaluation point particularly useful. In particular, "Level of online or offline impact, e.g. content created or participants reached" can't be measured exactly and is quite useless, in addition to coming third last in your own rank of criteria perceived as fair. --Nemo 16:16, 3 December 2014 (UTC)

This criteria is a reformulation of the previous “Activity within the Wikimedia movement” criteria, with the sub-dimensions of “Collaboration”, “Impact”, and “Community leadership” replacing the previous sub-dimensions of “Level of participation/involvement” and “Level of Impact”. Given one of the target outcomes for the proposed program is to encourage experience sharing among Wikimedians, this criteria will hopefully highlight applicants who have compelling experiences to share with the broader Wikimania audience. If the phrasing of “relevant experience within the movement” is confusing, any ideas on how this could be phrased better?

To your point on Impact, you’re right that a large number of contributions isn’t the only (or best) way to measure impact. An applicant could describe their “impact” either quantitatively and qualitatively, and both the offline and online dimensions will be considered. --Shouston (WMF) 23:18, 8 December 2014 (UTC)

Measuring offline activities is very difficult, but has to be considered equally simply on grounds of fairness. I think the current use of terms is fine, except that I would make "experience" plural, and would caution against getting to carried away with scholarship jargon. CT Cooper · talk 16:53, 2 January 2015 (UTC)

Re: minimize reader-only applications

On "minimize reader-only applications", that's not a goal. The goal is to have engaged Wikimania attendees. I agree it's rare that formerly-passive users become suddenly very engaged thanks to Wikimania, but it's not impossible. I have no idea where to draw a line, but requiring a username or 1 edit might be enough to catch the vast majority of obvious cases. --Nemo 16:16, 3 December 2014 (UTC)

Realistically, we have to have something substantial to go on in order to reward someone a scholarship. In my experience, the few positive outcomes of giving scholarships to those without a significant record of contributions has been heavily outweighed by negative outcomes in other cases, including people using WMF money to get to the host city then not attending Wikimania, with one reason I noted being that they just couldn't follow the conference sessions. Even for those who it did result in positive outcomes, giving someone a scholarships is very expensive way of motivating someone to contribute, and there are much cheaper alternatives.

Applications from readers and those with minimal contributions took up a lot of time last year, and efforts to minimise them will be welcomed. CT Cooper · talk 17:00, 2 January 2015 (UTC)

Phase 2: Evaluation

The biggest missing factor here is that you're not even trying to assess how sincere a Wikimedia projects interest the applicant has in attending Wikimania (as opposed to mere tourism interest), nor what concrete need for a scholarship. If you want to go beyond the usual brutally numerical metrics, what is needed first of all is that the applicants be allowed to post their applications in their own language. --Nemo 16:16, 3 December 2014 (UTC)

The topic of financial need has been a recurring topic of discussion, however this would necessitate collecting private financial information on every applicant, and would introduce a substantial complexity in trying to accurately compare financial need across applicants in different contexts. As such, we rely on applicants being considerate of other’s financial need when applying; the program overall includes the following statement to this end: “It is asked that applicants carefully consider their financial need against the needs of other applicants from all over the world. Many applicants do not have the financial means to attend Wikimania without this opportunity.”

On the topic of applicants being able to apply in their own language, as English is the primary language used at Wikimania, having sufficient English proficiency is an important aspect to consider. On the topic of “interest in attending Wikimania”, we can bring this up as a topic of discussion with the Scholarship Committee, to better understand if this question has been very useful in evaluating applicants in the past. --Shouston (WMF) 23:18, 8 December 2014 (UTC)

I agree that considering applicants' financial situations is simple impractical and would cause too many privacy concerns. While not ideal, asking people to self-assess their own finances is the only viable option here.

Last year, applicants could and did put applications in their own language, and I'm sure there were several successful ones which weren't posted in English. However, putting an application in English is helpful as it allows everyone on the committee to understand what it says, as well as allowing us to assess one's English abilities. Putting an application in a language other than English will inevitably mean that some on the committee will have to use online translators which rather defeats the purpose in my opinion. CT Cooper · talk 17:14, 2 January 2015 (UTC)

Re: size of their home Wikipedia

I don't like this special casing. It would be better to say "size of their home language community", including all the subdomains (language editions of all Wikimedia projects) for that language. Of course the size will be dominated by Wikipedias anyway, except a few cases of particularly big Wiktionary or Wikisources. --Nemo 16:16, 3 December 2014 (UTC)

Great comment on using "size of home language community" across all projects. The main reason the proposal uses just Wikipedia was to minimize implementation issues, as it seemed reasonable that the "size" of the language community (over all projects) would map pretty closely to the "size" of the Wikipedia. Given that "size" is a loose categorization (of just Large/Medium/Small), are there many language communities where the size of a sister project would move a language community from one category to another (e.g. from Medium to Large)? If so, then "size" over all projects might be better. --Shouston (WMF) 18:46, 3 December 2014 (UTC)

• I disagree with this criterion. I understand the intended effect, but the size of home Wikipedia in addition to Global North / Global South makes a systemic bias:
  • Most of Global North contributors edit large (English, French, German, Spanish etc.) or medium-sized Wikipedias, with very few small Wikipedias with active communities. In fact, the only active community having both a chapter and a small Wikipedia is a Macedonian one (and to the certain extent Cantonese), while other languages with small Wikipedias either have small Wiki community (Albanian, Bosnian, Latvian) or have relatively few speakers.
  • The only three languages having a large Wikipedia and not spoken outside Global South are Cebuano, Waray-Waray and Vietnamese (all three largely bot-created). At the same time, this group receives 1/3 of 75% scolarships. Thus most likely this group will be completed by users contributing to Wikipedias in their second language, e.g. native Tamil speakers contributing to English Wikipedia.
  • Thus it is quite unfair that the number of speakers of these languages and Wikipedians working in these languages is different in these three groups, if one considers Global North and Global South separately, although this division is less important if do not make a North/South division.
  • In addition, with generalisation of bot creation of articles, size of Wikipedia is no longer a good criterion. For example, this means that a contributor to Cebuano Wikipedia will be treated the same way as a contributor to English Wikipedia, but not the same way as a contributor to Tadalog Wikipedia. On the other hand, contributor to Volapük and Arabic Wikipedia will be treated equally, despite obvious differences in sizes of communities
  • Many users contribute to Wikipedias in several languages. For example, a person from Hong Kong can contribute to Wikipedias in English (large), Mandarin Chinese (medium) and Cantonese (small) — how can we treat these people? It is likely that these three groups will have significantly different profiles on the average.
  • I would also try to take into account languages a person is speaking but is not contributing. For example, a person speaking a small regional language but contributing only to Wikipedia in the national language may be a more interesting candidate than a monolingual speaker
• To sum up, in my view, more intelligent criterias are needed than simple size of Wikipedia. I would suggest trying some of the following: active users (e.g. clusters of 1000+; 100+ and <100), articles per speaker, users per speaker etc. Bringing 25% of users of large Wikipedias from Global South should not be a goal in itself, but respecting linguistic diversity should be more valued instead — NickK (talk) 10:11, 9 December 2014 (UTC)

Great comment, especially the point on systemic bias when using "size of WP" + Global North/South split. The reason for including the GN/GS split was that an applicant contributing to enWP from the US has a very different context & very different set of experiences than an applicant contributing to enWP from Malawi; as such, they should be considered differently. So while Cebuano, Waray-Waray, and Vietnamese are the only "Large WP" languages spoken in the GS, there are a number of contributors who live in the GS but contribute to enWP (for example, contributors from India). So 1/3 of 75% of scholarships would not go to just Cebuano, Waray-Waray, and Vietnamese community members. However it's a good point that many GN contributors will map mostly into the Large/Med WPs. One potential change is that the definitions of "Large/Med/Small" are different for GN vs. GS residents (e.g. maybe over 500K articles is a better definition of "Large WP" for those living in GS).

However, if I understand your general comments correctly, it seems like you agree that language community is a better split than geography (as it was previously), but that using only the "size of WP" isn't sufficient / the best way. If we were to replace "size of WP" with "number of active users by language community", what time period would be sufficient to define an active editor? A year? Two years?

in regards to your comment around how users contribute in multiple languages, it's a great point. However, to simplify the complexity of this system, using primary language community seemed the best path forward, as many users can choose their primary. --Shouston (WMF) 19:24, 11 December 2014 (UTC)

Hi Shouston (WMF) and thanks for your comments. Your suggestions made me think of the following groups that would more accurately represent users' experience:

• Global North:
  • Users contributing to a project in the main national language, e.g. users from Italy contributing to Italian Wikipedia or users from Belgium contributing to Dutch Wikipedia.
  • Users contributing to a project in a regional language not spoken elsewhere, e.g. users from France contributing to Breton Wikipedia or users from USA contributing to Navajo Wikipedia
  • Users contributing to a project in a minority/immigrant/international language spoken elsewhere, e.g. users from Poland contributing to Belarusian Wikipedia, users from the UK contributing to Bengali Wikipedia or users from Portugal contributing to English Wikipedia

  These groups should not necessarily get 1/3 each, the first one will be far larger then others

• Global South:
  • Users contributing to a project in a local language which is not a Global North language, e.g. users from Mali contributing to Bambara Wikipedia, users from Vietnam contributing to Vietnamese Wikipedia or users from Peru contributing to Quechua Wikipedia.
  • Users contributing to a project in a local language which is also a Global North language, e.g. users from Brazil contributing to Portuguese Wikipedia or users from Malawi contributing to English Wikipedia.
  • Users contributing to a project in a minority/immigrant language which is not a Global North language, e.g. users from Thailand contributing to Chinese Wikipedia or users from Saudi Arabia contributing to Bengali Wikipedia
  • Users contributing to a project in a minority/immigrant/international language not commonly spoken in the country which is a Global North language, e.g. users from Egypt contributing to French Wikipedia or any other similar uncommon combination

  Here first two groups will most likely be far larger then other two

Some of these groups may be split by language sizes, but number of active users (over any period, as it is highly unlikely that at any moment Volapük Wikipedia will have more users then Arabic Wikipedia) will be a better measure then size of Wikipedia (easily impacted by bots), as it is a better measure of both potential audience than can be impacted and of size of community user is representing.

In my view, such groups more accurately represent users' experience than simple number of articles, as three cases (English speaker in the UK), (Swedish speaker in Germany) and (Vietnamese speaker in the US) are definitely different cases, although based on Wikipedia size + Global North / Global South they represent very different experiences — NickK (talk) 10:47, 12 December 2014 (UTC)

P.S. I think that for more transparence it would be interesting to have a statistics in the following form: (country, language, granted / not granted), obviously anonymised. In my view, it would be interesting to analyse this data to find out which distribution would be more fair — NickK (talk) 01:29, 13 December 2014 (UTC)

Re the p.s., consider that users from discriminated countries (essentially Europe) often gave up applying in the first place. --Nemo 22:31, 13 December 2014 (UTC)

Do you mean there was an effect that a number of users decided not to apply because of Global North / Global South discrimination? — NickK (talk) 02:15, 14 December 2014 (UTC)

Sure. --Nemo 11:47, 14 December 2014 (UTC)

Re: applicants within 10% of the cutoff

The "cutoff" is not defined, what is it? The minimum score? --Nemo 16:16, 3 December 2014 (UTC)

Good catch - this part wasn’t very clear. All scholars will have a numerical Phase 2 score, and all scholars in a particular sub-group will be ranked based on this score. Scholarships will be awarded to the top X applicants, where this “X” can’t be defined yet as it will vary by the cost per scholar (e.g. hotel, flight costs which vary each year), and then that per-scholar cost is divided by the total budget, to determine how many scholarships can be provided in each group. As such, the “cutoff” is the score associated with the last awarded scholarship in that sub-group. --Shouston (WMF) 23:18, 8 December 2014 (UTC)

That seems like a reasonable approach to me. CT Cooper · talk 17:16, 2 January 2015 (UTC)

Re: Coordination with Wikimedia organizations

If the aim is to reduce duplication, I doubt the proposed steps help much. To reduce duplication, you'd first need to ensure that all the people served by the affiliates' programs, as well as all the criteria they use, fit in the WMF's application system. As for WMIT, for instance, there may be multiple missing pieces: a) reaching out to all Italian Wikimedia supporters, as per #Recruitment; b) being able to apply in Italian; c) asking the questions WMIT cares about, see #Phase 2: Evaluation; d) receiving all the information about the applicants, so that the affiliate is able to conduct the evaluation, which is not possible with the mere score. Other organisations may also use different eligibility rules. --Nemo 16:16, 3 December 2014 (UTC)

Good comments on coordination, but the goal is not to merge the affiliate's program into the WMF program. The proposal is focused on addressing two scenarios that exist today: (1) A Wikimedia organization who wants to run their own Scholarship program (e.g. defining evaluation criteria that they find most relevant), but wants to access all individuals who applied to the WMF Scholarship Program from their home country/community/group. This case refers to the first deadline, where WMF would simply hand over all applicants before any evaluation/scoring is done, and the organization would be responsible thereafter. (2) A Wikimedia organizations who wants to fund a larger number of applicants from their home country/community/group than would be funded through the WMF Scholarship Program. This case refers to the second deadline, where the all relevant, scored applicants would be sent over and taken out of consideration for WMF Scholarships. Any Wikimedia organization who doesn't want to access the applicants to the WMF program, would be able to move independently with their own processes, on their own timeline. --Shouston (WMF) 18:46, 3 December 2014 (UTC)

I agree there was some confusion, and even a little tension, regarding co-ordination, or lack thereof, last year so I think setting hard deadlines is a good idea. Inevitably, some affiliates will want to do there own thing, and I say let them be. If a user gets awarded multiple scholarships by applying through multiple programmes, then they are free to only accept one and turn down the others without causing too much hassle. CT Cooper · talk 17:21, 2 January 2015 (UTC)

Re: meeting the final recipients at Wikimania

This is very good. However, more personal interaction should be encouraged during the application phase as well. --Nemo 16:16, 3 December 2014 (UTC)

I believe that personal interaction with recipients during Wikimania is a great idea.

On personal interaction with applicants, this already happens to some extent as members of the Scholarship Committee (including myself) have frequently sent out e-mails to applicants asking for clarification on simple issues, such as verifying a person's username. I've always been mindful here that one has to be very careful what they e-mail applicants over in order not to give certain applicants an unfair advantage. For example, sending out e-mails asking someone to expand their application would not be appropriate. Regardless, I don't think it is practical for a team of volunteers to personally interact with even the shortlisted candidates, as there are simply too many, even if less than half the e-mails I send out actually get a reply. CT Cooper · talk 17:35, 2 January 2015 (UTC)

You talk of "shortlisted candidates" as if they were a huge mass of people. We're only talking of few dozens scholarships here, from a pool of few hundreds candidates (which IMHO should instead be thousands, but you disagree). More interaction is definitely possible, because we manage in (some) chapters: see also Wikimania/Scholarships/Vision 2015 where I and Martin outlined a possible process. --Nemo 17:34, 27 January 2015 (UTC)

On the current budget at least 75 WMF scholarships are given every year. I've not given a position on how many/what percentage of candidates should be short-listed if such a process were to be created, but going with 150% of that number, which would really be the minimum to have reasonable leeway, we're talking at least 112 people here. I do not see how interviewing each one etc. is practical. I do think reviewers should spend more time reviewing candidates though than they have previously as per #Re: Roles, Responsibilities and Incentives. CT Cooper · talk 21:07, 27 January 2015 (UTC)

Re: Overview

As promised, now that my holiday, Christmas, and New Year are behind us, I now have some time to give some detailed feedback on the proposed 2015 process. I hope I'm not too late as I believe some further feedback will be useful. I've divided my comments into sub-sections and signed every paragraph to make responding easier for all, and underlined key points.

"Started in 2005, the Scholarships program was initially handled by the Wikimania host country, but operation was transferred over to WMF in 2012. Since 2012, WMF has funded a total of 317 individuals."

I think that was good move as the hosts have more than enough to do as it is, and WMF leadership ensures some consistency from year-to-year. CT Cooper · talk 20:03, 2 January 2015 (UTC)

"Final decisions were made through joint discussion between the Scholarship Committee and WMF coordinators."

That was true in 2013, but not in 2014. The planned Skype chat between the WMF and the Scholarship Committee was cancelled and the results rushed out as quickly as possible. While I respected the concerns around releasing the results late, there were two unfavourable consequences of this rush. The first was that almost all of the spreadsheet annotations from the Scholarship Committee on the short-list of recipients was disregarded, meaning that myself and others sacrificed a significant amount of time for no benefit. Secondly, after the results were released, serious errors were spotted in the spreadsheet which couldn't then be rectified, meaning that a few applicants missed out on scholarships solely because of such errors. I believe this could have been avoided if either the Skype chat had gone ahead or phase three hadn't have been rushed. If there was only one thing I could change about last year's process, it would be avoiding this happening again. CT Cooper · talk 20:03, 2 January 2015 (UTC)

This looks indeed like a big failure of the process. On the other hand, it seems that WMF now wants to manage the entire process directly, which should clarify what (reduced) share of commitment and responsibility volunteers have? --Nemo 17:40, 27 January 2015 (UTC)

Re: Outcomes reported by 2014 Scholars

I think the decisions to publicly announce who receives a scholarship and to require the writing of a report are good ones. I would suggest though that such reports are better placed on the Wikimania wiki, as one of the wiki's purposes is to record outcomes from Wikimania – the wiki isn't locked until at least a year after the conference is over. I would also ensure that plenty of time is given for recipients to complete their report to a good standard, such as 90 days. CT Cooper · talk 20:03, 2 January 2015 (UTC)

On the contents of the reports, the outcomes are indeed interesting, particularly when split between small and large projects, but such methods of analysis do have shortcomings as already highlighted by others, so I wouldn't over read into them. CT Cooper · talk 20:03, 2 January 2015 (UTC)

Re: Feedback from 2014 applicants and recipients

Visa assistances seems to be the one big area we fell down last year, though I understand it won't be as bigger issue for Mexico. I would suggest including visas in the cost of scholarships to level the playing field for everyone, rather than having it on a request only basis, which I don't think has ever been very effective. CT Cooper · talk 20:03, 2 January 2015 (UTC)

I deliberately avoid the term "rejected" as it gives people an inaccurate impression of how the process worse, not to mention it also sounds rather harsh. Putting that aside, 46% of unsuccessful applicants believing the result was unfair is a much larger number than is desirable number. Making the process as open as possible, making the criteria clearer, and moving away from scholarships being a reward for contributions (per above discussion) will hopefully get this number down. CT Cooper · talk 20:03, 2 January 2015 (UTC)

I agree that scholarships should not be granted randomly, and attempts to put such a process in place would open a huge can of worms. On whether and how to favour new recipients over repeat ones has been a source of controversy within the committee and the community for a number of years. No formal process dealing with repeat recipients, such as deducting points, has ever been agreed on. Some clearer guidance on this would be helpful. CT Cooper · talk 20:03, 2 January 2015 (UTC)

Re: Summary of Issues

A fair and reasonable summary in my opinion. CT Cooper · talk 20:03, 2 January 2015 (UTC)

Re: Target Outcomes

I think having clear targeted outcomes is a good idea. I would perhaps also add that recipients also enrich Wikimania itself through their own participation, though I guess this is already said, but with different words. CT Cooper · talk 20:03, 2 January 2015 (UTC)

Re: Recruitment

Yes, I think the banner has frequently been misinterpreted as being aimed at readers as well as editors. I noticed that in both 2013 and 2014 the number of poor quality applications and/or applications from users without any chance of getting a scholarship went up significantly once the banners had gone-up. Setting a 50 edit baseline is a prudent move without being a case of editcountitis. CT Cooper · talk 20:03, 2 January 2015 (UTC)

I reject the supposed distinction between editors and readers. Wikimania is for members of the Wikimedia movement at large, which includes all supporters. --Nemo 17:40, 27 January 2015 (UTC)

I don't accept that definition of Wikimania, but even if I did, don't think giving travel grants to readers is a good use of finite funds. CT Cooper · talk 21:01, 27 January 2015 (UTC)

Re: Selection process

"Initial eligibility evaluation completed by WMF staff, rather than by the Scholarship Committee, to make better use of Committee members' time and focus their efforts on in-depth evaluation"

Yes, this will put the committee's talent to better use and make the role a lot less demanding, presuming there are the WMF staff available to complete this task. CT Cooper · talk 20:03, 2 January 2015 (UTC)

"Clear rubric against these criteria will be provided to the Scholarship committee, with scoring against each criteria returning to the 0-4 point scale, to simplify evaluation."

I don't think this is a good idea. The reason that the system was changed from a 0-4 to a 0-10 scale was to spread out scores a little better, as on the old system in practice almost everyone got a score of 3 or 4. It is also worth noting that the idea behind the 0-10 scale was that the main levels were 0, 2, 4, 6, and 10 – with levels 1, 3, 5, 7, and 9 only for when a reviewer couldn't decide between two levels, so removing it really just removes flexibility rather than making things simpler. CT Cooper · talk 20:03, 2 January 2015 (UTC)

I see your point. Perhaps the decision should be left to the committee once it forms? --Nemo 17:40, 27 January 2015 (UTC)

The Committee has now discussed these issues and it was the overwhelming consensus that the 0-10 scale be left as it was. CT Cooper · talk 21:03, 27 January 2015 (UTC)

Re: Eligibility Criteria for Phase 1

Yes, I think a clearer definition of Wikimedia contributions and activities is needed as there was a lot of confusion last year and people seemed to misread my list of examples in the reviewers' guide as being an exhaustive list, which it was never intended to be. This new definition seems reasonable to me. CT Cooper · talk 20:03, 2 January 2015 (UTC)

"automatic disqualification of 2014 recipients who did not complete a post-conference report."

Seems fair enough to me. How many 2014 recipients didn't complete a report? I would also consider expanding it to anyone who turned down a scholarship without a good reason or directed clear abuse at committee members, whether on or off wiki. CT Cooper · talk 20:03, 2 January 2015 (UTC)

Re: Evaluation Criteria for Phase 2

"While strong candidates were eventually selected using these criteria, both the Scholarship Committee and WMF staff felt the criteria lacked strong rationale and a clear and consistent rubric (especially for non-Wikimedia related activities)."

I was one of the main drafters of last year's criteria and I agree with this conclusion. I tried my best but the criteria needed a stronger baseline to be built from and more outside input on wording was needed. CT Cooper · talk 20:03, 2 January 2015 (UTC)

The non-Wikimedia related activities criterion didn't work at all, as it was never clear what was considered relevant or valuable, and having such a criteria effectively discriminated against those who gave all their time contributing to Wikimedia projects. I therefore welcome it's removal, which I presume means non-Wikimedia applicants will no longer be encouraged to apply, unlike in previous years, despite having no real chance anyway. CT Cooper · talk 20:03, 2 January 2015 (UTC)

The reason it didn't work is the lack of insight in the candidates, which is a failure of the process and not a given: see #Re: meeting the final recipients at Wikimania. --Nemo 17:40, 27 January 2015 (UTC)

Unfortunately, It doesn't matter how much insight you give the candidates if no one agrees on what is relevant or valuable. CT Cooper · talk 21:05, 27 January 2015 (UTC)

Re: Diversity system for Phase 3

The old region system didn't work very well and a replacement is needed, but I'm not sure whether this is a suitable replacement, to be honest. I think allocating scholarships based on the size of a recipients home wiki leaves the process open to gaming and the complications of it will outweigh any benefits. I do agree though with the 25/75 global north/south split, as long as minimum standards are defined and unused scholarships re-distributed, as has been proposed. I also welcome tipping the balance slightly in favour of non-male applicants, presuming that "decline to state" applications will be treated the same as male ones. CT Cooper · talk 20:03, 2 January 2015 (UTC)

Re: Roles, Responsibilities and Incentives

The proposal to define clearer roles is very welcome, as this seemed to cause confusion last year, particularly towards the end of the process. I believe bringing back a final discussion between the WMF and the committee on the recipient short-list to filter out any issues would be sensible, though I agree that the final decision should be the responsibility of the WMF and that fact should be made clear to applicants and recipients. CT Cooper · talk 20:03, 2 January 2015 (UTC)

"Offering of financial support to attend Wikimania (if needed), upon completion of duties"

Not withstanding the slight conflict of interest, I think giving committee members at least some assistance is justified. This has been proposed before and was a little controversial, but not doing so means Scholarship Committee members are almost guaranteed to miss Wikimania unless they can fund the trip themselves or get an affiliate scholarship which is disconnected from the WMF process (the only type which committee members can apply for without a conflict of interest). CT Cooper · talk 20:03, 2 January 2015 (UTC)

I agree with making more efficient use of volunteers' time, and I think 50 applications per person is reasonable and realistic, and will hopefully allow some high quality reviewing – I was rather alarmed last year to find that some reviewers were spending less than five minutes per applicant and were judging users purely by edit count or similar.

"Committee members will serve a more public role as ambassadors for the Scholarship Program, meeting the final recipients at Wikimania (at a mixer organized by WMF) and encouraging more individuals to apply next year."

I went out of my way to keep the community informed of scholarship developments last year, something which was very much appreciated by most concerned. Having the entire committee doing this will make the process more open and having committee members talk to recipients at Wikimania will help justify giving them support to attend. CT Cooper · talk 20:03, 2 January 2015 (UTC)