Wikipedia talk:Article Feedback Tool/Version 5/Help

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Regarding "What happens to my feedback?"[edit]

I think it is important to clarify to users whether feedback they give during the testing phase will ultimately be kept for editors to see, or whether it may be lost. Looie496 (talk) 17:40, 11 December 2011 (UTC)

That's a good point, and something I'm not too clear on myself. I'll find out the answer and stick it in :). Okeyes (WMF) (talk) 18:11, 13 December 2011 (UTC)
The answer is apparently "we're not sure yet". I'll let you know if I find out anything more specific. Okeyes (WMF) (talk) 19:54, 13 December 2011 (UTC)
I think it would be okay to say that we'll try to keep the feedback that comes in during the testing period but we can't guarantee that it will be possible. Looie496 (talk) 22:44, 13 December 2011 (UTC)
  • That'd work! Thanks for the awesome phrasing suggestion :). Okeyes (WMF) (talk) 04:24, 14 December 2011 (UTC)

In Hebrew Wikipedia I have implemented in July 2011 a script for article feedback (he:MediaWiki:HelpLinks.js which is referenced from common.js) which allows adding feedback to article on the talk page. When posting a feedback it kept as a regular edit on the talk page, and adds an empty template that allows this edit to be identified (using abuse filter). You can then see article feedbacks in recent changes (see here). If you are interested how the tool looks like, you can browse a random page in Hebrew Wikipedia and see it on the bottom (autogrowing textarea). Eran (talk) 06:07, 22 December 2011 (UTC)

Thanks for pointing this out, dude! I'll pass it on to the developers :). Okeyes (WMF) (talk) 13:05, 22 December 2011 (UTC)
Similar point - having just noticed one of these feedback boxes on an article on english WP - my first thought was, if i was doing this, where would my comment end up, i mean is it going to be immediately displayed right there where the box is, or added to the talk page in some way? i suppose also wondered whether the comments wlil be moderated. Not sure if anyone had already made a comment via it. Eversense (talk) 23:50, 6 January 2012 (UTC)
So, the details are at WP:AFT5, and are still being discussed (I'm working as hard as I can at getting the community really, really involved) at the talkpage; if you have any opinions on what we're doing, I'd love for you to get involved). In order:
  1. The feedback is going to be displayed on a special feedback page, which we're designing at the moment. It's not currently live - because the tool is being deployed for testing purposes only, so we can work out which form works the best, we don't currently want the feedback displayed.
  2. Whe're still deciding on whether things should be directly syndicated to the talkpage, or dealt with separately, or copied over message-by-message as editors think it's justified, and so on and so forth. If you have any comments on the subject (or on any other subject!) the talkpage is a-callin' ;).
  3. Comments will definitely be moderated; we'll have the Abuse Filter and spam blacklist, plus a variety of moderation tools (both hide and oversight buttons) available to editors. There's no intention of having an unchecked stream of reader comments around. Okeyes (WMF) (talk) 08:32, 7 January 2012 (UTC)
Thanks for taking the time to get back on this here, I'll check out over there too (did before but not the talk). It might help other editors if the test boxes included a little note that for now it's only a test and any comments won't be deployed yet. Every (talk) 03:04, 8 January 2012 (UTC)
  • That's an interesting idea; I don't think we're likely to go for it, but I'll bring it up next time I meet with the guys who make these decisions :). My reasoning is that an "it's only a test" box is likely to substantially alter the nature of who gives feedback and what feedback they give, which completely undermines the purpose of a test. Really we want to keep things as "real" looking as possible. But we'll see! :). Okeyes (WMF) (talk) 06:02, 8 January 2012 (UTC)

How do I get rid of that ^%!@#$!^%$# thing?[edit]

Making people aware that they can play a role in making this a better resource for all is something I favor, in principle. The manner in which this is being done is annoying. How the heck do I get rid of that annoying thing that says "Improve this page" and is constantly blocking something? Besides going through the rigamarole of clicking on it, making a pro-forma response, dismissing the window, and dismissing the next window that pops up, that is. Which is also annoying. Lovibond (talk) 05:00, 20 February 2012 (UTC)

See the "how do I disable it?" section. Okeyes (WMF) (talk) 06:46, 20 February 2012 (UTC)
How long are you going to be testing the floating "Improve this page" tab? Kaldari (talk) 21:27, 26 February 2012 (UTC)
Sent an email out 3 hours ago to find out :). We were meant to turn it off in early feb, and I know there's a planned bugfix that introduces a close button. Not sure what the delay is. Okeyes (WMF) (talk) 23:54, 26 February 2012 (UTC)
That ^%!@#$!^%$# thing has been a pain for ages on a GA article as it appeared to contradict the article's standing. Perhaps if the tag had been on every article then it wouldn't have appeared to be such a judgement. Anyway its gone, gone, gone, yipee. --Rskp (talk) 06:24, 1 March 2012 (UTC)


When will verbal feedback be an option on every page and become accessible to everyone? --Toccata quarta (talk) 06:46, 28 February 2012 (UTC)

If by "verbal feedback" you mean the feedback form with the text box: we're looking for a final, full deployment on 7 March :). Okeyes (WMF) (talk) 12:36, 28 February 2012 (UTC)
Thanks! --Toccata quarta (talk) 13:39, 28 February 2012 (UTC)
No problem! :). Okeyes (WMF) (talk) 17:47, 28 February 2012 (UTC)

General comment about the survey[edit]

I'm not sure if there is much value added in having this do you have any recommendations box for 3 line stubs like Charlie Hoffman. I recommend tweaking this to either only do it for Start class or higher or else by the number of characters on the page (I would suggest excluded categories, Quotations, templates and the like). ShmuckatellieJoe (talk) 03:32, 1 March 2012 (UTC)

Well, surely tiny stubs and start-class articles are (theoretically) the ones where we should be asking for recommendations the most? :). Okeyes (WMF) (talk) 06:10, 1 March 2012 (UTC)
So the trial appears to have ended as the tag has gone from the Battle of Romaniarticle. Hooray. No explanation. It was on a GA article and seemed to contradict that by enticing editors to improve it. I'm sure that this article is not perfect but its also not in urgent need of improvement which is what the tag appeared to be saying. If the tag had been on every article then it would not have seemed to be such a judgement. Anyway, glad its gone. --Rskp (talk) 06:24, 1 March 2012 (UTC)
Er. It still appears on that article for me, and should still be there. I have repeatedly explained what the idea of the feedback tool is. Okeyes (WMF) (talk) 06:33, 1 March 2012 (UTC)
My comment wasn't meant to offend anyone so sorry if I did. I only meant that to me, the fact that an article only has 3 or 4 lines should make it clear that the article needs to be expanded. Although I can see where its useful to know what articles people are looking at so they can be improved. ShmuckatellieJoe (talk) 12:45, 1 March 2012 (UTC)
Exactly :). Plus, we don't just expect (and we haven't just got) "this article needs expanding"; we're looking for "why doesn't this article include information about X?", where X is something specific. So even with short articles, it can be useful by pointing to the direction it needs to be expanded in. And you didn't offend anyone - I was replying to Roslyn :). Okeyes (WMF) (talk) 17:15, 1 March 2012 (UTC)

The 'Improve this page' thing needs to go[edit]

The 'Improve this page' link in the bottom right corner is an annoying, distracting, and confusing piece of clutter. It needs to go... badly. Jason Quinn (talk) 05:17, 9 March 2012 (UTC)

Have you read the help page? Okeyes (WMF) (talk) 06:36, 9 March 2012 (UTC)
By "help page", do you mean Wikipedia:Article_Feedback_Tool/Version_5/Help? Yes, I have read it. Are you suggesting I turn the AFT tool off? (I don't like having to guess what you mean by your reply but that is my interpretation.) I am not interested in turning the article feedback tool off. I like to keep my interface similar to the default. My complaint is that the default way (with AFT on) is producing a poor user experience. My top level comment was talking about a specific aspect of the tool that shouldn't be there in the first place, the "Improve this page" CSS element anchored in the right-hand corner that, A) covers up article text and and, B) violates the generally "static page" look-n-feel of Wikipedia. I have reservations about the AFT in general but version 5 is acceptable to me except for this component. It shouldn't be there.
In general, I think suggesting to editors who complain about the AFT that they should turn it off (which seems to be prevalent) is a wrong approach. It is systematically silencing those rare editors who are concerned enough to comment about things like the AFT in the first place. Here a choice between the V5 tool with or without the "Improve this page" thing is under discussion. Maybe AFT V5 is better with it or maybe it's better without it. I think it would be better without it. I think this is obvious as per the reasons above. Encouraging me to turn off the tool altogether is completely side-stepping the question at the heart of the issue. Jason Quinn (talk) 23:51, 9 March 2012 (UTC)
Argh, sorry; I thought I'd updated the help page (must have slipped my mind :S) to make clear we're introducing a way to turn off the "improve this page" thing on its own ASAP. I'll go do it now.
Now, personally, I don't like "improve this page"; I think there are much more elegant ways to increase the prominence of the box and the amount of feedback we get. But there is some genuinely fantastic research we're doing with it that has the potential to completely change how we approach stuff like this (I'll give details as soon as we have more data - right now I don't want to raise hopes!) and so there is a good reason we keep mucking around with a docked button. Hopefully the ability to switch it off on its own will be introduced very soon. Okeyes (WMF) (talk) 00:05, 10 March 2012 (UTC)
Thanks for the reply. I'll glance at the research results when they are published. I admit I'm skeptical that it will change my mind. Past discussions on the AFT tool voting results and other issues like the "gender gap" have sort of undermined my faith. As for an option to turn "Improve this page" off, in the long term, that option should only exist if the studies definitively show the "button" should be there in the first place. I worry already that some overly simplistic quantitative statistic like click-rate will be used to justify its existence. Of course, having that "in your face" button is going to cause people to click on it. But if click rate is all that matters, might as well make the entire page a giant blinking "Improve this page" button. My point is that a simple analysis would overlook completely the qualitative effects such as I mentioned above. My past perception of WMF studies is that they are interpreted in the most straight-forward way possible, often to suit some preconceived notion while counter interpretations are ignored. Maybe I'm too cynical. Jason Quinn (talk) 18:00, 10 March 2012 (UTC)
We don't actually measure click rate at all; it'd have privacy implications. But the data we're using is varied. At its most base level we've got the number of people who provide feedback, as well as the quality of that feedback. Then we've got the number of those people who attempt to edit (which, really, is secondary). Then we've got the number of people who complete that edit (also secondary), and then we'll have data on whether the presence of the box has an impact on editing on the page as a whole (maybe it reduces anonymous edits? We need to find that out). If this has a potential impact on how readers interact with us that we can test, we are doing our best to do so. Okeyes (WMF) (talk) 19:10, 10 March 2012 (UTC)
I also find the "Improve this page" link quite annoying. I suggest adding the following to Wikipedia:Article Feedback Tool/Version 5/Help#Why is there an "improve this page" link?:

Registered users can remove the link without disabling the whole feedback widget by adding this line to Special:MyPage/common.css (applies to all skins) or Special:MyPage/skin.css (your current skin):

PrimeHunter (talk) 23:15, 21 April 2012 (UTC)

This is superfluous. There is a talk page.[edit]

Educate the people to use this talk page. Furthermore what happens to the feedback? I will always only look for feedback on the talk page. If there is another feedback source somewhere else beside the talk page i promise i will not look at it. The talk page is a simple and good centralized way to leave feedback - and it is the only one that i accept. You may work on precenting the talk page or give easier access to it by a "feedback tool" but the talk page is the only page i will look for feedback for my articles.-- (talk) 11:50, 13 March 2012 (UTC)

You're perfectly welcome to do that, but it's simply not feasible. Most readers cannot get wikimarkup, and we should not demand they do so. We could totally take the comments and just feed them to talkpages - but have you seen the volume? On high-profile pages it is simply not feasible. Okeyes (WMF) (talk) 00:07, 14 March 2012 (UTC)
This tool produces redundancy. When you look at a talk page you see whether the "problem" has already been adressed or not. You read before you add a new comment. But with this feedback tool you just set a comment regardless whether there is already an ongoing discussion. This leads to redundancy - this is why i do not support adding the comments blindly to the talk page. The tool is just superfluous. PS: there is a project, that works on replacing the wikimarkup ( - there is really no need for this tool here. Please help on these WYSIWYG projects instead of producing data that are 1) very possibly redundant, 2) not accessible in a good way. — Preceding unsigned comment added by (talk) 13:06, 15 March 2012 (UTC)
Sure; we're working on a visual editor now. The expected completion date is some point in late 2012, assuming nothing goes wrong. In the meantime, this provides a decent mechanism readers can use to leave feedback, and corrects deficiencies with talkpages along the way. Okeyes (WMF) (talk) 15:52, 15 March 2012 (UTC)
Hey, i dont want to annoy you. I am not against you as a person - but against this project, because... see above ;)-- (talk) 12:53, 16 March 2012 (UTC)
@Okeyes. Is there a study that shows "most readers cannot get wikimarkup"? What demographics are these people? Are they people that we really want to be contributing? The people who cannot get markup are largely going to be children and, well, dumb people. If the goal is to create a high quality encyclopedia, perhaps things are better off without their contributions. Perhaps the barrier of wikimarkup self-censures Wikipedia in a beneficial way. A natural outcome of a WYSIWYG editor seems to me that vandalism will increase and the quality of the prose (and source) will be lower. Our slogan is "the free encyclopedia that anyone can edit" but that doesn't mean the encyclopedia that everybody should edit. It seems to me that WMF has always put openness on a pedestal (largely to satisfy big donors) but neglects to consider the quality impact that openness initiatives will have. ALL initiatives should study the quality impact they will have. If it's true that a WYSIWYG lowers the quality of articles, it should not be used. By no means am I saying that the WYSIWYG is not an interesting project. It definitely is. I am raising a legitimate concern about quality. Actually, I am more concerned about institutional behavior that seems to ignore such questions because of their slightly non-PC overtones. Talking about openness and blah blah blah is great bureaucratic-level thinking but it often glosses over the side-effects of openness. As any social media site from Digg, Reddit, Slashdot can tell you, with greater participation, the quality of user-content tends to go down. I submit that as a hard fact and one that should be investigated for the WYSIWYG editor. The truth is that the barrier to participation should be high enough that it "weeds out" people who would tend to contribute noise. Jason Quinn (talk) 21:45, 22 March 2012 (UTC)
The usability initiative highlighted markup as one of the things that people struggle with the most. I agree that there are many concerns with the impact WYSIWYG would have on the community and on quality - we are taking them seriously (and there is, internally, a debate as to how the heck we even go about switching it on without making everything blow up).
I do not agree that "The people who cannot get markup are largely going to be children and, well, dumb people". In fact, I find it offensive. Over the last few years, the nature of the internet and of users' expectations of it has changed, fundamentally. One of the largest changes is that people expect some form of WYSIWYG editor. Wordpress offer it, Facebook offer it, pretty much any site that allows user contributions to the corpus allows it. We are an exception in that we do not. That's both something that renders us unique and something that renders us increasingly out of date; internet users do not, by and large, expect to have to learn a markup language to contribute. No other site, really, requires them to. And so instead of the pool being "children and dumb people", it's "children and dumb people, and anyone who is intelligent and wishes to contribute but thinks, on the face of it, think that learning an entirely new pseudo-html language to correct a typo or add a citation is stupid as hell". And for the record, I agree with that group of people. In 2006 when I joined, I'd grown up as part of a generation with dial-up connections and html as a requirement to do pretty much everything. Learning markup wasn't a big leap then. If I'd grown up in the "facebook generation" - as part of the group of people who are used to WYSIWYG, to low contribution barriers - I can't say I'd be around. Okeyes (WMF) (talk) 18:01, 23 March 2012 (UTC)

──────────────────────────────────────────────────────────────────────────────────────────────────── My assertion is either true or it isn't. It should be studied. I could have easily dressed up my comments in formal language but they would have been expressed much less concisely. I find it tiresome that you would claim offense. I wonder how much of it is feigned simply because you felt the need to reply thusly in your capacity of a WMF official. Regardless, your reply mostly sidesteps the issue I raised about quality. Note that basically nothing of what you wrote supports the premise, "A WYSIWYG editor will increase the overall quality of Wikipedia articles" or even the weaker "A WYSIWYG editor will not lower the overall quality of Wikipedia articles". That is the issue at hand. It would be more interesting to learn about the quality research you mention. I also fail to accept to premise that wikimarkup is so obtuse. The basics of wikimarkup can be mastered in five minutes. Compared to the controls in a modern video game, wikimarkup is 100x easier to learn. Hell, an editor doesn't even need to know wikimarkup, people can and do create articles using just plain text. I am not against the idea of the WYSIWYG editor. It will be an interesting experiment. I am claiming that it is not necessarily a good thing from the viewpoint of quality. [Note that this is not the same as saying the editor will be bad for quality.] I worry that poor bureaucratic thinking will falsely conclude having a WYSIWYG editor will make the encyclopedia better simply because it makes editing more accessible. That is a non-sequitur. Your comment seemed to show you accept it as a given. What's important is the ratio of "good" edits-to-"bad" edits that will result from its introduction. I fear that this notion will not be properly researched by the WMF because everybody is too lily-livered to entertain it. Ironically, if the debate is worded in terms of increasing "expert" contributors instead of discouraging "stupid" contributors, everybody is all gung-ho; yet, they are two sides of the same coin: trying to increase quality. It's a rather interesting quirk of human psychology. Jason Quinn (talk) 17:03, 28 March 2012 (UTC)

  • So, in reverse order; this is something we're going to study and that we are worried about, as mentioned above. And as also mentioned above, we're getting data on the impact prominence makes on quality, too; a second round of the study begins this week :). A first round actually concluded that improving the prominence and ease of input methods had a positive effect on the quality of submissions, but it was a very small impact (albeit a statistically significant one). We're running a second round to see if it was just a fluke. You are welcome to wonder what you will; as both an editor and as a contractor, I think your initial statement is problematic. If you read the usability studies you will see exactly how easy or hard people find wikimarkup - people of a wide variety of professions, ages and levels of "real world" experience. Generally I think the biggest issue we have as a community is with perspective; from our perspective, sure, markup is easy. But my perspective is as someone who has been using wikimarkup for six years - yours is as someone who has been using it for eight, and back it up with fluency in multiple programming languages and a DPhil degree. Neither of us are really in the position to start saying "newbies will find this hard/not hard", because it's been a good while since we were newbies, and because our perspectives are not likely to line up with those of the vast majority of new editors, or even the vast majority of new editors capable of making high-quality contributions. The people qualified to make those judgment calls are those who participated in the usability study, which I would urge you to read. Okeyes (WMF) (talk) 18:03, 28 March 2012 (UTC)

Using feedback[edit]

I'm just curious; if somebody gives feedback about an article that I am helping to support, where can I find the feedback, or is it not publicly available yet? Thanks, Falconusp t c 13:07, 28 April 2012 (UTC)

At the moment we're still building the page, so it's very hush-hush. But we expect a more public release in a few weeks :). Okeyes (WMF) (talk) 05:10, 29 April 2012 (UTC)
Alright, thanks. I'll be looking out for it :-). Falconusp t c 06:52, 29 April 2012 (UTC)
Awesome! Okeyes (WMF) (talk) 15:23, 29 April 2012 (UTC)
I had the same question as the user above. I'm excited to see the final product. Good luck. ~Adjwilley (talk) 21:03, 31 May 2012 (UTC)
It's going to be good to release it! On the down-low; Special:ArticleFeedbackv5/whateveryourarticletitleis. Give me a poke if you need any help or have questions :). Okeyes (WMF) (talk) 21:16, 31 May 2012 (UTC)
Thanks. It looks like it's going to be great. If there's anything at all that a lowly editor like myself can do to help, do let me know. ~Adjwilley (talk) 22:54, 31 May 2012 (UTC)
Poke it with sticks, tell us if anything breaks as a result (and what you were doing to make it break), and offer any suggestions you have :). And hey, most of the time I'm just a "lowly editor" too. We're the guys who make this place tick :P. Okeyes (WMF) (talk) 00:52, 1 June 2012 (UTC)

Newer AFv5 vs. Older AFv5[edit]

I noticed that this AFv5 doesn't have "suggestion, praise, problem, question" and is much simpler. I currently have the newer AFv5 installed, but I want to change it to this simpler one. Is there a way to do that? — Preceding unsigned comment added by (talk) 18:39, 7 August 2012 (UTC)

Hmn; I'm not sure! I would recommend emailing; he's our dev on this :). Okeyes (WMF) (talk) 15:59, 9 August 2012 (UTC)

Who is this "Article Feedback V5" guy?[edit]

In the "activities log" to one particular feedback post, I just happened to come across the entry "Article Feedback V5 marked this post as helpful". Normally, the subject of such an entry is a named user, who has judged a post as helpful or unhelpful. So, who is this "Article Feedback V5" guy, and why has he got no user page but a link to the tool's general help page instead? (Alternatively: how on earth does the tool itself determine whether a post is helpful or not?) Fut.Perf. 22:02, 14 August 2012 (UTC)

Update: I've now seen activities of this type on multiple other entries, and they appear to be pretty much random – flagging harmless comments as abuse, marking the same comment alternately as "helpful" and "not helpful" several times in a row, and so on. What's going on here? Fut.Perf. 06:47, 15 August 2012 (UTC)
Okay, I now found that this is what happens when an IP editor rates the feedback post of somebody else. A rather annoying bug. Where can I report it so it can get fixed? Fut.Perf. 10:25, 15 August 2012 (UTC)

Broken link[edit]

In Wikipedia:Article Feedback Tool/Version 5/Help#Can I enable/disable AFTv5 on individual articles?, there is a link "small subset of articles", piped to "m:Article feedback/Samples". That link leads to an empty location, with only a deletion notice of a recently created "test page" on Meta. Fut.Perf. 22:06, 14 August 2012 (UTC)

Empty entries[edit]

The article feedback tool shouldn't save empty replies, or replies full of repeated characters or conspicous (algorithmically-detectable) "filler".--Wtshymanski (talk) 16:02, 20 August 2012 (UTC)