Wikipedia talk:Content assessment/Archive 1

Questions
Some questions. Jdorje 19:27, 11 January 2006 (UTC)
 * Where is discussion on these classes done?
 * How do articles that don't fit in the normal set of classes get assessed? For instance disambiguation articles or category articles probably shouldn't be assessed on this scale.  It's easy to say that they just shouldn't be assessed but I would like to be able to categorize them somehow.


 * The assessments themselves are done on individual project pages such as here. As for discussing the grading scheme itself, this talk page here is where discussion should take place.  The scheme evolved over at WP:Chem and has worked very successfully over there for several people regularly assessing 380 articles over 7 months.  Since I am a great believer in using a system that works well in one place and seeing if it can work elsewhere, I proposed its use in WP 1.0 here, and that's how it came to be used.


 * I read some of the comments over on WP Tropical cyclones and your assessment page. Your group seems to have an excellent grasp of the issues, and of what we intended for the assessment criteria.  User:Maurreen (the founder of the WP1.0 Editorial Team) and I have had a talk about a possible C-Class between B and Start, or about the renaming of start and stub as letters, but we didn't come to a decision.  My original scheme for WP:Chem was based on A,B,C,D,E, but the others on WP:Chem revised that to the current system.  I suspect this was in order to get away from the idea that we were assessing primarily quality, because in fact these "grades" reflect mainly completeness rather than quality (though hopefully the latter is a factor!).  The system is now used in several places as is, but if changes are definitely needed, particularly for levels below B-Class, that would be feasible now (probably not in a few months time!) I would say:


 * If you and others at WP Tropical cyclones feel that a revision is urgently needed, propose it formally here and we'll discuss it.
 * By all means adapt the scheme to your own field, create different "grades" if that helps the assessment process on a particular topic.


 * Finally, regarding dab pages (lists are another thing, and things like "Science events of 1999"), the standard assessment scheme doesn't apply. Maurreen has spoken of having an "atlas" within Wikipedia 1.0, clearly this is also to be evaluated differently.  As we move forward on WP1.0, I'm becoming aware that we will need all of these things as well as new organisational material; currently these "non-articles" simply fall outside of the assessment process.  However their quality and completeness will need to be assessed - if you want to make proposals on this also, I'd appreciate this. Thanks for your trenchant remarks. Walkerma 22:48, 11 January 2006 (UTC)


 * Interesting. Currently for tropical cyclone articles we use Stub-Start-B to distinguish content and B-A-FA to distinguish length.  So once an article has sufficient content to be considered "complete" we mark it as B.  Then it's a matter of improving the quality (including more content, of course) to get it to A and then FA.  &mdash; jdorje (talk) 08:56, 19 February 2006 (UTC)

Ok, I have another Question. *Hand shoots up* Who goes to a "Discussion" page for info, unless they wish to discus the info given?

Lists
How do Featured lists fit into the assessment criteria? There is no featured list which is also a featured article, but they all meet the featured list criteria (that is, they have a good lead, are useful, comprehensive, factually accurate, stable, and well-organised, and have references, etc) so I would say that they all are in Class A. Comments? -- ALoan (Talk) 15:03, 10 February 2006 (UTC)


 * This question was also raised here recently. We don't have a separate template for featured lists in tables, but we use the FA list.  For example the element folks have a couple of "FA" lists, see our listing here, and the isotope people have a couple of "A-Class" lists on the table below that one.  This issue still formally needs to be resolved, but I'm almost certain we'll include plenty of lists in 1.0. We'll update the assessment criteria accordingly, hopefully very soon. Cheers, Walkerma 16:31, 10 February 2006 (UTC)

Examples
While listing the examples, it would have been more helpful if it was mentioned which version of the article it was referring to. Antarctica has been edited 300 times since it was put up as an example of a 'B grade' article. Tintin (talk) 00:31, 11 February 2006 (UTC)


 * Changed the example links to specific revisions of that time. Femto 12:15, 11 February 2006 (UTC)

Another take on color coding of classification
see WikiProject_The_Beatles/Article_Classification... interested in comments and feedback as well as seeing if unification makes sense? Ours are templatised so are easy to change around. + +Lar: t/c 04:26, 3 April 2006 (UTC)


 * I don't have strong views about colors, myself, as long as they are clearly different. I like the Beatles colors because you can see them going from red (bad) to green (good).  However I think the A-Class and FA-Class colors are a bit similar.  Any other views?

GAs
I'd like to suggest adding a level for designated good articles. Maurreen 02:46, 7 April 2006 (UTC)
 * That is an excellent idea. Any chances of making WP:GA a formal guideline? ≈ jossi ≈ t &bull; @


 * I'm not sure what you mean. Maurreen 03:39, 7 April 2006 (UTC)


 * I'd be happy to see GAs listed as such in our tables, though do you think we should wait till GA becomes policy instead of just a proposal? I think the system is still evolving somewhat, that's all.  I would also make the case that we should not scrap A-Class; in effect A-Class becomes what an assessor believes to be (roughly) equivalent to GA.  We and others can then easily see articles that might be submitted as GA candidates.  By the way, sorry I've been quiet, but things have been crazy at work lately. Walkerma 04:00, 7 April 2006 (UTC)


 * I understand. :) And I wasn't meaning to scrap the A class, maybe just divide it. Maurreen 07:54, 7 April 2006 (UTC)

So now there's a separate GA-Class? And it's *below* A-class? That does not make sense to me. Once a GA is upgraded to A-class it loses it's GA template and then gets nothing in its place? — jdorje (talk) 18:14, 29 April 2006 (UTC)


 * It shouldn't be losing the GA template until it becomes featured, no? These assessments aren't meant to replace other Wikipedia article recognition programs. Kirill Lok s  h in 19:31, 29 April 2006 (UTC)


 * Not at all! We keep our existing A/B/S/S system for WP1.0, but we note if something is also a GA.  We are regarding the GA review is an independent review, not as a replacement for our own reviews.  We have come to the conclusion (elsewhere) that GAs are very variable in quality, and so we can't make simply equate A=GA as I suggested above. Look at the table for core topics to see this. Walkerma 19:39, 29 April 2006 (UTC)


 * I see your table describes GAs as sort of 'only slightly deficient' articles that other encyclopaediae could do better. Although many of them are such, many of them are excellent short articles that broadly meet FA criteria but are too short to be realistically featured, and so are probably rather better than your description here implies.  I'm trying to resolve this dichotomy now - you might be interested in the discussion going on at Wikipedia talk:Good articles.  Worldtraveller 10:42, 12 May 2006 (UTC)


 * Quite right, I will fix this. I was looking at the discussion yesterday - it's hard to resolve, with such a variety of articles, we have a huge range at the B-Class and Start-Class level too.  "How long is appropriate" is impossible to answer! Thanks, Walkerma 15:49, 12 May 2006 (UTC)

"Stub Class"
Any chance someone working on this could clarify the difference between a "stub" and a "Stub Class" article, if any? If they're different, this is a terminological train-wreck, and a likely to cause all manner of confusion down the way; if they're essentially the same, the relationship between assessing the grade on the one hand, and bog-standard stub-sorting on the other should be clarified. If an article self-classifies itself in one way, and the associated talk page makes some seemingly-inconsistent grade assessment (and puts it in a category with some similar but distinct-but-overlapping scope), heads are going to be scratched. Alai 00:51, 10 May 2006 (UTC)
 * By most means, a stub, as sorted by WP:WPSS, will be assessed as a stub in our assessment scale. The only difference is that a long article of inadequate quality may not be marked as a stub by the Stub Sorting WikiProject, but it will be marked as a stub by us if it is basically useless. Tito xd (?!? - help us) 01:30, 10 May 2006 (UTC)
 * There's no maximum length defined by the stub guideline, so that's by no means clear. Where in any case is the 1.0 definition that makes this explicit?  Given your comment on SFD, I rather anticipated your answer on this, but as I say above, this clearly needs to be clarified one way or another.  Alai 01:34, 10 May 2006 (UTC)
 * From this page's companion: "The article is either a very short article or a rough collection of information that will need much work to bring it to A-Class level [almost identical to the Stub guideline]. It is usually very short, but can be of any length if the material is irrelevant or incomprehensible [emphasis mine]." Tito xd (?!? - help us) 01:38, 10 May 2006 (UTC)
 * I'd read that, I was looking for something a little more... definitional.  Anyway, points and questions still stand.  Alai 01:49, 10 May 2006 (UTC)


 * I think these things differ principally in their purpose. If your focus is on stub-sorting, you can ignore the assessment categorisation, and vice versa.  I don't see any need for head scratching.  In the few cases where there are differences, most people won't even notice, and those who do may double check to see if "stub" is appropriate, thereby helping both activities.  I expect that by the end of 2006 most Wikipedians will be familiar with this automated assessment system, and many will be using it, because of the great benefits it brings to WikiProjects and WP:1.0. Walkerma 02:20, 10 May 2006 (UTC)


 * It appears this terminology and separate classification is being defended alternatingly on the grounds of being the same, and of being different, and I see nothing to change my original contention that this is as clear as proverbial mud. Given that this is flagged as a "proposed" system, it seems remarkable that the suggestion of renaming or removing a "grade" meets with such entrenched conservativism, and insistence that it'll be imposed on everyone else, especially when it's not even clear what the purpose of said class is, given the similarity in how the bottom two are "actioned" ("major editing" vs. "almost any editing"). If it actually is necessary, and its meaning is simply "useless", and not "stub", why not "useless grade", "placeholder grade", or some other less overloaded synonym?  Alai 16:49, 12 May 2006 (UTC)


 * The meaning of the two is similar, but the purpose of them is different. Walkerma 18:56, 12 May 2006 (UTC)

Question
How does an article/group of articles get nominated for assessment by the Editorial team? I have noticed that nearly every article in Category:Tropical_cyclones has been assessed, but I have been unable to find one article in general meteorology which has a ranking on the assessment scale. There are several articles I have in mind for assessment&mdash;can I nominate them somewhere or does the team go in some type of order? (posted by User:Runningonbrains)


 * Several answers needed here!


 * The cyclone assessments have been done by members of the project themselves, which is of course how it should be. The WP1.0 team members typically don't have the specific expertise needed for good assessments on specialised articles. Several projects have done similar things, but mostly they have just come up with a handful of decent and/or important articles (still very helpful).
 * The cyclones group is unusual in that they are also participating in testing out a "bot", you can see all the results here.
 * We have contacted all the WikiProjects that were on an "active list" in October 2005, and meteorology was contacted here in April.  Based on replies we fill out a table such as this table for Meteorology.  Often people are a bit reluctant to assign grades to the articles, so in that case team members do their best to look at easy things like layout, references, etc.
 * We will be re-contacting all the projects over the spring/summer, in order to request lists of key articles, and to encourage use of the bot for tracking assessments. The bot will make it easier both for the project and for us at WP1.0 to keep track of thousands of articles.
 * On the subject nominations, we will be taking nominations very soon at WP:V0.5N, I hope you and others will nominate some of the important meteorology articles!
 * If you want to start using the bot straight away, reply here and we'll get you up & running. Thanks for contacting us, Walkerma 19:13, 23 May 2006 (UTC)

Antarctica
Antarctica is here given as an exapmle of a B-Class article and lists problems with it. However, it is a Featured Article. I am confused--Freiberg, Let's talk!, contribs 21:44, 1 June 2006 (UTC)
 * Right now it is featured, but at the time the snapshot included in the template was taken, it was not featured, and had some problems with it. Tito xd (?!? - help us) 23:57, 1 June 2006 (UTC)


 * I've been thinking that maybe we should update our examples to limit such confusion - thankfully all of them have improved since the example was taken! Thoughts anyone? Walkerma 01:26, 2 June 2006 (UTC)
 * Link to the specific revision number, not the article name. Simple.Stevage 08:09, 14 June 2006 (UTC)
 * That is the way it is at the moment. Tintin (talk) 10:11, 14 June 2006 (UTC)

Editor's experience of featured articles
The claim that the editor's experience of featured articles is that "No further editing necessary, unless new published information has come to light" is simply preposterous, and is one of the main reasons I consider the entire Featured Article concept to be misguided. I have never seen a featured article that couldn't stand to be improved on, whether for style or content. However, any attempt to do so is immediately reverted by the article's "protectors" who actually believe the above statement. The sentence "Once an article reaches the A-Class, it is considered "complete", although obviously edits will continue to be made" is just as bad — in fact, no article can ever be considered "complete". Science is never finished. There is no area of human endeavor in which all questions have even been asked, let alone answered. Whoever thought it would be a good idea to release a paper and/or CD version of Wikipedia was simply being extremely short-sighted. User:Angr 12:47, 19 June 2006 (UTC)

"Start"

 * The article has a meaningful amount of good content, but it is still weak in many areas, and may lack a table.

It may lack a table? What? I know including a picture, no matter how uninformative and useless, is required to get an article to FA status, but since when is a table an indication of anything? I'm sure some articles that should have tabular data lack it, but as a general measure of quality it looks silly. 82.92.119.11 16:50, 20 June 2006 (UTC)
 * I think "a table of contents" is what was intended. User:Angr 18:01, 20 June 2006 (UTC)
 * Aha! Yes, you may be on to something there.
 * TOCs are generated automatically, so that would mean the article is either too short or not properly sectioned. Now that I could see as a criterion. 82.92.119.11 18:06, 20 June 2006 (UTC)
 * But not a significant one. Almost all 'Start'-class articles have a TOC. Saying that Start-class articles "may lack a table" is not only potentially confusing, but also not really relevant or accurate. It's too trivial and obscure a problem to be worth mentioning. -Silence 20:12, 20 June 2006 (UTC)

Question
It seems to me that the "stub-class" is actually two classes in one? Shouldn't the short article, and the less than totally usable classes be separate? Say, a C-Class? - just some of my thoughts. 132.205.45.148 01:31, 8 July 2006 (UTC)
 * Not all short articles are useless, though. Short, but useful articles fall under the traditional definition of a stub, though. Tito xd (?!?) 03:43, 8 July 2006 (UTC)
 * I meant that not-short articles seem to be classifiable under stub, from the description of what "stub-class" is. Sorry for any confusion in my query. 132.205.95.25 20:44, 8 July 2006 (UTC)
 * Yes, they indeed can, if they are useless. Short articles, that can be stubs, can be assessed as a higher-class article if they're more useful. Tito xd (?!?) 20:46, 8 July 2006 (UTC)

Question
"Classes lower than this" ... would be? IS there a D-class or an F-class? 132.205.45.148 01:33, 8 July 2006 (UTC)
 * Where's that? Tito xd (?!?) 20:46, 8 July 2006 (UTC)

Tone and consistancy
Can I question the tone of some of the text on this template. Also the consistancy between the Criteria and the other columns. this is particularly true in my view on the "Start" class

Reader's experience
Not useless.
 * surely a might negative - combative in fact - some editor's take offence at this type of thing (and have).

Some readers will find what they are looking for, but most will not.
 * surely "other may not" would be a better phrase.

Most articles in this category have the look of an article "under construction" and a reader genuinely interested in the topic is likely to seek additional information elsewhere.
 * I am myself unclear on how to improve this but again the tone is in my view unhelpful. I'll think a bit more about this clause.

Editor's experience
Substantial/major editing is needed, most material for a complete article needs to be added.
 * Would not "more material for a complete article" be better.

This article usually isn't even good enough for a cleanup tag: it still needs to be built.
 * What a put down, this is not an encouragement to editor's efforts.


 * Sorry I see what is being attempted here but there is a disparity in experience between "Criteria" whic is good "stuff" and these other columns. :: Kevinalewis  :  (Talk Page) / (Desk)  12:27, 11 July 2006 (UTC)

Response
You make some good points. But I don't know how to edit the box. Maurreen 15:31, 11 July 2006 (UTC)
 * I would "Be Bold" but I thought that with such a key document caution was more in order. If there is agreement on content I would be happy to implement the changes. Tables are not a problem for me. :: Kevinalewis  :  (Talk Page) / (Desk)  15:34, 11 July 2006 (UTC)
 * I appreciate your asking. No one has disagreed, so please go for it. Maurreen 12:39, 12 July 2006 (UTC)


 * Very helpful, thanks! You are wise to proceed with great caution, the Start tag is already applied to 6,729 articles that we know about!  We should probably see if User:Titoxd has any comments, too.  You are quite right about the general tone, though I don't agree with everything you mention - it needs to remain distinct from a B-Class.  It should certainly be written from a viewpoint of a glass half full, not a glass half empty.  Some detailed comments:


 * In place of "Not useless" (reminds me of "mostly harmless"!) and the rest of the reader experience, how about "Useful, provides a moderate amount of information, but many readers will need to find additional sources of information. The article clearly needs to be expanded."
 * For the "editor's experience" section, changing "most" to "more" is too vague and it also makes it sound too much like a B. You can think of most Stubs being <10-20% complete, most B-Class being >70-100% complete (depending on breadth of topic, POV, language quality etc), and Starts fill in the ground in between.  We could perhaps change "most" to "much" without hurting things too much (though I agree it's also a bit vague).  That would be more appropriate for those Starts that are close to B-Class.
 * "This article usually isn't even good enough for a cleanup tag: it still needs to be built." I found this so upsetting I took the liberty of changing this to read, "This article still needs to be completed, so an article cleanup tag is inappropriate at this stage." I understand what was meant was not insulting, but the "not good enough" part comes across badly.  Please feel free to discuss this wording further, but I didn't want to leave such a negative phrase up any longer.
 * What do you think, Kevin et al? Walkerma 17:52, 12 July 2006 (UTC)
 * I'm not sure. There needs to be some sort of "kick in the keister" phrasing, or otherwise editors may not feel encouraged to make edits to an article. At the same time, the language does not need to be unnecessarily harsh. Tito xd (?!?) 18:44, 12 July 2006 (UTC)


 * I think Walkerma has good suggestions overall.
 * But I lean toward deleting the cleanup reference entirely. No. 1, I think a cleanup tag could apply to a short or very short article (although if it's tiny, it might just be easier to do the cleanup than put the tag on). No. 2, my understanding, fuzzy as it is, was that "Start" class was meant to indicate length and breadth more than quality. But maybe my understanding is wrong.
 * As an aside, the "Start" and "Stub" labels aren't clear about how to classify items where other quality aspects are more of a concern than the amount that the article covers. Maurreen 15:30, 13 July 2006 (UTC)

Request for clarification
I note the descriptors for FA class:


 * Definitive. Outstanding, thorough article; a great source for encyclopedic information.

and


 * No further editing necessary, unless new published information has come to light.

Do all FAs qualify for this category, and if so, do these descriptors purport to apply to all FAs? Many, even most FAs require further editing, and are not outstanding" by any stretch of the imagination.

By locial inference, does this mean that an article that is not an FA is necessarily in need of minor edits, as the descriptor for A class says?

Tony
 * What it means is that the articles that qualify for the FA class are any and all that went through WP:WIAFA and passed, and as a result, need no major editing unless significant new information comes to light. (Of course any article will need minor edits every now and then.) The current language was adopted from there. If FAs are not outstanding, that's a discussion that should be done at WT:FA, or WP:FAR, as it is more appropriate there..
 * As for an A-Class article, it is pretty much equivalent to FA-Class, yet it is almost a sure fact that an article that is sent to FAC will receive a few more significant edits as part of the process. The key phrase there is that the article has a passing chance at FAC in its current shape. Tito xd (?!?) 18:40, 12 July 2006 (UTC)

The example chosen of the ideal FA-class article
To take the previous matter further, I note that Medal of Honor is displayed as the example of FA-class articles, next to the statement that these articles require:

"no further editing ... unless new published information has come to light."

I wonder why, then, a cursory reading of the first few paragraphs of the article revealed the following problems.

*"All branches of the U.S. military are eligible to receive the medal, though each branch has a special design." "Though" in wrong here, since it doesn't contradict the preceding clause.

*"The Congressional Space Medal of Honor is a separate award and not equivalent." Insert "is" before "not"; "equivalent" to what needs to be explicated for ease of reading.

*"Scott did not approve the medal; however, such a medal found support in the Navy." Either "approve of the medal" or "approve the proposal" is required here, whichever conveys the intended meaning.

*"In the rare cases (19 so far) where a service member has been awarded more than one Medal of Honor, regulations specify that an appropriate award device will be centered on the MOH ribbon and neck medal." The parenthetical phrase would be less intrusive if place after "Honor" (i.e., before the comma). Remove "will".

*Stubby, one-sentence paragraphs, including one in the lead and quite a few further down.

I haven't read the whole article, but these problems suggest that it needs a close copy-edit if it's to continue to be held up as a shining example. I'm listing it for a Featured Article Review to provide an opportunity for this to be done.

A related issue is my slight discomfort that the topic may be construed by some readers, both within and outside the US, as too culturally and/or politically sensitive to serve as an example of the pinnacle of Wikipedia's achievements.

Tony 16:14, 13 July 2006 (UTC)


 * Tony, you know the FAs better than most of us. Can you suggest a good example? It's easy to change this example, since no one can debate whether or not an FA is an FA.  Thanks, Walkerma 22:33, 13 July 2006 (UTC)
 * Sanssouci is very nicely written, but may be a little image-crowded. What do you think? Tony 02:28, 14 July 2006 (UTC)


 * Support: Looks nice to me, though I may be accused of bias as I hale from Potsdam, New York! We'll see if anyone here objects, then change it. Thanks, Walkerma 02:50, 14 July 2006 (UTC)
 * Just by looking at the references, they're not properly formatted. Any other examples? Tito xd (?!?) 04:15, 14 July 2006 (UTC)
 * Everyone has their own ideas about how references should be formatted - what do you think is wrong with the reference of Sanssouci? Presuambly such a small technical matter is easily fixed? -- ALoan (Talk) 11:20, 14 July 2006 (UTC)


 * How does Sikhism look to you? I just reviewed it for 0.5, it certainly "felt" like a very nice article to me. The English wasn't absolutely perfect, but was generally a high standard. It may not be good to have a WP:FAR article as our example. Walkerma 03:58, 20 July 2006 (UTC)

Let me ask one question
Are these assessment tags meant to be placed on article Talk pages? --WikiCats 14:57, 17 July 2006 (UTC)
 * if you look at the article Medal of Honor you will see how it is done. And here Talk:Medal of Honor :: Kevinalewis  :  (Talk Page) / (Desk)  15:34, 17 July 2006 (UTC)


 * Yes, obviously. It really doesn't work any other way (we've tried doing lists by hand, and they simply become unreadable past a few hundred articles). Kirill Lokshin 12:42, 18 July 2006 (UTC)

Next question. Who chose the terms quality, importance and class? --WikiCats 14:34, 18 July 2006 (UTC)
 * Importance should in my view be Priority as a less loaded term. And more clearly linked to a specific task rather than an overall statment about the article. :: Kevinalewis  :  (Talk Page) / (Desk)  14:39, 18 July 2006 (UTC)

At last, someone with some insight into this issue!

My fear is that there was no debate into what these terms should be. --WikiCats 14:52, 18 July 2006 (UTC)


 * It depends on the context. These templates are generally being used by Wikprojects in conjunction with a subproject of the overall Version 1.0 Editorial Team. My limited understanding is that the Wikprojects are setting the levels within their own projects.
 * Then, for example, there is Version 1.0 Editorial Team/Core topics, which is concerned with top priorities in general. We realize that value, priority, etc., are relative. But the core topics project is mainly working to bring needed attention to more general articles such as Culture, Humanities and Technology, geographic areas, etc., topics on the higher levels of a general tree of knowledge. Maurreen 14:47, 18 July 2006 (UTC)

Where can I find the debate in which the terms quality, importance and class were chosen? --WikiCats 14:56, 18 July 2006 (UTC)


 * Aside from "importance" (which has been discussed in a variety of places, including individual WikiProjects), there wasn't any debate. Nobody, to my knowledge, has ever objected to "quality" or "class" (or proposed any better wording, for that matter). Kirill Lokshin 15:01, 18 July 2006 (UTC)


 * There was no debate, per se. Something close to what you are looking for is [] about creation of the core topics subpage.
 * There seems to be something larger behind your questions. Maybe we could help more if you tell us more about where your questions are going. Maurreen 15:08, 18 July 2006 (UTC)


 * This seems to be the larger part of it. Kirill Lokshin 15:12, 18 July 2006 (UTC)

Thanks, Kirill. WikiCats, now some questions for you (and anyone else):
 * 1) How about changing "Start" class to "Needs expansion" or something similar?
 * 2) How about more explantion on the tags?
 * 3) Does anyone know of other offenses taken because of the tags? Maurreen 15:41, 18 July 2006 (UTC)


 * Also, the quality levels are based on those used by the chemistry wikiproject. Here an early version from Walkerma. -- Maurreen 16:24, 18 July 2006 (UTC)
 * Why change Start to "needs expansion"? It's shorter, and it gets to the point. Tito xd (?!?) 19:39, 18 July 2006 (UTC)


 * On the last point, we should stick with "Start" for the reason Tito states. Walkerma 02:34, 19 July 2006 (UTC)


 * Some history

The terms importance and quality naturally evolved on the Wikipedia 1.0 project. Quality assessments were part of Jimbo's proposals, and obviously importance lies at the heart of the Core topics proposal. As for WikiProjects, many besides chemistry developed assessment schemes quite independently, see this list for some examples (Pokemon, Medicine, Anti-War Movements and the Music listings are some of the older listings). In our plans at WP1.0 for WikiProjects we began  by considering quality first, once we started contacting WikiProjects for quality articles many (quite naturally) asked, "Do you want a list of our most important articles? or "What criteria are there for inclusion?"  Some typical examples are this (from October 2005) and this.  Going back through the archives it's clear that this idea came from the projects to WP1.0, rather than the other way round (though we at WP1.0 should have thought of it!).  We then began to discuss this in earnest here and then here. When I stumbled across the Math importance criteria I got very excited, though in fact they had themselves taken the idea from the Computer and Video Games Essential Articles List. When we began to talk about using this bot, it was natural by that point to try including it, though personally I had thought it would be entered manually - but projects such as Military History expressed a desire for the bot to read it directly from article pages. (Comment:I had a lot of fun reviewing the history of all this!)

It is apparent that try as we at may (at WP1.0) to come up with bright ideas, many of the best ideas (and much of the impetus for assessment) are coming from the projects themselves The WikiProjects are the reason why we now have over 20,000 articles assessed by over 40 projects (and many more not using the bot). Setting priorities (importance) is a natural question for any WikiProject, IMHO, and if a project neglects this they may often find themselves with a lot of specialised or crufty articles, while their key articles may be quite weak. Likewise, a project that ignores quality standards is failing in its duties to the community. Walkerma 02:34, 19 July 2006 (UTC)

See also this comment.
 * On Importance vs Priority

I think maybe we need to rename "Importance" as "Priority". "Importance" does seem to cause upset in some circles, and priority is perhaps less emotive (though not totally neutral). It would emphasise (more clearly than does Importance) that this is a rating within the related project, though (as has been pointed out) is should not imply a priority in the sense of "This one needs a lot of work." We need to reword the description for "Low Importance" also (barring any objections). "Quality" is pretty clear, and I think the assessment terms used are working pretty well now. Walkerma 02:34, 19 July 2006 (UTC)
 * Changing "importance" to "priority" is really easy. It literally requires just one edit, and that would take care of the listings themselves; the categories could be renamed/deleted, but if the main table says priority, no one would care about the name of a category. Still, it is a question whether we should actually do it; the real question doesn't change: "Who says that this article is more important than the other?" would be replaced by "Who says that this article is a higher priority than the other?" It literally is an endless can of worms similar to the dreaded notability/inclusionism/deletionism debates on WP:AFD. I've never liked to classify articles by importance, but to be honest, it does give many benefits. A article that is  is always an encouragement, to the same degree a  article that is assessed as a  is a wake-up call. By the way, the article I'm the most proud of is classified as . Does that hurt my feelings? Absolutely not.
 * Sounds like a good place to restate my first law of Wikipedia: "No matter what you do, someone, somewhere, won't like it." That is certainly what is happening now. Tito xd (?!?) 05:38, 19 July 2006 (UTC)


 * If things are being renamed, I'd like to suggest that low-Class is a prime candidate to get rid of first - it appears by name on lots of the bot-generated pages and is the thing I've felt is worst-named. low-importance or low-significance would be fine. "Priority" to me has more of a sense of now, and priority will change later, compared to importance or significance being an enduring property of the subject. I'd suggest losing the word "Class" completely in favour of quality and importance (or significance) respectively. It looks like the template has already been renamed, so it's primarily Bot output or hand-generated pages that still use Low-Class. Is the bot-owner listening here?


 * While I'm grumbling, I'd like to see a summary table (for each project, subproject, and overall) with importance on the x-axis, quality on the y-axis, and each cell has a count of articles as a link to the table/list showing all the articles in that cell, their rating date etc. Is this something the bot(s) could generate?


 * I echo WikiCats' comment below - thankyou for the history lesson. --Scott Davis Talk 11:37, 19 July 2006 (UTC)

Thanks
Thanks guys for the help. I was looking for some background information for my debate at Wikipedia talk:WikiProject Australia/Assessment. I find it exciting that others can see the potential harm that can be caused by the choice of the wrong words. --WikiCats 07:27, 19 July 2006 (UTC)

This is the problem as I see it. From the figures I've seem the vast number of articles are going to get the lowest rating. That means hundreds of thousands of articles with the judgment of low class, poor quality and little or no importance. Some people will get angry and complain but what I am concerned with the people who be disheartened and feel no encouragement to work on articles that the administrators have written off.

Scales that use the words quality, importance, class and low-significance have to re-classified.

If you look at the so called "Quality" scale Editor's experience column talks about the degree to which an article has reached completeness. So you could use terms like stub or half-finished or completed.

Iorek85 has suggested terms like Foundation, High, General, Detail, Specific.

Another term that would be quite OK is Key article, a term that is already used. --WikiCats 15:09, 19 July 2006 (UTC)


 * Not really, no. While "importance" is debatable, "quality" is a perfect description for what we're assessing.  I can't imagine what's objectionable about "class" either; it's a useful shorthand.  (Would you prefer "A-quality" and "B-quality" articles?  That seems like a fairly trivial point of semantics.) Kirill Lokshin 14:56, 19 July 2006 (UTC)


 * Agreed "importance" change to "priority" is the only term change I see as needful. :: Kevinalewis  :  (Talk Page) / (Desk)  15:08, 19 July 2006 (UTC)


 * Look at the "Criteria" column instead; that's what really defines the scale. The other columns are just there for further guidance. (In any case, "completeness" is an aspect—but only one—of quality; see featured article criterion #2b. An article can be "complete" in the sense that it includes all the necessary material and still be horribly written; conversely, even FAs are not "complete" in the sense that they never need to be edited again.) Kirill Lokshin 15:06, 19 July 2006 (UTC)

The problem with using the the word "quality" is that over half a million article will be put in the class "poor", sub-standard" or "rubbish" quality. The problem with this is obvious.

We have systems to ensure quality. So don't get yourself worried about that. As articles are reworked the quality improves. It's a fact of life.

Substituting the work "quality" does not mean the end of the world! --WikiCats 15:24, 19 July 2006 (UTC)


 * Err, so what's the problem? Half our articles are "poor", "sub-standard", or "rubbish"; are you suggesting that we use a different term in order to conceal this?  The answer to low-quality articles is to improve them, not to change the metrics so that they don't look so poor anymore. Kirill Lokshin 15:38, 19 July 2006 (UTC)


 * Agreed. Until someone gives a good, convincing reason why "quality" is inaccurate, I'll oppose any change to that. Tito xd (?!?) 21:57, 19 July 2006 (UTC)

I'm going to have to resort to the guidelines on this. The guidelines for Stubs make absolutely no mention of quality with regard to stubs. In fact they speak quite highly of them. It says "The community values stubs as useful first steps toward complete articles." So it is all about the degree of completeness. It talks about their size and about the amount of information they contain.

Do you have a guideline that says that Stubs are all about quality? --WikiCats 15:44, 19 July 2006 (UTC)


 * Stub talks about quality at a number of points, actually: "Stubs... do not yet contain sufficient information on their subject matter. In other words, they are short or insufficient pieces of information... A stub is an article that is too short to be genuinely useful..." Thus, while not all poor articles are stubs, stubs are, by definition, poor articles. Kirill Lokshin 15:49, 19 July 2006 (UTC)

Doesn't mention the word quality. It talks how big they are. Doesn't mention poor articles.

Kirill, you are a majority of one for those who can't see the problems with the terminology. Particularly at the base end where the majority of articles are.

But I'm happy to address your concerns.

What is this obsession with quality. Most people think they know quality. But quality is subjective, everyone has there own idea what constitutes quality.

We don't edit according to quality. We edit to comply with the guidelines. We don't have guidelines that explains how we define quality or how we determine what is quality and what is not.--WikiCats 13:22, 20 July 2006 (UTC)


 * You're still playing at semantics. How does "too short to be genuinely useful" not relate to quality?  If it's not useful, it's a bad article by definition; are you suggesting that articles can be not useful yet still of high quality?
 * (Not to mention that we do have explicit guidelines for quality: WP:WIAFA and WP:PERFECT!)
 * As far as my being a majority of one: you might have noted that there are 44 different projects using this system through the bot (and a number more doing it through the older style of manual lists); that there are more than 20,000 articles that have already been assessed; and that, aside from yourself, not one person has complained that we shouldn't assess "quality"! To put it quite bluntly, it is you who is alone on this issue. Kirill Lokshin 13:39, 20 July 2006 (UTC)


 * Hi. Welcome back. Neither of your references mention the word quality. We don't have guidelines that explains how we define quality or how we determine what is quality and what is not. --WikiCats 13:49, 20 July 2006 (UTC)


 * More semantics. What do you think "very best work" and "perfect" are in reference to, the weather in London?  This isn't all that complicated, really. Kirill Lokshin 13:53, 20 July 2006 (UTC)

If you can find a word that is defined in the guidelines then we can start from there. --WikiCats 13:54, 20 July 2006 (UTC)


 * "Perfectness"? But I don't think that's an improvement.  Kirill Lokshin 15:04, 20 July 2006 (UTC)

It's not semantics. Everything we do is determined by the guidelines. --WikiCats 13:56, 20 July 2006 (UTC)


 * Wikipedia is not an experiment in rule making. Kirill Lokshin 15:04, 20 July 2006 (UTC)

Let me understand what you are proposing. If you were assessing articles and you came across an article that was very short, say, one sentence but brilliant or quality that you would not give it a base rating. You would give it a higher rating or nearer the top. --WikiCats 14:41, 20 July 2006 (UTC)


 * In theory, yes. In practice, I can't imagine a single topic where a one-sentence article would be of high-quality.  Kirill Lokshin 15:04, 20 July 2006 (UTC)

You've painted yourself into a box there.

Ok. Lets take Perfectness. What wording would you use when rating something at the base of the scale? It is only the two base ratings that concern me. I don't want editors and new editors to be disheartened or turn their backs on articles because they have been judged of little or no importance and of trash quality. The vast majority of articles will be given the base ratings. --WikiCats 15:51, 21 July 2006 (UTC)


 * "Insufficient", perhaps? Or "Not useful"?  (For that matter, "Start"?)  But again, this is merely semantics.  The complaint seems to be a more general one about the existence of "negative" ratings, and I doubt any amount of rewording will change that.
 * Perhaps a better explanation is this: most editors, when seeing that others believe that their articles are deficient in quality, are motivated to improve them; only a handful seem to prefer angry rants against the rating system instead. It is my opinion that this latter group should not be mollified at the cost of depriving those of us who actually care about improving the quality of the encyclopedia—rather than sticking our heads in the sand and pretending it is not, on average, abysmal—of extremely valuable information. Kirill Lokshin 15:58, 21 July 2006 (UTC)

I'm having a problem understanding why you are the only one debating against this proposal. I'm assuming good faith but at the some time I can't help feeling that there is some other motive. There seems the be a drive to support a rating scale that directs more honours on to articles at the top of the list without regard to consequences. Being the creator of featured articles may or may not be a factor in this. I would hope that this not a case of manipulating Wiki to achieve a particular result in a similiar way to WP:DISRUPT. --WikiCats 14:47, 22 July 2006 (UTC)


 * Perhaps it is because—how should I put this—nobody takes your proposal seriously? The only reason I've kept up the discussion was because I considered it impolite to leave you without a response.  If you're going to make accusations and veiled attacks, however, I shall leave you to your own devices; you're free to sit here screaming into the silence for the next six months, for all I care. Kirill Lokshin 15:03, 22 July 2006 (UTC)


 * Yes, I had thought the discussion was over, I think other do too. I suppose we have taken for granted that some form of assessment is needed, but that's because we as editors want to know where our articles stand, and where improvement is needed.   I don't see any advantage in changing the system on quality, it has been refined over more than a year so the designators are about as neutral (yet clear) as you can get IMHO.  Every day, dozens if not hundreds of Wikipedians are quietly assessing articles for quality this way, and that should speak for itself.  Walkerma 17:17, 22 July 2006 (UTC)
 * And for what it's worth, I don't buy the argument that editors are going to be infuriated when someone says their article needs a bit more, because from personal experience, it just pushes editors to improve articles. I had said that before already, and since I see that someone else is saying the things I would say anyway, I don't know what benefit it would be for me to keep arguing. Tito xd (?!?) 18:03, 22 July 2006 (UTC)

I'm happy to accept Kirill's explanation.--WikiCats 01:53, 23 July 2006 (UTC)

A possible solution
I came across this terminology on the Computer Project. e.g. Talk:Half-Life_2:_Lost_Coast

It says:

Stub This article has been rated as Stub-Class on the assessment scale.

The term "quality scale" has been replaced with "assessment scale" but this does not stop the Projects talking about quality as much as they want.

This could be a viable solution to problem terminology turning up on talk pages. --WikiCats 02:14, 23 July 2006 (UTC)


 * Yes, I think we could allow that as an alternative (IMHO), though it is certainly much less specific than quality. It would work best in projects where they only assess on quality (though the list you cite was in fact the first to start using importance) I think we can offer the following alternatives to WikiProjects when they set up their templates:


 * QUALITY or ASSESSMENT
 * IMPORTANCE or PRIORITY
 * We should let the WikiProjects themselves decide what to use (they will anyway!). Walkerma 02:55, 23 July 2006 (UTC)

I would be very happy with ASSESSMENT and PRIORITY appearing on Talk pages.

They would be only link names with the links going to QUALITY and IMPORTANCE respectively in the Project spaces.

I would be very happy with this as a solution to my concerns. --WikiCats 04:03, 23 July 2006 (UTC)

I have to go interstate for a few days but I want continue this debate when I return. --WikiCats 04:23, 27 July 2006 (UTC)

I propose that the link name on the Talk page template be changed to Assessment scale with link still going to Quality and the other link name changed to Priority scale with its link still going to Importance.

This will only affect the wording that appears on the Talk pages. I make no proposal to change anything that would affect how the projects are operating at present. --WikiCats 14:37, 8 August 2006 (UTC)

Article progress grading
I don't know whether to commend or condemn your attempts at "grading" articles. Seriously, the results will be very crude, at best. I work with marking rubrics that have some striking similarities with this one. I know the problems with them inside out (I have done hundreds of analyses of data collected for this kind of format -- and no, I'm not kidding). Do you have evidence for the implied correspondence of the criteria with the descriptions of 'reader's experience' and so on? Do you have evidence for reproducibility of assessments across users and times? (rhetorical). It is more than a little ironic that measurement has been tagged as a "core topic" :-) Anyhow, if you're interested in a really stringent and highly precise process of assessment, let me know. It would take quite a bit of work though. For that matter if you're interested in tightening up the process at all, let me know, but anything will take a bit of data collection. It's fine as long as the info is treated as very coarse-grained and unreliable, but please don't presume it's anything else. Cheers Holon 14:08, 19 July 2006 (UTC)
 * I don't think we have ever regarded these assessments as anything but coarse-grained. At WP:WVWP have had cause to look over assessments done by scores of people from all sorts of subject areas, and (for myself) I've been surprised that we can elicit any agreement!  A few points:
 * The info is somewhat unreliable, but it is so much better that no system at all! At least the wiki system does allow unrealistic assessments to be corrected by others and to evolve towards consensus.
 * The "grades" are of necessity broad, to reflect the coarseness of the system and the variation of subject matter (an article on psychometrics vs Angel (Buffyverse))
 * The system is straightforward enough to allow pretty much anyone do assessments quickly. More sophisticated systems such as Article assessment have been worked on, but I don't think we have enough people available to assess a significant number of articles that way.
 * We don't have any objective "assessment of our assessments" at the moment. Maybe one day!
 * If you can see any simple, specific tools we could use to enhance our procedures, or anything you think that is glaringly wrong, I think we'd like to hear from you!


 * BTW, I'm a college professor, so I'm very familiar with marking! Perhaps you can do some assessments for us!  Thanks, and keep in touch, Walkerma 14:37, 19 July 2006 (UTC)


 * Yes, that's right, the difficulties I'm speaking of are wide-raning (if that's what you're pointing to). Depending on what you mark I can give some tips. I'll think about simple speficic tools. The most obvious thing is more examples. For many reasons, exemplification of anything possible (specific criteria, not just the whole) is valuable in giving clear reference points. Pairwise comparisons of articles in rleation to criteria is really the most precise way to go, however. It would be possible to build up a table of info where everyone has a job to just compare a couple of artilces with respect to criteria. You'd begin to build an order and a scale. Then you can set some (arbitrary) cut-points and place dessriptions around them. The problem is people try to do it the other way -- that is, use a handful of words of description and try to categorize accordingly. Entities such as articles are their own best description and are best compared directly. Grading "scales" such as this are redundant. It's not that hard, just takes a methodical process and some coordination. Think about it anyhow Holon 15:12, 19 July 2006 (UTC)


 * I would go one step further and actually suggest that a more stringent assessment system would be less successful. On a practical level, if an assessment cannot be eyeballed in under a minute for the vast majority of articles, the average editor simply won't bother with it; he is, after all, a volunteer.  This is why Article assessment has gone through a few dozen articles while we have more than 20,000.
 * The problem is you're assuming it's doable. There is no reason at all to think you can get information that is worth even the miminal efforts. Ther are so many problems with this appoach, it's difficult to know where ot start, though I'm happy to send powerpoint outlining five years of research into the waste of time that is 'grading' using this sort of schema. No, they won't do it. Doesn't mean it's worth doing badly.


 * As far as all the exhaustive descriptions of "reader experience" and so forth, I'm pretty sure they aren't really needed—we could just as easily have described the article levels as "bad", "not so bad", "decent", "good", and "great", and most people would understand what was meant—but I doubt they're actually harmful in any way. Kirill Lokshin 14:53, 19 July 2006 (UTC)
 * My point wasn't the descriptions are poor -- they're not too bad. The point is there is an implied correspondence between sections of the rubric (such as criteria and reader's experience) for whcih there is simply no evidnence. To claim a reader's experience is even statistically (e.g. most likely) x given a criterion of y is nonsense. Holon 15:01, 19 July 2006 (UTC)
 * Well, remove the entire "reader experience" thing, then. As I said, the primary objective here is neither to determine the exact quality of an article relative to other articles nor to obtain a canonical description of how the article is regarded by any particular reviewer, but merely to determine whether a given article is halfway-decent or not (with a bit more precision than a simple two-level "garbage"/"not garbage" scale). Kirill Lokshin 15:22, 19 July 2006 (UTC)


 * I think that things like "reader experience" are helpful because (a) they provide more guidance to the assessor and (b) they remind the assessor (who is most likely to be mainly an article editor more than a reader of such material) that most people seeing the page will see it differently from the assessor. The fact that we haven't verified that such experiences are valid doesn't destroy their value as basic guidelines.  We're not making any claims about them!  Maybe we could introduce a few "should"s into them, if you're concerned, but I think they are implied.  I think the idea of a range of examples is an excellent, and I would be very happy to see someone work on this!  In my teaching organic chemistry, I've found students find worked examples to be an excellent way of going from the abstract & general to the concrete & specific - great idea! Walkerma 16:24, 19 July 2006 (UTC)


 * I'm not saying things like reader experience aren't helpful. The problem is that when you construct a matrix like this, it is logically implied that for a given classification (e.g. "B-class") the criteria and the editor's experience correspond. This is an empirical question. If they don't correspond in the actual articles, then it leads to confusing. On a more constructive note, it is quite straightforward to establish a set of reference points to clearly guide assessments. Simply select 4 or 5 articles in each category. This will give about 24 to 30 articles in the six classifications for which there is an implied heirarchy. Pring out the assessment criteria. Carry out pairwise comparisons of each article with each other. If someone lists the articles and the classifications, I can design a matrix with only 80 comparisons that will give highly efficient information. If you can get 80 comparisons (which is better, A or B?) I can pretty much guarantee you'll order the articles well. Then use these as reference points in order. To assess a new article, someone should judge which classificaiton they think is roughly correct, then compare with one of the chosen examples. Select which the new article is most like, and slot it in at that classification. The ordered articles then provide the clear set of reference points for any assessor. In my experience, a little work like this at the beginning makes life a lot easer when it comes to decisions because ambiguity makes people unsure and the process is actually very inefficient. Let me know if anyone's interested, I'm quite happy to design the pairs and to do the analysis at the end -- it's very quick. Cheers 03:59, 20 July 2006 (UTC)


 * Sounds very interesting! I'd like to take you up on the offer, and I'd like to get help from other interested parties.  We can at least do a thorough job on our examples, even if we can't on the other 20,000+! Thanks, Walkerma 04:37, 20 July 2006 (UTC)


 * Strictly speaking, we only need examples—at least for comparison purposes—on five of the classes ("A", "GA", "B", "Start", and "Stub"). The "FA" designation (a) is assigned on a pro forma basis to those articles (and only those articles) that get promoted to featured status and (b) is assumed to be the highest designation available; hence, there's never any need to compare articles when assigning that class. Kirill Lokshin 04:46, 20 July 2006 (UTC)


 * Understood. However, at least a couple of examples from the classification will provide valuable information about the relative quality of articles in the next classification down. In fact, the FA classification is likely to provide the clearest information becasue they've been through a more comprehensive process. Even if you don't want to have multiple exemplars, I'd still select two or three FAs if you want to order through pairwise comparisons. Perhaps people can suggest articles they think are the clearest (possible) examples of each classification, and if we get four or five in each, I'll mail the info to those who are will to do comparisons. I'm guessing based on pairwise with essays it should take less than half an hour to do ten comparisons. Maybe less than that, but about three minutes per comparison to scan over and look for a few key elements. Holon 07:12, 20 July 2006 (UTC)

Depth and 'Quality'
I note that 'good' articles appear to be those restricted to the 'where','what','when' and 'who' of a subject, expressed in lists of phrases such as one might employ to teach a parrot to recite. Is this really knowledge? Where is the 'how' and 'why'? I cannot see how any deep understanding, particularly of science, engineering or mathematics could possibly be imparted by such bland articles. Gordon Vigurs 10:15, 20 July 2006 (UTC)


 * You may be right, that's probably a major difference between Good Articles and Featured Articles. Your comment reminds me of my (history?) teacher's comment {paraphrased): "In O-level we teach you what, but if you want to learn why you have to take A-levels."  I think there's a good reason for it - "What" is easy to pull out of a book and is usually uncontroversial, whereas "Why" will often involve a POV.  Try writing about "Why" the stock market and the federal budget have suffered during George Bush's term and you'll see - but if you list the relevant numbers in the appropriate articles with citations no one will question you!  You really need to raise this on the GA talk page, though - that is where GAs are actually voted on. Walkerma 04:23, 23 July 2006 (UTC)

Opt out
May editors remove the assessment tag from a page if they don't want the article assessed? What I mean is, if all WP articles receive assessment tags, which certainly seems to be the direction things are going in, but are not then assessed, what is the point? -Acjelen 20:29, 23 August 2006 (UTC)


 * Well, they're eventually going to be assessed if there's a WikiProject for them, so it seems somewhat counterproductive to remove the tags. Much better would simply be to assess the article; most projects are very liberal about having people—even those who don't consider themselves members of the project—helping out with their backlog of unassessed articles. Kirill Lokshin 20:33, 23 August 2006 (UTC)


 * Honestly, what will that achieve, aside from exterminating every red talk-page link, and bloating the size of the database with another several hundred thousand talk pages with no information content? Wouldn't it be more sensible to have an "opt-in", where only articles with a meaningful assessment are tagged, on the basis that the remaining vast wadge of stubs are never going to make in into WP1.0 anyway, especially if no human being has ever actually "assessed" them for such?  (I'm assuming for the purpose of this message (though not necessarily otherwise) that 1.0 is an actual product that will actually appear at some point.)  Alai 16:42, 28 August 2006 (UTC)

Little question
If I'm correct, then assesment for 1.0 may replace previous assesment boxes for 0.5 on talk pages? Please answer on my talk page.--Steven Fruitsmaak (Reply) 23:20, 29 August 2006 (UTC)
 * What if another project has already assesed an article, such as on Talk:Cancer?
 * Should it replace ? And  ?Or --Steven Fruitsmaak (Reply) 00:09, 30 August 2006 (UTC)
 * Not quite sure what the problem is here. There's nothing wrong with having multiple boxes, if they're from different sources; the WikiProject templates are entirely unrelated to the V0.5 ones. Kirill Lokshin 00:11, 30 August 2006 (UTC)

C-Class
I would like to propose a class for articles that are in between B class and Start class. In my opinion, a C-Class assessment can serve as a borderline between B and Start, since the current ranks go from B and drops off steeply to a Start class. I know that there are six ranks in the assessment table, but before the article can go from B class to GA, it has to go through some kind of nomination process. That means that there are only three classes to put articles without putting it through a week or month-long, headaching process; B class, Start class, then Stub class. And since I think Start and Stub class articles are the lowest ranking articles in Wikipedia, and B class is "okay, almost there (Since it is one level lower than GA)," I think a C class can be created to describe "good, but not quite." An example of a C class article in my opinion is Son Goten, which is currently rated as "start", since it couldn't be rated as "B-Class." Goten's article, in my perspective, is slightly higher than the expectations of a start article (especially when it was "started" in July 2003), but falls a little bit short of the expectations of a B class, therefore assessing it as a C class. But, other than Goten, I believe there are thousands of articles in Wikipedia that are listed in the B class or the Start class that should really be listed in the hypothetical C class. 65.8.35.224 22:01, 31 August 2006 (UTC)


 * We have had a variety of suggestions in the past, including this one, but the consensus reached has always clearly been the same - the current simpler system works well. Interestingly in the original proposal to WP:Chem there was a C-Class like you suggest, but then things were switched over to the system we currently use.  In principle, therefore, I wouldn't have minded a C level, but at this point after 70,000 articles have been assessed using the current system, it wouldn't be appropriate.  We haven't seen any strong desire to change from this system. Of course an individual project can adopt their own system, it just couldn't be read separately by the bot.  Thanks, Walkerma 00:56, 1 September 2006 (UTC)
 * Obviously, the AfD class exists - a lively and contentious category. 24.148.93.88 08:29, 24 October 2006 (UTC)
 * Just to clarify - the A-Class doesn't require GA-status. Girolamo Savonarola 18:53, 24 October 2006 (UTC)
 * Also, since this original discussion took place, the Mathematics WikiProject has started using a B+ assessment level. This shows up in the bot list just as a B, but on the article talk page it appears as a B+ indicating what we tend to call "a good B".  This seems to be working well - so instead of a C-Class, your project could simply use a B- (going into bot category as B-Class) or even a C (going into bot category as Start-Class).  Walkerma 19:11, 24 October 2006 (UTC)

editing classification table
On the table listing grades and the assessment sidebar on some pages there are no links to Good Articles and Featured Articles as there are for A class articles and so forth. It would be nice to see some consistency, I just can't figure out how to edit this table myself. RichMac 22:46, 15 September 2006 (UTC)
 * Done. Seems like a really good idea, especially for consistency's sake.  -Runningonbrains 01:44, 16 September 2006 (UTC)

bug 4288
Please vote for bug 4288. This would allow you to tag an individual revision of the article. I think that would be helpful to your project. --Gbleem 18:09, 16 September 2006 (UTC)
 * Voting on bugs isn't the most effective thing to do, as developers seldom pay attention to that. However, stable versions, as well as version tagging, will be implemented soon, as Brion Vibber indicated during Wikimania. Tito xd (?!?) 18:23, 16 September 2006 (UTC)

FF class?
The talk page for Robotech: The Shadow Chronicles has a box that says, "This article has been rated as Future-Class on the assessment scale." There is a reel of film symbol that says 'FF'.

However, I am not able to find any references or explanations for FF, FF-class, or Future-Class either on the main page or the discussion page of the assessment scale. Shouldn't this be fixed? --Robotech_Master 17:42, 17 September 2006 (UTC)


 * It seems to be a custom class used by WP:FILMS; please discuss the issue with them. Kirill Lokshin 18:27, 17 September 2006 (UTC)

Criteria changes for Good articles
There quite a heated debate at Wikipedia talk:Good article candidates, Wikipedia talk:Good articles over changing the criteria, especially a proposed requirement for inline cites. 1.0 members might like to join the discussion to help ensure consistancy between the projects. --Salix alba (talk) 19:46, 28 September 2006 (UTC)


 * The speed with which this dispute is spreading to vaguely related places is somewhat disturbing; here, it wouldn't even have been an issue if we hadn't shoehorned the GA level into the assessment scale in the first place. Kirill Lokshin 19:55, 28 September 2006 (UTC)


 * The GA discussion will definitely have repercussions here, we will probably need to discuss it once it's resolved. It seems that GA standards getting tougher and tougher, and I think many A-Class would now have a hard time as GA candidates - not the case 6 months ago.  Also see this related discussion - it's already affecting us.  The person posting the question on that page is someone who recently set up a bot assessment scheme for WikiProject Space exploration.  Walkerma 04:20, 29 September 2006 (UTC)
 * Well, aren't A-Class articles supposed to be close to FA status anyway? I don't see how changes in GA would affect the vast majority of them. Perhaps a few of them may be changed, but otherwise, I can't foresee a significant change to this scale. Tito xd (?!?) 04:25, 29 September 2006 (UTC)


 * The main effect is how we deal with the ranking of B/GA/A. I'm hoping we don't need to make any changes, you can see my own views in more detail at the Grading scheme page.  I think we will need to clarify our consensus view for the hundreds now using this assessment scheme, or Mark's question on the Grading scheme page will become very common, and it will hinder the assessment process. Walkerma 04:32, 29 September 2006 (UTC)


 * Given GA's track record, I very much doubt it is in our interests to make arbitrary changes in an attempt to tag along with whatever new standard they're trying to impose this week. I would go so far as to suggest we simply remove the distinct GA level from the scale here (as it's not something that the WikiProjects can work with themselves, and lacks the existing gravitas of FA) and let every project decide whether they will give articles externally marked as GAs A-Class status or not. Kirill Lokshin 04:39, 29 September 2006 (UTC)


 * I could go along with that change myself, Kirill, but since it's in place I think I'd like to let projects make up their own minds whether or not they use it. I also think that the standard will stabilise in time, and that the GA tag already carries some gravitas, so it does serve as a useful benchmark.  If I were setting standards for a particular WikiProject I'd put it in the bot tables in the comments table rather than as a "grade" in this scheme.


 * I wasn't so much thinking of a discussion here about the effect of inline citations - I agree that would be silly. But there has been a big change in GA standards over the last few months, and some GA reviews I've seen recently (I read a lot when doing assessments for V0.5) seem to have been more thorough than some FA reviews!  That does raise issues when the GA level does appear in our scheme.  Walkerma 05:00, 29 September 2006 (UTC)

In many ways, i think it would make more sense in a 6-tiered grading system to have the top two tiers reviewed independently, rather than the 1st and 3rd.. so maybe it would be better for the GA standards to be high, leaving no need for a tier between GA and FA. Mlm42 08:36, 29 September 2006 (UTC)


 * Hmm, I'm torn. I'm one of the ones advocating inline cites, but I didn't think it meant ipso facto that this meant that would kick it up the assessment scale. I definitely don't want to cause that kind of trouble. My view was that this helped place articles squarely between being a B and an A. I see A's as potential FAs that just haven't had a peer review to make sure there's nothing we've missed, etc., whereas GAs might not have a tight, well-written lead, but a decent lead; it might not cover some things that should be covered, etc. I guess I saw the difference as a difference in prose and how well-written it is, than just that it has inline cites. I think the reader's experience line in our assessment scale for GA is apropos "Adequate for most purposes, but other encyclopedias could do a better job." I guess I see cites as a basic requirement, I didn't see it as something that elevates it up the scale -- even WP:DYK gives preference to articles that have inline cites as an effort to encourage this. --plange 15:39, 29 September 2006 (UTC)
 * I agree - an uncited article shouldn't even receive a B-Class assessment, IMO, but I do think that there are a few differences between GA and FA, and that an FAC-quality article without the little star is A-Class. While some projects don't make use of the difference, for others there is a certain difference between A- and GA, and changing the scale would be slightly disruptive. WP:TROP, for example. Tito xd (?!?) 19:57, 29 September 2006 (UTC)

I've come here because I was looking for an answer to the question about whether or not editors doing assessments can class an article as GA-class without it having had a formal GA. I've been doing this so far at WikiProject Filmmaking, because to my eyes, the grading scheme examples do clearly progress significantly enough to warrant a middle-classification between what is regarded as A-class and B-class. I would strongly disagree with the notion that B-class articles should have citations, simply because we have plenty of great articles here whose only major flaw is the lack of references, but are otherwise brilliantly written and contain plenty of depth and nuance. I would say that some of them are clearly (to my eyes) A-class. I wouldn't disagree, however, that B-class articles should at least have some external links if nothing else in the way of outside sources.

At the end of the day, however, the assessment process has nothing to do with the Good Articles project in and of itself, nor should we feel constrained by it. Personally, I feel that renaming the GA-class to something else will be more productive because it will allow assessing editors to give a middle grade between A and B without having to worry about whether or not it has gone through some other distinct process that most editors will not be inclined to submit a random article to, just for the purposes of assessing it at a slightly higher class. The solutions to me would be: 1. raise GA class above A-class, 2. rename GA class to something else which will remove entanglement with the GA project, or 3. abolish the class entirely for assessment purposes. I think that option 1 is flawed because it seems to imply that once an article is at A-class, it has two reviews it can go through - why bother to use GA then when you can simply leapfrog it to FAC? (Which already happens in droves.) I also disagree with option 3 because I think that the difference between A class and B class is too distinct and there needs to be a middle level.

There is another way around this problem, which is simply to argue that if the assessing editor has not worked significantly on the page that they are assessing, then they could count themselves as a WP:GA reviewer and thus judge the article to have passed GA standards. This will help both projects because it will greatly increase the number of Good Articles while at the same time keeping it easily integrated within the assessment system. And of course, if someone disagrees with the GA status, it can always be brought up for review. How does this all seem? Girolamo Savonarola 20:27, 1 October 2006 (UTC)


 * Having the projects automatically add GA tags as part of their assessment process isn't a bad idea in theory, but I wonder if it could actually work in practice, particularly insofar as the actual GA process picks up more structure. Kirill Lokshin 21:17, 1 October 2006 (UTC)

Articles covered by multiple WikiProjects
I've noticed that many articles will be covered by more than one wikiproject, and hence may have more than one banner.. and in some cases this will mean more than one assessment given. so the obvious question is whether or not the bot counts these articles once or twice? should i worry about this when tagging my WikiProjects banner (Template:WP Space exploration) to pages that have already been assessed? Mlm42 10:15, 29 September 2006 (UTC)


 * As far as I know, the bot has been counting such articles only once for quite some time now; that was something specifically requested when the issue first came up. Kirill Lokshin 13:03, 29 September 2006 (UTC)


 * Then the next obvious question is, if there are two differing assessments, which one does it count? Mlm42 14:15, 29 September 2006 (UTC)
 * IIRC, both, just one for one project, and the other for the second project. If there are differences in assessment, though, projects should discuss it between themselves on a case-by-case basis. Tito xd (?!?) 19:59, 29 September 2006 (UTC)


 * There is also a problem with lots of space being taken up at the start of the talk page with lots of direrent boxes. For example Talk:Carl Friedrich Gauss has 7 boxes on it, and I've seen pages with many more. I've now redone Template:maths rating so it only used two lines, which should save a bit of talk page clutter. --Salix alba (talk) 21:28, 29 September 2006 (UTC)


 * It's not too bad, I think. While the tags to eat up a certain amount of space (which isn't all that significant on talk pages that are quite long to begin with, incidentally), the benefits of having them—which go beyond assessment (and, indeed, are primarily not about assessment at all)—generally outweigh the minor inconvenience when multiple ones appear on a page. Kirill Lokshin 21:38, 29 September 2006 (UTC)


 * One solution to this is at Mini Talkpage Template where a smaller version of the templates is being investigated. You can see its use at Talk:Eigenvalue, eigenvector and eigenspace. --Salix alba (talk) 10:23, 15 October 2006 (UTC)

Colors
Am I the only one who doesn't like the new colors? Tito xd (?!?) 19:32, 4 October 2006 (UTC)
 * No, you're not. I agree that the colors should be changed back (the new color for Start-class is too close to the color for Template-class, for instance). The new ones look too startlingly different. --Core des at (talk) 19:34, 4 October 2006 (UTC)
 * (edit conflict) Nope, not the only one; they're too neon-ish for my taste. Is there any reason we can't just use a normal six-color spectrum here, though? Kirill Lokshin 19:35, 4 October 2006 (UTC)
 * Well, IIRC, we had green and blue on one side of the spectrum, to give the impression that they would be what was acceptable for 1.0, and the other ones were reddish to indicate they're not... Tito xd (?!?) 19:38, 4 October 2006 (UTC)

For the record...

The only major changes has been to make A-class cyan rather than green, and to make GA-class green rather than a pale wishy-washy green. The other colours have been changed only so that the shift in colour is the same from one to the next. Tompw 19:48, 4 October 2006 (UTC)

Further to the above... a pastel version would look something like the following, which to my mind just doesn't seem to work very well. (Pure personal opinion that). PLus the above colours can be roughly replicated using red/orange/yellow/lime/cyan/blue. Tompw 19:53, 4 October 2006 (UTC)

Who said anything about pastel? ;-) I would prefer something rather darker/duller:

The colors are approximate, but the general sense of my idea should be clear, I think. Kirill Lokshin 20:02, 4 October 2006 (UTC)
 * The trouble with darker colours is that the contrast between background and text isn't wonderful, which doesn't help acessability. Tompw 20:05, 4 October 2006 (UTC)
 * I think we should go back to the original colours as a starting point for discussion. The only changes I would support would be for GA and A (I prefer the new colors for those two), but I'd also be happy with the way things were.  There were several ideas when the colours were originally set up: (a) The "spectrum" from red to blue representing the spectrum of assessments; (b) The idea of bright red being a warning (STUB WARNING!) and the green of GA was supposed to go with the green icon that is used.  At some point someone made that green more pastel, presumably for the reasons Tompw gives in the post before this one.  So, is there any powerful reason we need to change? Walkerma 21:16, 4 October 2006 (UTC)

I say revert back to the old colours for now and discuss on here without further changes. Just try and avoid a truly massive discussion on this topic, it has happened before. My opinion is that changing A class to a light blue is reasonable allowing a darker green for GA; but apart from that no change please.--Nilfanion (talk) 22:38, 4 October 2006 (UTC)


 * We also need to sonsider what they look like on the standard class="messagebox standard-talk" background which is where they frequently appear so:

Less Neon and dulled B's don't have much contrast with the background. Personally I prefer a purpilish FA as it has a more rainbow feal, so I've tweeked new colours and added a more purple FA. --Salix alba (talk) 00:32, 5 October 2006 (UTC)


 * Wow... I honestly didn't think it would generate so much discussion. I feel the purple is too dark for good contrast, plus I feel that FA should be a similar colour to A-class. Purple and cyan are too different. By the way, I changed template-class to magenta, which solves the problem of looking like any of the colours above. (Also, I remember the Tropical storm colour thing... I lacked time to contriubute, as the discussion took place over christmas/new year.) Tompw 09:18, 5 October 2006 (UTC)


 * i'm not a fan of the purple here because it overlaps with the article 'importance' colour scheme, which are purply colours. i like having both A and FA class as blue. so i'm in favour of the 'new colours'. Mlm42 12:20, 5 October 2006 (UTC)


 * I vote in favor of the "new colors." 38.100.34.2 19:04, 24 November 2006 (UTC)

Images on classification templates
Could someone please create versions of the images with transparent backgrounds rather than white? Also, why were images added to the other classes? Tompw 13:15, 7 October 2006 (UTC)


 * although i'm not flat out opposed to each class having an image, these ones are no good.. especially the stub big red X symbol! it's not like we're handing in these articles to be marked, this is a work in progress.. teachers don't come around and draw big red X's on your essay after you've been working on it for three minutes! :) Mlm42 14:32, 7 October 2006 (UTC)


 * The GA and FA templates have an icon attached because (a) those icons are strongly associated with the GA or FA tag and (b) they represent an external "award" rather than a WikiProject assessment. I think it was an interesting experiment, but with over 130,000 articles now using these templates you're likely to ruffle feathers with any changes - IMHO it's best to discuss first any template changes here (or on the template talk page), and reach consensus before applying changes.  Walkerma 03:38, 9 October 2006 (UTC)
 * Actually, here. That way, it gets more eyes that can discuss it. Tito xd (?!?) 21:42, 12 October 2006 (UTC)

New GA Green proposal
I apologize for revisiting the colors issue again, I'm sure everyone was glad that issue had dropped off everyone's radar... I'd just like to propose that we change from the current sea-foam green to a more neutral green that doesn't conflict so much with talk-template backgrounds such as on ...

Current:

Proposed possibilites:

I can't edit the template myself, and discussion should probably occur first anyway. Thoughts? -- Renesis (talk) 23:32, 7 November 2006 (UTC)
 * The first proposal is better than the second one, although I'm still undecided whether it is better than the current one... Tito xd (?!?) 23:56, 8 November 2006 (UTC)
 * The exsisting colour scheme was carefully chosen so that the colours used for the various grades were as disinct as possible (by being equally spaced within the RGB colour space). Therefore, if you wish to propose a new clour for any grade(s), please show *all* the grades/colours togther, rather than one in isolation. Also, if it ain't broke, don't fix it. :-) Tompw (talk) 00:52, 15 December 2006 (UTC)

More thoughts on GA standards, citations, and assessment...
Please join the discussion, as it pertains to the Assessment project and several issues raised or broached here. Thanks! Girolamo Savonarola 00:16, 9 November 2006 (UTC)

setting up templates
Ello. I would like to set up this assessment machinery for WP:TREK. Is there an easy howto guide somewhere or does some kind soul mind helping us out with this? Apologies if I am missing something obvious. Morwen - Talk 11:41, 20 November 2006 (UTC)
 * See Version 1.0 Editorial Team/Using the bot, the easiest way to set up a template is to copy an existing one and modify it as necessary. If you get stuck you can ask here or on Wikipedia talk:Version 1.0 Editorial Team/Index. --Salix alba (talk) 13:18, 20 November 2006 (UTC)


 * Alternately, just follow the instructions here. :-) Kirill Lokshin 14:24, 20 November 2006 (UTC)

request for change to criteria
Currently pages that have "significant gaps or missing elements or references" or need "substantial editing for English language usage and/or clarity, balance of content, or" contain "other policy problems such as copyright, Neutral Point Of View (NPOV) or No Original Research (NOR)" are eligible to be graded as B-Class. Copyright problems are a serious matter; therefore, I request that articles with copyright problems not be eligible even for B-class. 19:08, 24 November 2006 (UTC)

"Template-Class articles"
Would it be churlish to point out that category names like are rather oxymoronic? Why are they even necessary? Even if templates have to be tagged with wikiproject templates at all -- the need for which in the first place escapes me -- what possible purpose is there in that template transcluding a "class" category that seems to have no actual meaning whatsoever? (If yet another #if is necessary to stop this category name being formed by the existing code, I'm not sure it'd make much difference to the overall esotericness thereof.) I have a distinct feeling that if I ask these assorted wikiprojects why, or take these to CFD, the familiar mantra "necessary for 1.0" will be deployed, so I'd like someone to give me the "1.0 take" on this. Alai 20:53, 8 December 2006 (UTC)


 * It's not primarily a 1.0 issue at all (although some projects are likely to have copied it under that assumption). Aside from the whole rating scale thing, WikiProject tags generally appear in two places:
 * Articles - this is pretty simple
 * Not-articles - includes templates, categories, and a variety of things in articlespace that aren't really "articles", such as disambiguation pages
 * The question, then, is this: if the second group is to appear in a category (which is quite convenient), what category should that be? Putting them together with actual articles is confusing, so most projects create a separate category for them (e.g. Category:Non-article military history pages).  Once this category grows to a certain size, a project may decide to split it up by the "type" of non-article, creating separate categories for templates, disambiguation pages, etc.  Hence, the existence of the cateories you refer to.
 * (The names are, admittedly, quite strange in many cases, and I suspect that the only reason for the prevalence of the more bizarre variants is the wholesale copying of template code. I doubt that anyone would object to changing the names to be more logical.) Kirill Lokshin 21:01, 8 December 2006 (UTC)
 * If they're a) to be tagged, and b) the tags are to categorise these on a by-namespace basis -- neither of which I'd consider a given; the first of these I noticed had about three templates in it, so it's hard to see size as a pressing concern (certainly in every such case), and the need to tag them at all seems like a solution in search of a problem;  are these denoting "scope", which is generally thunderingly obvious, or 'ownership'? -- at a minimum the category should be "-Class Blah pages", or less tortuously, just "Blah ".  So I hope you're correct about the last part...  I'll drop notes at a couple of randomly-slected wikiprojects.  Alai 21:57, 8 December 2006 (UTC)

Re-assessment?
I've spent some time cleaning up an article previously graded at Stub class, and I'm pretty sure it doesn't fit into that class now. How do I go about getting it re-assessed so I can see the fruits of my labours? Tbone762 13:29, 14 December 2006 (UTC)
 * Since it's under the purview of WP:CVG, you should bring it up in their talk page. Tito xd (?!?) 19:52, 14 December 2006 (UTC)

Article progression from GA to A to FA???
Can someone confirm that articles should progress from GA to A to FA? (This seems strange considering that GA and FA articles have in my experience so much buzz compared to A Class. I've heard of Wikipedians trying to make an article GA or FA but not A. I've never even heard of A Class until now.) --Tiger Marc ROAR! 00:49, 30 December 2006 (UTC)


 * Yep, that would be the full progression (although it's quite common for articles to skip levels—B-Class to FA is a common jump, for example). As far as A-Class, most projects don't (yet) have any formal process for assigning it, which makes it somewhat less process-driven than GA. Kirill Lokshin 00:53, 30 December 2006 (UTC)