Proposal talk:Comparison with printed encyclopaedias

From Strategic Planning

response and comments

Thanks for translating on your proposal into English. My German is passable, but I lack much of a vocabulary. Besides, unless I am looking for a hooker and needing to give sexual instructions, what German I do have is utterly useless! ;)

I understand the merits of your proposal, but I am not sure why its applicability needs to be formalized. Cannot we just do this by ourselves. Like a paper-based encyclopedia, Wikipedia is meant to be a compendium of knowledge, but unlike one, it is web-based which gives it a flexibility and an expansiveness that the paper-based encyclopedias do not have. Why just compare Wikipedia to a book-type (BT) encyclopedia simply to get an article count? Consider the following questions for your proposal:

  1. Can we develop a set of encyclopedic standards based on BT encyclopedias?
  2. If so, which of these standards can we use to increase the quality, credibility, and reliability of Wikipedia?
  3. Are there things common to BT encyclopedias, such as layouts, methods of citations, governing structures, interaction with students, that we could also incorporate into Wikipedia?
  4. Could the logical conclusion to your proposal be the development of written and published version of Wikipedia which is tied to the web-based version?

Hope this helps and does not confuse.GMJ 20:18, 15 August 2009 (UTC)[reply]

Don't want to skew independence of evaluations

Newspapers, magazines, and popular press web sites often independently commission such studies, selecting articles on various topics and asking experts local to the reporter author to compare them. The English Wikipedia, at least, has been fairing increasingly well in such studies. I am somewhat concerned that if the Foundation sponsored such studies, which I do support in general, that they might be seen as potentially skewing the results in favor of the Foundation. Would using a third party organization with an impeccable reputation for objective analysis of reference sources be more appropriate than sponsoring such studies directly? For example, would it be reasonable to contract out such studies to a newspaper conglomerate, an academic press, or the review board of an English humanities journal specializing in reference works? 01:40, 16 August 2009 (UTC)[reply]

Well, as far as i see my proposal, it's primarily for an internal use. But if we use that to advertise wikipedia as the best encyclopedia out there, then we must ensure that our rankings are correct. This could be done by the need of a double-check of the respective ranking or that only a wikiproject/wikiportal member could do so. The one way or the other, a system to control the rankings must be installed and of course the help and knowledge of externals should not be denied, but used to its fullest.

I actually don't see the problem of the WMF trying to bend the results, but I agree that this could be one point to be criticized by others (mainly the publishers of BT encyclopedias). Mad_melone -- 08:48, 16 August 2009 (UTC)[reply]

Some more thought about it

It's me again, the proposer.

I put some more thought in the issue yesterday evening and I thought about the technical realisation of the project. Something like that came across my mind:

1) First we need to OCR a complete BT encyclopedia (or multiple, or read out other internet based encyclopedias). By that we get all the articles they have and moreover the WMF could set up an internal wiki which would be feeded with the OCR-ed data. By that we would be able to see, if they changed anything from one printed version to another. 2) As we have all the articles now, we link them to the respective articles in the wikipedia. First a simple "same name" comparison, then eventually get rid of the "(disambiguation)" pages and use the appropriate wikipedia article. Then we have articles that don't have an equivalent in wikipedia or are simply listed under a different name. 3) Categorize all the articles. I would propose that all topics are assigned to a specific wikiproject or wikiportal, so that the wikipedia experts can decide about the status of the article. Thex especially have the knowledge if the article is listed under a different name.

I think that this approach would also have following advantage: lots of times our writers are engaged in some kind of wikiproject or wikiportal, most of them in multiple. And very often you have time to write a new article or upgrade a stub, but you don't know where to start. And with this approach you will have a huge list of topics that would really bring the wikipedia forward to get the best encyclopedia out there. 4) The usage of the traffic light system could also be done by the wikiprojects/wikiportals as it seems to be already quite common to do so in the english wikipedia. Again, in the projects/portals you have the wikipedia experts and they should be able to rank the articles as following: red: article in the wikipedia either non-existent, a stub or substantially worse than in respective BT encyclopedia. yellow: article at same level as in BT encyclopedia, but could (easily) get better. This should be the most common ranking in the beginning and should be enough for the proposal as it ist. green: article in the wikipedia is better than in the BT encyclopedia. This could be the case for excellent articles or stubs in the BT encyclopedias.

Surely OCR of other encyclopedias which are then uploaded to Wikimedia (or anywhere else) would be absolutely illegal for copyright reasons. We already have our own article rating system and prioritisation system on en:wp. I'm not sure another is needed. --Bodnotbod 17:48, 18 August 2009 (UTC)[reply]
If you don't make them available for the public. but the OCR-ing is just optional, we could just buy the BT encyclopedias in order to be able to compare the quality. As for the "we already have our own rating system": that's true and i don't want to see that replaced, but my idea actually goes a totally different way. The key question is: where is the comparison for the rating system we already have? It's an internal one, and I don't think that this will help wikipedia's quality standards. By accepting the proposal we would be able to compare the quality of our articles with the quality of the BT encyclopedias, not based on our millions of articles (which is already our strength in the comparison), but based on THEIR articles (that means their strength). By trying to beat them there, we would show the whole world, that wikipedia is the best encyclopedia in every field. --Mad melone 06:16, 20 August 2009 (UTC)[reply]

Just a few more ideas, feel free to discuss

Mad_melone -- 08:48, 16 August 2009 (UTC)[reply]


Some proposals will have massive impact on end-users, including non-editors. Some will have minimal impact. What will be the impact of this proposal on our end-users? -- Philippe 00:07, 3 September 2009 (UTC)[reply]

If you understand end-users as readers, there will certainly be an impact. Readers are looking up stuff in the different wikipedias because they use it as a free source of knowledge - but they may have heard that not everything on wp can be trusted (which is true). Or they were told that wp doesn't qualify as an citable source (which is also true). or they might think that they can't trust wp because it is for free and "if it's free, then it probably isn't worth any money". but by showing everyone, that the articles in the wikipedia are on a higher level than the very same articles in other encyclopedias, users will trust wp not only in these "certified" articles, but will also acknowledge the standard of quality of wp in other topics, that aren't covered by other encclopedias.

Last point is that topics covered by printed and sold encyclopedias are most times the important and central topics of knowledge, whereas wp also covers detailed knowledge about topics of special interest. The goal of wp should be to lead among the central topics and there we have a wonderful tool of comparison level - printed encyclopedias. Use it and beat them and they won't be needed anymore, as everyone sees, that the intelligence of the masses beats the intelligence of some professionals. --Mad melone 22:34, 5 September 2009 (UTC)[reply]

Comparison is tricky

Comparison with "real" printed encyclopedias is a tricky subject. We certainly invite ourselves to comparison and our articles are oftentimes very reminiscent of printed articles. The issue with comparing ourselves to printed encyclopedias is that we are a constantly changing and evolving encyclopedia, for better and for worse.

To present some obvious problems with this, Wikipedia has such policies as "Neutral point of view" and "original research". Printed encyclopedias throughout time did not care about these. Original research is an easy case: printed encyclopedia articles are often written by experts that can use their own research to back up scholarly claims, while, for obvious reasons, Wikipedia editors are forbidden from doing this. Neutral point of view is a relatively new concept in an encyclopedia. Older encyclopedias are filled with overt points of view, and I feel their writers wouldn't say that a "neutral point of view" would be a valid objection to their work, after all they were the ones who wrote it. Flip through the freely-available 1911 Britannica for some POV ridden articles. So basically, we have our niche. We are similiar to printed encyclopedias but our final product cannot resemble them fully due to the structure of our editing body. Themfromspace 03:29, 12 September 2009 (UTC) But does that constrain us from comparing two articles about the same topic? We should be able to say which one is better in terms of giving information. Articles in "real printed encyclopedias" normally aren't referenced as we usually do it in wikipedia, this is also not a quality issue when comparing the informations given in two texts. --Mad melone 06:28, 21 September 2009 (UTC)[reply]