User talk:Kokiri/WQA

Weird, two links show up in red... I've used copy & paste :-( Kokiri 18:30, 25 Nov 2003 (UTC)

---

By extrapolation, we have about SIXTY THOUSAND "real articles" in the English edition. That is awesome for less than three years of volunteer labor! I don't see why the fact that we also have many less mature entries is seen by some as detracting from that laudable achievement; cf Sturgeon's law... --Brion 19:18, 25 Nov 2003 (UTC)

Wikipedia Quality Assessment

I've done a Wikipedia Quality Assessment. Kokiri 18:32, 25 Nov 2003 (UTC)

Hi Kokiri, interesting survey. I fixed the links for you. Welcome back. Angela 20:23, 25 Nov 2003 (UTC)

An interesting read Kokiri. Thanks for taking the time to do that. More non-text media is probably something no-one is going to argue about. The other most major point you raise is the stubs issue. Stubs that _can_ grow but just haven't been written yet are great IMO. I am one (surely not) who gripes about stubs that _can not_ really grow because their domain is too small.. I think the needs of an encyclopedia reader would be better served by fewer, but longer, articles than the current trend as a point of 'policy' (meant in a vague sense!)... especially since we can redirect pages to particular parts of other pages using the [[...#...]] syntax. NB I am aware the wiki is not paper and that we don't necessarily want to slavishly copy Britannica. Now someone can show me where this issue has been raised ten times before! Pete 22:22, 25 Nov 2003 (UTC)

I note that when this topic was added to the pump it was not only without a heading (which was easily corrected) but was also marked as a minor edit (which I can't fix). So this edit is partly to bring the article to the attention of others who might have missed it because of this.

While I'm at it though, I'm not completely convinced of the methodology. My first impression is that there seem to be a lot of assumptions which the data itself contradicts. Is Peter III of Portugal really a stub, for example? It's marked as one now, possibly because of this survey. But what extra information should be added? I notice you say that Britannica has a short article. Is it significantly more complete than this one? It seems to me that this might be a classical case of an article that we want to keep despite its being very short.

If there's a problem with this article, it's that there seems to be some sort of battle over his name, Peter or Pedro. Andrewa 03:17, 26 Nov 2003 (UTC)

I've now reworked the Peter III of Portugal article on the assumption that it's essentially complete. I think it is. I'd now like to remove the stub warning, which I see you put there yourself, Kokiri.

But I have added very little information, and what I have added is entirely duplicated in other articles that were already directly linked to from this one. I've just rephrased things. Andrewa 08:48, 26 Nov 2003 (UTC)

Firstly, apologies for marking this as a minor edit. Secondly, I have mentioned in the introduction of the test that I don't consider it a good one myself. For example, as you might have noticed, I compiled the data rather quickly. As for the Peter III article, Britannica doesn't have more on him and I have removed the stub alert now. Please feel free to edit the assessment I've done (isn't this Wikipedia?). Maybe one day I'll do another one where I take more time. Kokiri 11:34, 26 Nov 2003 (UTC)

Very nicely done. It is very easy to read. And wonderfully interesting. I have often been disappointed and frustrated at the frequency of Rambot articles that come up when using the Random function. Sometimes I get three in a row.

I wonder if you could try the assessment again someday utilizing a different way to obtain your data. Rather than use the Random page feature, maybe you could ask a small sample of people to list specific things they would be interested in reading about. Afterall, wikipedia serves human readers, not Randomizers :) Kingturtle 10:46, 27 Nov 2003 (UTC)

I agree, it's an interesting survey. An additional interesting stat might be the number of articles that link to an article. That would support your comment that some articles seem to exist to support just one other.168... 20:26, 27 Nov 2003 (UTC)

I've been thinking about the Peter III issue coming out of my attempt to assess Wikipedia's quality of the articles. My feeling is that many stubs shouldn't exist as there just isn't enough for an entry. However, there are stubs we want to keep, just like the Peter III of Portugal. What about having a boilerplate text such as Mini Article to show that this article may be very short, but it's sort of finished (as finished as a Wikipedia article gets)? After all, Britannica does have very short qarticles too. Kokiri 09:50, 28 Nov 2003 (UTC)

I think this is an excellent idea, Kokiri. Stubs can be divided into 3 types: delete because of no real value, potential long articles, and whole short articles (non-stubs). A way of marking this last group so that they do not clutter up Wikipedia:Cleanup and Wikipedia:Votes for deletion would be a good thing. Bmills 09:59, 28 Nov 2003 (UTC)