Talk:Twitter/GA2

GA Reassessment

Article (edit | visual edit | history) · Article talk (edit | history) · Watch

I am opening an individual reassessment following issues raised at the recent FAC. Thanks to improvements made since the review of the good article nomination (and in particular improvements made during the FAC), it seems to me that the main outstanding GA issue is the use of unreliable sources to support material in the article. In particular, for an article to be GA standard, the criteria require that

• 2a. it contains a list of all references (sources of information), presented in accordance with the layout style guideline;
• 2b. reliable sources are cited inline. All content that could reasonably be challenged, except for plot summaries and that which summarizes cited content elsewhere in the article, must be cited no later than the end of the paragraph (or line if the content is not in prose).

A list of possibly unreliable sources was provided by Ealdgyth at the FAC. Some of these examples are clearly GA issues: for instance, SuicideGirls.com is not a reliable source for a direct quotation from a living person!

In each case the issue can be resolved in a number of ways:

• Argue that the source is sufficiently reliable for the material it is supporting.
• Provide (additionally or instead) a more reliable source for the information.
• Remove the unreliably sourced material and the corresponding citation.

For example, regarding TechCrunch.com, it may well be possible to use this as a reliable source for some of the material in the article, and some arguments for this were made at the FAC. However, the fact that reliable secondary sources use TechCrunch does not necessarily make TechCrunch itself a reliable secondary source: Reuters and Forbes may have checked the facts, rather than trusting TechCrunch to do it. Also, the fact that TechCrunch is successful (as shown by newspaper articles and hit counts) does not necessarily make it a reliable source. On the other hand, the GA criteria are not as demanding as the FA criteria. In particular, for material that does not require citation per 2b, the requirements on the source are not as high as is required at WP:FA. Note also that a source can be reliable enough for citation X, even if it is not reliable enough for citation Y.

I'm not going to delist right away. I'll check back on Sunday. If significant progress has been made, or a clear timescale for fixing the above issues is in place, then I can be quite flexible, as I'd prefer not to delist an improved article. Geometry guy 22:09, 15 June 2009 (UTC)

It looks like a few of the controversial links are to summary articles. I'll take a pass at a few of them and change them to links to the direct article they are quoting (ex. the Ruby and Starling article can be change to a direct link to the Starling website. Or just ditched since there's also a link to the announcement on Twitter's website).

For the TechCrunch citations, it seems that most of those are TC linking to someone else's report and providing a brief commentary. I'll change this article so that those citations point to the sources TC is linking to. While there may still be some controversy over how valid those sources are, this will remove the somewhat suspect TC from the picture. We can then do a review of those sources and see which ones stand as valid citations.

Regarding the SuicideGirls citation, I think that's valid since SuicideGirls is in fact the primary source; they are the ones who did the interview with him. While interviewing people isn't their primary business, it is something they have done before and the site is popular enough that (a) it's credible that they have access to him for formal interviews and (b) if the interview was faked, word of the fraud would have gotten back to Williams who would have contested it. Thus, I think their claim to have interviewed him is credible and their transcript of their interview can stand as a quotable source. I agree that if they presented a one-liner paraphrase from him in an article on another topic their credibility would be suspect.

Jopo sf (talk) 08:03, 17 June 2009 (UTC)

I just went through the list provided here of bad references and did a fair number of them. Mainly, I just removed the citation because there was already another there, or I removed the section it was with because it was poorly sourced and not notable. A few I left behind though:

• http://blog.compete.com/2009/02/09/facebook-myspace-twitter-social-network/ - Compete.com measures web traffic. The source is used as stating how much traffic different websites get. Therefore, Compete.com are surely a decent source in this situation.
• http://www.techcrunch.com/ - Jopo sf aims to sort this out, though I'm still of the belief that it's a decent source, at least in the contexts we've used it. It's written by a specific group of writers, not the populous at whole, whereas I felt earlier calls against its reliability were questioning who wrote it.
• http://www.webmonkey.com/blog/Twitter_Vulnerability:_Spoof_Caller_ID_To_Take_Over_Any_Account - I understand general concern but, in the context, I feel this suitably proves that such a vulnerability exists, and that the details of how to do so are correct. This source may be preferable, but is written by the subject.

Other than that, the list provided have been taken care of. Like I say, many of them were on double-sourced statements so simply removing the dodgy one worked out well. Greg Tyler ^{(t • c)} 08:28, 17 June 2009 (UTC)

In reviewing the TechCrunch articles, they are often part of a double-cite or paraphrases of someone else with some snark thrown in. So I didn't think it was worth resolving the discussion about how reliable TC is; it was easy for me to just go ahead and change the TC citations to be single-article or to point to the more neutral-toned source. In one case where there wasn't an easy fix and that was the TC article that said Twitter was moving away from Ruby on the front end. I yanked out that line since Twitter reports contradict this and I don't think TC has a sufficient reputation for research for us to cite their article that bases it's perspective on a claim of having found anonymous sources that contradict the official announcements.

I agree that Compete.com and webmonkey are sufficiently authoritative in the context they are cited.

Jopo sf (talk) 03:44, 18 June 2009 (UTC)

The article has been much improved and so I will close this reassessment shortly. A couple of comments: first, it is fine, in my view, to back up uncontentious material from a primary source with a less-than-stellar secondary, so some of the TechCrunch cites could have been kept. The secondary source provides independent verification of the primary and helps to ensure we don't write articles based on original research. I'm glad the SuicideGirls.com interview has gone: I agree it is a primary source, but if the interview were notable and added insight, then there should be a secondary source quoting it. Please always bear in mind that this is supposed to be an encyclopedia.

Double-sourcing is often helpful to readers: conversely, the webmonkey.com cite could be backed up by the primary; it doesn't have to be either-or. Such uses of primary and less than ideal secondaries will certainly be challenged at FAC, but clearly explained responses should satisfy the best reviewers. Finally, attributing a statement to a notable source is another useful tool: while Compete.com might be dubious as an unqualified secondary, I agree it is notable enough to use as an attributed source for web traffic.

Good luck in continuing to improve the article. Geometry guy 20:06, 21 June 2009 (UTC)

I fail to see why some editors are obsessed with TechCrunch being unreliable. It is. If it weren't called a 'blog', I doubt we'd have any objections. Computerjoe's talk 21:35, 22 June 2009 (UTC)

The issues raised about TechCrunch in this case were that they do insufficient fact checking and that they give more credibility to anonymous sources than is appropriate for encyclopedia content. There's no general issue with using blogs as sources as long as those blogs are sufficiently authoritative; note that there are a number of blogs referenced in this very article.Jopo sf (talk) 06:40, 26 June 2009 (UTC)

The section titled "Use in criminal proceedings" - The second paragraph in reference to the July 2009 libel suit: This is unfinished. Even though verifiable, it does not say how the case was disposed and therefore should be noted as such. Mnemnoch (talk) 06:41, 24 August 2009 (UTC)