Wikipedia talk:Countdown deletion/What constitutes significant improvement?

What constitutes significant improvement?

While each voter will judge whether any improvement is significant or not, it must be clear that the test is not whether there have been any improvement, but rather whether there has been significant improvement. OK, simple cleanup is not significant (presumably, if the article was good enough before the cleanup, it should not be threatened with deletion anyway), but writing brilliant prose, providing context and adding sources is significant. How much prose? How many references? Would an extra sentence be significant? Or one external link? -- ALoan (Talk) 12:06, 6 Apr 2005 (UTC)

There is somewhat of a circularity, yes. How can we talk about voting on "significant" improvement if we define "significant" as "what you think is significant"? As you say, don't we need to specify it some more? Shouldn't we provide some bottom line?

This immediately raises a simple question: what bottom line do you have in mind? Are we going to write "one external link is not significant improvement"? Is that debatable? If it's not, why are we adding instruction creep? If it is, why are we putting it in there? To make it foolproof? But then we'll just get better fools who demand we also lay down how many words added are significant, how many improvement points we assign to a reference (0 for weblog, 1 for website, 2 for website with high Alexa ranking, 3 for book...) That way lies madness.

It boils down to this: if you think "significant" is too ill-defined in the present version (and I'm not saying it couldn't be made a bit more explicit without immediately falling into the instruction creep trap), go ahead and edit! This proposal is not written in stone. The version that's there at any time is not final. If we break it, that's fine, we can just un-break it afterwards. JRM 14:42, 2005 Apr 6 (UTC)

I think that 'significant' may be a red herring here. The key issue is whether the article is sufficiently improved to merit survival. Each person passing comment is making a judgement about whether the article has been adequately improved; has the article changed enough to satisfy the individual making the judgement? --Theo (Talk) 15:41, 6 Apr 2005 (UTC)

Good - that is the point that I was trying thrash: it is a personal test of "significance" (although this means that some people could find relatively inconsequential edits "significant"). -- ALoan (Talk) 16:56, 6 Apr 2005 (UTC)

I agree. "Significant" suggests an objectivity there is not. I don't like it that we've added "significant" to the vote options, however. This is obviously not akin to "strong" on VfD, expressing that the voter believes policy fundamentally agrees with them, because there is no real policy on improvement. So "significant" would instead just strengthen the vote—but really, this complicates makes vote tallying for no obvious reason. Does a vote of "not significantly improved" mean I agree it has improved somewhat, and that my vote may be overridden if enough people vote "improved"? I've removed that for now, but feel free to put it back in if you can add clarification to it. JRM 18:15, 2005 Apr 6 (UTC)

Sorry, the point is whether a voter should be deciding whether an article was improved or not, or significantly improved or not (I was not seeking to have four voting options, "Improved", "Not Improved", "Significantly improved", "Not significantly Improved"). Perhaps it is a red herring, and people will vote how they want anyway, but I think a vote "Improved" should indicate that the voter thought that the article was significantly improved (not merely improved). Although the test is still subjective, it is clearly a higher hurdle. -- ALoan (Talk) 19:01, 6 Apr 2005 (UTC)

Would 'material' be a more neutral word than 'significant'? The voter is noting that the article has undergone improvement that is relevant to this process (with format improvements being irreleveant in this context) rather than assessing the level of importance of the change. --Theo (Talk) 19:33, 6 Apr 2005 (UTC)

The only thing that is slightly jarring to me is the unusualness of the word "material" in this meaning, which is repeated throughout. Maybe that's just a consequence of my personal idiom. How about "substantive", which is subtly different yet subtly similar?

That we're talking about what words to use may also suggest we're still not quite where we need to be in terms of clarity. It's better to explain than to rely on precise meanings of individual words. I don't see an obvious way of improving what we have, though. If it boils down to whether to use "significant", "material", "substantive" and whatever anyone else can come up with, we might do with a quick straw poll—but maybe that's taking things too far. JRM 22:17, 2005 Apr 6 (UTC)

This use of the word 'material' is legal. It means relevant to the matter at hand; in this case, 'relevant to the decision to delete'. To my mind 'substantive' means 'essential', 'fundamnental', 'heavy duty' or 'pertaining to the heart of the matter', which is rather tougher than mere relevance. Loosely speaking, a material change affects the effect of something (it changes the way that the thing works); a substantive change affects the nature of something (it changes what the thing is). The choice of such words seems important to me because they establish the tone of the policy and so influence the approach taken by the implementors. Choosing the most appropriate word and then explaining it seems like the best approach to me—all I am changing here is the key word. --Theo (Talk) 22:50, 6 Apr 2005 (UTC)