Wikipedia:Articles for deletion/Homogeneity (statistics)

The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.

The result was Keep and cleanup. – Avi 04:18, 31 July 2006 (UTC)

[edit] Homogeneity (statistics)

This article seems to be totally unintelligible to some mathematically astute WP editors, I have serious concerns about the history of edits of this article (some major contributions by 'Cruise' or 'David Cruise' and links to external sites which mention 'D. Krus') In its current state, the article is quite possibly OR, and fails to give any sensible definitions of the concepts it claims to be about. Reason the page should be deleted Madmath789 22:07, 19 July 2006 (UTC) and Salix alba (talk) 16:13, 20 July 2006 (UTC)

This was previously listed as Wikipedia:Articles for deletion/Homogeneity however the content of the page in question was cut and pasted into Homogeneity (statistics) and Homogeneity was turned into an uncontriversal disambig page. It seemed simpliest to start a new AfD discussion here rather than have a possibly confusing one at the old AfD. --Salix alba (talk) 16:13, 20 July 2006 (UTC)

Content of previous AfD listed below

Conditional keep I'm not an advanced math guy, but the history suggests that this has been taken seriously by a number of editors and expanded. It would be good to get a statistics expert to take a peek at it. OhNoitsJamie ^Talk 22:39, 19 July 2006 (UTC)

• Weak Delete it sounds like an application of Homogeneous space to stat but my statistics classes weren't that good and they were a while ago now. --Pboyd04 01:23, 20 July 2006 (UTC)
• Comment I have moved the material that WAS at homogeneous to homogeneity and redirected homogeneous to homogeneity, and moved what WAS at homogeneity to homogeneity (statistics) and added a cleanup tag. That that is how the material should be apportioned among the article titles is obviously required by simple common sense. So a question arises: Should this AFD tag be put on homogeneity (statistics)? I've only superficially glanced at the material; it appears to be one of those things used in statistics applied to psychology that are better-known among statisticians in psychology departments than anyone else. More later... Michael Hardy 14:47, 20 July 2006 (UTC)
• Confused issue Hmm - I think that has somewhat muddied the waters, and I am not sure that moving an article in the middle of an AfD discussion was best thing to do. The material which I think is utter gobble-de-gook, and which caused the original AfD is now at: homogeneity (statistics). I am a reasonably competent mathematician, and have taught statistics for many years, but this article seems to be almost meaningless pseudo-maths, and has already been described by other editors of maths articles as "whacky" and "... like abracadabra ... ". What is the next step? should I tag homogeneity (statistics) as AfD? Madmath789 14:57, 20 July 2006 (UTC)
• Comment probably simplest to withdraw this AfD and start a new one for homogeneity (statistics). The two current votes could be copied across. I generally agrees with Hardy that Homogeneity should be a disambig page or a redirect to one. --Salix alba (talk) 15:50, 20 July 2006 (UTC)

• Week keep and extensive rewrite. Basically it looks like there is not much stats going on here. We have a set of fairly standard correlation matricies correlating the variables p,q,r. The authors have then taken examples of a few special cases where there is a logical relation between the variables, say p implies q and given examples of the resulting correlation matricies. The illustrations don't really illustrate the mathematics well. Some of the matricies look suspect, no examples of real data are given, just synthetic data, giving 0 for some correlation coefficients where you would really expect some small non zero numbers for real data.

There does seem to be a core of something worthwhile here, there is certainly important statistics concerning hierarchal statistics, which I brushed across in my PostDoc days.

I'm less certain that this material is correctly titled, or the concept of homogenity in statistics is adaquatly represented by this page. --Salix alba (talk) 19:59, 20 July 2006 (UTC)

• Comment I have just contacted User:David Cruise who seems to be still around. (Igny 14:18, 21 July 2006 (UTC))

I looked at Cruise's edits and found out that he made significant contributions to a number of articles (see User:David_Cruise/Contributions), some of which contained references to Krus' publications. He decided to leave Wikipedia around February and deleted many of his contributions. Since then his edits were mostly minor (or so I think) (Igny 17:39, 21 July 2006 (UTC))

There seem to be 2 different accounts here: Cruise and David Cruse, and I have some serious concerns about another contribution: Canonical analysis, which 'looks sensible' (and has references - but almost all to sites connected with David Krus of 'Cruise scientific') - until you try to read it very carefully! Madmath789 17:49, 21 July 2006 (UTC)

I am going to raise the issue of Cruise's contributions to math articles at Wikipedia_talk:WikiProject_Mathematics (Igny 18:07, 21 July 2006 (UTC))

• Very weak keep I have no idea, having read it twice, what the objects called "lattices" are or why they are being called that; my best guess is that they are functions to {0,1}. This also looks like a miscellaneous collection of information. But I think those problems can be taken care of by normal editing; if normal editing isn't working, I would change my !vote. Septentrionalis 14:50, 21 July 2006 (UTC)
• Neutral. I can't make heads nor tails of it; however, I just use statistics in my work. Can anyone confirm whether Psychometrika is a legitimate journal, and whether this concept is actually used in that manner in that journal? In any case, if kept, I'm going to add it to category:Pseudoscience, it doesn't describe why the technique might be used. — Arthur Rubin | (talk) 18:11, 21 July 2006 (UTC)

  I have confirmed that Psychometrika is a legit journal. It is not available as an e-journal from my university library, so I would have to physically go to the library to check the references. Link here. --Aguerriero (talk) 20:59, 21 July 2006 (UTC)

According to Google Scholar [1], Krus has many publications in the fields of Psychometrics, Psychology. Could it have sense as homogeneity(psychometrics)? (Igny 19:45, 21 July 2006 (UTC))

• Delete unless some-one completely rewrites the article in such a way that (a) it makes sense to at least mathematicians and (b) starts with a definition of homogeneity in statistics that has some understandable relationship with homogeneity in general. If it's pseudoscience, then why should we keep it? --Lambiam^Talk 19:00, 21 July 2006 (UTC)

Comment a quick google finds An explanation of the statistics used in the Meta-analysis: A meta-analysis combines the data from several studies about the same subject. Homogeneity measures the differences or similarities between the several studies. which I think is an acurate definition of the term (well it seems reasonable to me). Taking this definition the article begins to have more context. If we have two studies which largely agree then will have logically consistant data matricies.

• Comment I don't know this topic. But I do know that (1) statisticians in psychology departments know lots of techniques relevant to the application of statistics in psychology that are seldom seen outside psychology departments, and (2) I suspect that this is about assessing consitency of responses to surveys. Michael Hardy 23:44, 21 July 2006 (UTC)

• Keep Walk into a math library, pick up a journal of a branch you do not specialize in, and it is likely (very) to look like gobbledygook. Ditto for electrical engineering, chemistry, or virology. The appearance of gobbledygook to a non-specialist should not be a criterion for deletion. The journals and authors cited are certainly not made up (Chronbach! He wrote the book used in a tests and measurements book I had decades ago.) This article appears to be perhaps written by an expert in a subfield. Meta-analysys is real. However, the article is way less approachable than the article on, say analysis of variance, which is far less technical but of far greater application. I agre one can have more than a nodding acquaintance with stats, read the article, and still not be sure what it is about. More definition and some simple explanations would be good.Edison 01:15, 30 July 2006 (UTC)

• Delete, per arguments made by Edison, immediately above. In fact, when I go into a math library, I make a point of picking up a journal in an area I do not specialize in, and I make a concerted effort to read and understand it. And, amazingly, I often can understand it, because it is usually well-written. If the topic is particularly difficult, good articles do provide overview sketches for the required background. The problem with this article is not that it deals with a difficult, abstract topic, but rather that it is so poorly written, that one cannot make sense out of it. linas 18:03, 30 July 2006 (UTC)

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.