Talk:Homogeneity (statistics)

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Statistics (Rated Start-class, Low-importance)
WikiProject icon

This article is within the scope of the WikiProject Statistics, a collaborative effort to improve the coverage of statistics on Wikipedia. If you would like to participate, please visit the project page or join the discussion.

Start-Class article Start  This article has been rated as Start-Class on the quality scale.
 Low  This article has been rated as Low-importance on the importance scale.
 


Confusing tag[edit]

I have tagged this article as confusing for a number of reasons:

  • "coefficient of homogeneity" mentioned but no precise definition is in evidence.
  • "..depart from this ideal lattice form.." - not clear what is meant by 'ideal lattice form'
  • Fig. 1 - what on earth does this represent? what are the scales on the axes? etc.

I have many other concerns. Madmath789 13:15, 15 July 2006 (UTC)

THe diagrams on this page do not relate to the 'truth tables' associated with them, and have absolutely no explanation of why they are relevant - unless anyone can see a good reason for keeping them, I propose deleting them. Madmath789 08:01, 4 August 2006 (UTC)

Madmath, over the long time, censorship does not work. Soon all you'll see will be a mirror reflection of your face distorted by hatred. David Cruise —The preceding unsigned comment was added by LCF (talkcontribs) .

I have absolutely no hatred, and I am not interested in censorship. If you can give any mathematical justification for the the correctness and usefulness of the diagrams I removed, please feel free to do so - I will be more than happy happy to see any additions which make this article a positive contribution to the encyclopedia! Madmath789 23:35, 19 August 2006 (UTC)

Nomen omen[edit]

The usefulness of figures you deleted is obvious to anyone who can mentally expand a series. For instance, to fully describe the Figure 1 would necessitate a table with 32 rows. Anyone who is not mathematically challenged can expand a pattern in the table adjacent to Figure 1 from 8 by 3 to 32 by 5 matrix. However, after you deleted the relevant figures, I attached a link to an article which fully describes the discussed patterns and a link to an article describing the relevant truth tables. However, your "knee jerk" action was to immediately obliterate this second link that, in many ways, may be helpful to people with only a cursory knowledge of Boolean algebra to fully follow the subsequent discussion. What propelled you to do that I find hard to judge, as it does not even enter my mind to censor and to destroy someone's bona fide work. David Cruise —The preceding unsigned comment was added by LCF (talkcontribs) .

The figures lack the captions to describe what they represent. What are the x, y and z axis? How do these relate to the truth table, which has three variables p,q,r, or the text which describes five variables? Whats the non-standard notation (p 1 q), should that be (p | q)? --Salix alba (talk) 09:57, 20 August 2006 (UTC)

Salix, the operator you suggest, the Sheffer's stroke, is sometimes used for indirect expression of tautology within the compound statements. There is no consensus about tautology's operator. Often you'll find the infinity sign, though this sign has no appealing counterpart for contradiction. I myself prefer the operators 1 and 0 for these functions, patterned after the matrix algebra notation for unit [1 1 1 1 ...] and null [0 0 0 0 ...] vectors. However, the Wikipedia reference Mathmath substituted (even though a well written one) does not list the operator for tautology and defines tautology only indirectly LCF 18:31, 21 August 2006 (UTC)

I also beleive before we allow these diagrams back, that we need to see some sources which show that these things are in common usage in statistics. So far, the only sources I can find for them are all connected with David Cruise, D. Krus, Cruise Scientific, or Visualstatistics.net - without some independent sources, they must surely fall under the banner of original research? Madmath789 10:33, 20 August 2006 (UTC)

Madmath, on Wikipedia, the last refuge of the uninformed is the OR. In the link I attached are many published references in major journals which are well known to psychometricians, but which you likely did not bother to consult before elevating yourself into the position of the arbiter with the statement like "before we allow." The other day I left a message at the Salix Alba's talk page and there was a letter of a little child about the treatment of his grandmother on Wikipedia. Even a little boy can see that on Wikipedia, the emperor has no clothes. LCF 17:57, 21 August 2006 (UTC)

To sum things up, I discovered Wikipedia by chance and thought that it was an interesting idea to disseminate knowledge. At the beginning I had a good fortune (or, in the long run a misfortune?) to work with two reasonable and helpful administrators, Mark Sweep and Michael Hardy. At that time I had no idea that Wikipedia, as the little boy on the Salix Alba's talk page observed, is the domain of "yound aggressive males," where not logic and resson, but a stubborn persistence, vile accusations, and packs of predatory animals rule. Ever since I tried to extricate myself from this hostile environment by using several strategies, even nominating my own articles for deletion, but to not avail. By some sort of perverted reasoning, once you enter, you are subject to abject abuse with no way to leave. Salix Alba will probably say that I have been forewarned by the opening statement that people who do not like to be edited should not publish here. Edited? What a misnomer. The correct warning should be a verse from Dante's Inferno - "you who enter, abandon all hope." LCF 17:57, 21 August 2006 (UTC)

You're possibly using notation and terminology which is common among psychometricians, but is considerably different from that used by logicians or statisticians. (Yes, I am implying that psychometricians are not statisticians. Make of it what you will.) You need to define and source the notation in an appropriate article (psychometrics)?) before you use the notation elsewhere. — Arthur Rubin | (talk) 18:47, 21 August 2006 (UTC)

Circulus vitiosus[edit]

Arthur, I did that, but Madmath deleted it. LCF 20:19, 21 August 2006 (UTC)

When? I'll check on it, and reinsert it if it seems appropriate. — Arthur Rubin | (talk) 20:34, 21 August 2006 (UTC)

Arthur, that link is about seven steps back in history, However, more interesting is your assertion that psychometricians are not statisticians. This is usually framed as data analysts are not statisticians, a controversy that was briliantly summed by Cooley and Lohnes (1962, Wiley) in the preface to their classic Multivariate Data Analysis. Needless to say, data analysts are the leading force in the computer age with algorithms by Cooley, Lohnes, and many others, includine me, embedded in statistical programs as SPSS, SAS and in NASA computer programs, trailblazing the path toward the stars.LCF 21:05, 21 August 2006 (UTC)

For what it's worth, in the late 1970s and early 1980s, the tracking code used by JPL for tracking deep space probes had serious statistical flaws. We could patch around it, but couldn't get funding to rewrite the code cleanly. So, yes, whoever designed that code was not a competent statistician, whatever type of data analyst he or she may have been. That error may have cost 5 years of progress. So don't talk to me about data analysts.
Having gotten that out of my system — the correct analogy is probably data analyst:statistician::engineer:scientist . There are purposes to all of them, but, just as an engineer is not a scientist, a data analyst is not a statistician. — Arthur Rubin | (talk) 21:21, 21 August 2006 (UTC)
And I can't find anywhere in the history of the psychometrics article where that notation is referenced. It really needed to be there, rather than in this article. — Arthur Rubin | (talk) 21:21, 21 August 2006 (UTC)

Wow, talking about logic and generalizing from few to many. I looked at your user page and to my astonishment found many issues we agree upon, alas you like cats and I like dogs. Thank you for an interesting discussion Arthur. LCF 22:28, 21 August 2006 (UTC)

What to do[edit]

The other day we talked about more interesting issues than what to do with this entry. Contrary to Madmath's opinion, I do not intend to wait for his permission to put back pictures he deleted, as a matter of fact I do not intend to put any more work to this article, as I learned the hard way that on Wikipedia, hours of work can be wiped out in seconds. However, Madmath's destructive editing distorted this article beyond belief that now indeed, as he claims, it does not make sense. I intend to convert the article into a stub which Madmath is invited to elaborate into an acceptable article. If this strange character instead of constructive editing will again resort to distructive edits and reverts, I intend to keep my options open. LCF 19:03, 26 August 2006 (UTC)