Talk:Semivariance

Can you please tell my what the POV issue is? It will never be fixed if the tag is left unexplained; & if the issue is the controversy section, well, then its no longer POV....Bridesmill 16:25, 5 July 2006 (UTC)

Sources

I slapped a POV on, as 3 of the 4 source cited are not actually the articles they say they are, but ratehr unpublished criticisms. Forgive me if I'm wrong, but that is patently WP:OR, and either needs to be quickly fixed or removed. In addition, the way these cites are mis-attributed is dishonest to the max; curious, considering the user in question is slamming geostats and semivariance on the grounds of it being inherently dishonest....Bridesmill 00:12, 6 July 2006 (UTC)

Oops! I goofed!! Just copied the references to these textbooks linked to my retro-reviews but have made corrections. Don't read what I wrote. Or try to forget if you did! Please peek at Clark's Practical Geostatistics, which is posted on her website. I wouldn't even throw out an ugly baby with the bathwater. What I do want is reunite each distance-weighted average with its variance and get rid of an ugly science. Surely, you're not the only Wikipedian who wants to bring sound knowledge and scientific integrity to the world. JWM. --Iconoclast 16:08, 6 July 2006 (UTC)

Right, wrong, or otherwise it is not your job here to get rid of anything or create anything; just to deal with what is known. WP:NOR and all that. If your work is published in a peer-reviewed work, then it's admissible. Until then, nobody cares how much you dislike the field or what you think is wrong with it. Again, you seem to be taken in by very passionate beliefs, not least influenced by various scandals and heinous abuses. Sometimes, though, it works - as in applications to criminology. So until your work is 'published', it has no place here. Sorry.Bridesmill 16:47, 6 July 2006 (UTC)

Howdy Bridesmill, I decided to post some peer-reviewed work. "Abuse of statistics" was reviewed by IAMG's present President himself.In contrast, Stanford's Journel thrashed "Precision Estimates for Ore Reserves" because we applied "classical Fischerian [sic!] statistics" whereas Erzmetall praised and published it. Do you want to know what Journel wrote about our paper? Visit spatial dependence and click on Journel's letter! Do you agree or disagree with the junk science of interpolation without justification? Do you want to know what JMG's Editor wrote to me? Visit my website and look under Correspondence! Surely, you'll have more suggestions on how to become a good Wikipedian. JWM. --Iconoclast 22:41, 6 July 2006

deleted nonsense

I just deleted this statement:

In mathematical statistics, sets of n measured values give df=n-1 degrees of freedom whereas

Quite aside from the fact that

df=n-1

should have been typeset as

df = n − 1

the statement itself is idiotic nonsense. Michael Hardy 21:38, 17 December 2006 (UTC)

Variogram, semivariogram and others

I first cite: The semivariance is calculated in the same manner as the variance but only those observations that fall below the mean are included in the calculation. It is sometimes described as a measure of downside risk in an investments context. For skewed distributions, the semivariance can provide additional information that a variance does not. This cannot be correct. I have removed it. --Bachmai (talk) 15:39, 23 February 2008 (UTC)

I have described the context between variance and "semivariance" in Bachmaier and Backes (2008) cited in the article. This context is not yet written down in this article. (Martin Bachmaier) --Bachmai (talk) 15:39, 23 February 2008 (UTC)

Someone wrote: The statement that only measured values below the mean are included in the semivariance makes no statistical sense (see Ref 4). Clark, in her Practical Geostatistics, which can be downloaded from her website, proposed that the factor 2 be moved for mathematical convenience and berates those who refer to variograms rather than semi-variograms. The first part is o.k. And I think it's me who removed what you criticized. But as I wrote in my paper cited in the article, to which one reviewer said that it blaims professionals, variance and variogram are the only terms that should be used. The problem is that the variance formula which has the factor 1/2 and is the basis of the variance at given separation distance (unfortunately called semivariance) is less known. See the German empirische Varianz, where this alternative is described. My paper confers all standpoints. It needs to be known that the semivariance is nothing else than a variance restricted to certain pairs with a distance of h (that has nothing to do with negative deviations or similar). --Bachmai (talk)

Clark can be right when she consequently calls the squared sum the variance of the differece of z values. However, the problem is that people only say that it is the semivariance without saying of what. It's me who has added it now. People say that gamma is the semivariance of yield values for example. However, gamma is the semivariance of yield value differences.--Bachmai (talk) 19:51, 23 February 2008 (UTC)

Therefore:

gamma = 0.5 sum... variance of z at given h = semivariance of z differences

2gamma = sum... (according to Clark) = twice the variance of z at given h = variance of z differences (which is usually not said)

--Bachmai (talk) 19:51, 23 February 2008 (UTC)

Before I wrote here, gamma = sum... was defined. (gamma instead of 2gamma!!!) This is nowhere (perhaps in Clark, which is one of a few books which I have not looked at). These people normally say 2gamma = sum... --Bachmai (talk) 19:51, 23 February 2008 (UTC)

The main problem is that "semivariogram" is endemic in the general scientific literature, and not just in one field of science. I can mention two additional books:

• Stein ML (1999) Interpolation of Spatial Data. Spinger.
• Schabenberger O., Gotway CA. (2005) Statistical Methods for Spatial Data Analysis, Chapman&Hall/CRC.

The statement "Bachmaier and Backes (2008), who discussed this confusion, have shown that..." should at least be toned down to something like "say they think that". See also variogram. Melcombe (talk) 15:52, 17 April 2008 (UTC)

Comments for improved clarity

The header "The neutrality of this article is disputed..." appears to be resolved. I don't see anything remaining in the article that appears non-neutral. Not knowing anything previously about the topic, I'm not going to make the final judgement call to actually remove it...

A simple illustrative example with, say, 10 data points would help to make it a little more readily accessible to those of us who haven't looked at this previously.

Is there a reason that "Semideviation" doesn't link here. Is Semideviation the square root of semivariance?

Is this the same concept that Markowitz explored originally in his 1950s work on capital asset pricing models? I have heard he originally focused on semi-deviation to capture risk, but eventually settled on standard deviation in his now famous CAPM work. I got here because I was curious about this, but I'm not sure whether this is the same concept.

At the top is says "the empirical semivariance is...". Does this imply that a non-empirical definition exists, for example, given a continuous distribution P(x) is there a notion of the semivariance of this analytic distribution? If so, the dof controversy would be a bit more concrete also, since we could talk about whether the empircal semivariance is an unbiased estimator of the underlying semivariance. Otherwise, it sounds like vague handwaving. —Preceding unsigned comment added by Ldc (talk • contribs) 18:41, 19 May 2010 (UTC)

The variables in the equation are not described, making it difficult for those without a statistical background to understand. — Preceding unsigned comment added by JWS (talk • contribs) 21:17, 6 July 2012 (UTC)

Other definition of semivariance

Someone wrote above six years ago

I first cite: The semivariance is calculated in the same manner as the variance but only those observations that fall below the mean are included in the calculation. It is sometimes described as a measure of downside risk in an investments context. For skewed distributions, the semivariance can provide additional information that a variance does not. This cannot be correct. I have removed it.

Actually it is exactly right for a different usage of the term. I'll put in something to disambiguate this when I get a chance. Loraof (talk) 17:21, 27 October 2014 (UTC)

Intro

Someone should add an intro explaining what the semivariance is useful for (e.g. with an example). Ben Finn (talk) 11:14, 27 August 2015 (UTC)

External links modified (January 2018)

Hello fellow Wikipedians,

I have just modified one external link on Semivariance. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20020424165227/http://www.geovista.psu.edu/sites/geocomp99/Gc99/044/gc_044.htm to http://www.geovista.psu.edu/sites/geocomp99/Gc99/044/gc_044.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 10:10, 22 January 2018 (UTC)