Talk:Standard score

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Spoken Wikipedia
WikiProject icon This article is within the scope of WikiProject Spoken Wikipedia, a collaborative effort to improve the coverage of articles that are spoken on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
 
WikiProject Statistics (Rated C-class, High-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.

C-Class article C  This article has been rated as C-Class on the quality scale.
 High  This article has been rated as High-importance on the importance scale.
 

Other scores[edit]

  • Just one quick addition/question. Standard score in statistics is not related only to (z) scores. "It should include (z), (t), deviation (IQS), stanine, sten scores, (others...)"(Drummond, 2004). —Preceding unsigned comment added by 65.182.95.67 (talkcontribs)
  • Where can we add the value range standardization following

 y = \frac{x - minimum}{maximum-minimum} - herr_blaschke 14:22, 3 Jan 2008 (UTC)

Modified content[edit]

I added stuff about standardization and standardized because a while back I created an article called "Standardized" that I'm now trying to get rid of (merging, then REDIRECTing -- note that the only page history Standardized had was the initial page creation by me on March 9, 2004 — I figured that wasn't important to keep).

I removed the following sentence because it was rendered mostly redundant by my discussion of comparing 2 students' exam scores.

A standard score is a way of placing a raw score in context. It is often used to compare test results within and between groups, and especially with reference to a norm group. - dcljr 21:11, 9 Aug 2004 (UTC)

I also removed the "external link" section and moved it to main text, and changed "\overline" back to "\bar". Not only do I prefer the look of "/bar", but it makes a little more sense, since the symbol is usually read, "x-bar". - dcljr 21:22, 9 Aug 2004 (UTC)

Question about normalising[edit]

How is normalising data different from standardising data? The Normalization (statistics) pages doesn't give much (enough) information about this. 129.97.80.128 21:51, 7 April 2006 (UTC)

Symbols question[edit]

Hi. Could somebody explain to me why sometimes symbols (like the Greek letters named mu and sigma) appear in pages while other times there are what seem like LaTeX codes (like \mu and \sigma)? I wonder why the codes appear, instead of their compiled output or the appropriate symbols. Thank you. Xinelo 11:11, 29 September 2006 (UTC)

Suggestion: Separate entry for [Normal Score]?[edit]

There should be a separate entry for [Normal Score]...maybe even a disambiguation page for [Normal Score], to separate [Standard Score]s or [Z-Score]s (same thing) from the [Normal Score]s of van der Waerden or Blom (altogether different from standard- or Z-scores...see http://www.statsdirect.com/help/data_preparation/normal_scores.htm or http://www.dnr.state.ak.us/ssd/whtest/sashtml/stat/chap47/sect17.htm for definitions.) -- —Preceding unsigned comment added by 68.33.196.22 (talkcontribs)

External Link Added[edit]

I added a link to a z-score table. z-score tables, though archaic, are definitely a part of the curriculum and should be mentioned somewhere. 68.19.125.93

That link is a commercial link and does not belong in Wikipedia. Please find a non-commercial link with the same or similar content. -- Chris53516 20:53, 25 October 2006 (UTC)

Granted, but, as the site itself admits, this is not proprietary material. (edit: I fail to see how this is not a fair use, considering, but I will admit that wiki has far more stringent policies on fair use than is normal, and that I may be entirely incorrect due to a faulty interpretation of wiki policy) Anyway, its not worth my time to comb google for a better one, so I'll leave it in the talk page and someone else can find something better. http://www.statsoft.com/textbook/sttable.html#index 68.19.125.93

Oh, it's fair to use, it's just that the website is a commercial entity, and linking to it may be seen as advertisement. See Wikipedia:External links for more info. -- Chris53516 21:37, 25 October 2006 (UTC)

Population versus Sample[edit]

"A key point is that calculating z requires the population mean and the population standard deviation, not the sample mean or sample deviation. It requires knowing the population parameters, not the statistics of a sample drawn from the population of interest."

My professor (Ph.D Electrical Eng.) says this is wrong, ether type can be use... Anyone care to offer up a third opinion or clarification?


The official definition of z involves the difference between a particular score and the population mean. So in order to calculate z you must know the population mean. Knowing the population mean is very easy with standardized testing because the population is all of the tests administered. For most natural phenomena, knowing the population parameters such as the population mean is practically impossible which is why sampling is done to try to estimate the population parameters by calculating sample statistics. If the sampling is done properly and the population has one of statistically correct types of distributions then the sample statistics can be used to approximate the population parameters. z score calculation requires that the population distribution be a normal curve.

So basically your professor is sort of right and sort of wrong. He is right if the population of interest is normally distributed and the sample statistics are good approximations of the population parameters. In many cases people will make that the distribution is normal and that the sample statistics do approximate the population parameters well enough. The nice thing about have the entire population such as the results of a standardized test is that the researcher can calculate the population parameters and can also check that the distribution is actually normal.

see http://mathforum.org/library/drmath/view/68384.html

Richardelainechambers 02:47, 22 April 2007 (UTC)

Spell Check[edit]

The legend underneath the bell curve says "cummulative precentages".--Clinton57 (talk) 00:46, 1 January 2008 (UTC)

Totally incomprehensible jargon to this layperson[edit]

... and I say that as a Ph.D. who received good grades in her two graduate statistics courses. Is there anyone here who can decipher and "normalize" (sorry! :) the dense math-speak in this article to be a little more accessible? There's only so far it can be simplified, of course, being a mathematical concept, but surely it doesn't have to be this hard. As a guide, I clicked on this wikilink from an article that mentioned that the movie criticism site Metacritic is different from Rotten Tomatoes in that it uses normalized scores, so that may give you a clue as to how non-mathematical a person's motives might be in exploring this article. Lawikitejana (talk) 05:17, 8 February 2009 (UTC)

z-scores are not just for normal distributions[edit]

I think this article should mention that z-scores are applicable to ANY distribution where the population mean and standard deviations are known. All a z-score is, by definition, is the number of standard deviations away from the mean. It does not only apply to the normal distribution! This article intertwines z-scores and normal distributions which only leads to more confusion about z-scores. Z-scores make sense in uniform distributions, triangular distributions, and anything else you can think of. —Preceding unsigned comment added by ChevyC (talkcontribs) 16:35, 11 July 2009 (UTC)

Capitalisation[edit]

I came to this article to check to see whether "Z-score" should have a capital "Z", or a lower case "z". Both are used in the article, away from full-stops/periods. Is there a correct notation? Does the capitalisation refer to population, rather than sample properties?

Cheers! Buzwad (talk) 10:33, 6 May 2010 (UTC)

Table change[edit]

Hi, can someone please change the first row of the little table with Z scores to be 0.68ish and 1? I stopped in because I'm reverse engineering Excel's NORMINV() function (this version of Excel doesn't have function help) and I'm too busy to figure out how to update a wiki page (I never have before). The other three Z scores look right (the 95%, 99%, etc), but it looks like someone typed 0.68 and 0.67 or something accidentally, rather than 1 (68%ish is 1 std dev from the mean).

Yes check.svg Done Qwfp (talk) 19:22, 21 March 2011 (UTC)

Usefulness of the standard score?[edit]

Hi there, I read this article with interest. However, what seems to be missing to my (non-mathematical) mind is a section on the actual use and usefulness of the standard score. My very limited understanding is that this score is useful because it tells you how meaningful your results are: the more they deviate from (expected) normal distribution, the more likely it is that your data has something special to tell you. Now I am sure this is somewhat reductive or just part of the story, but more on this would be important, I think. -- Cheers, --PhilipWinter (talk) 11:16, 19 June 2011 (UTC)

There is an obvious lack of reference to the use of standard score with a standard normal table. There is no better use than I can think of. Stephen Charles Thompson (talk) 15:01, 7 May 2012 (UTC)

Help in correcting the diagram[edit]

Diagram showing the cumulative distribution function for the normal distribution with mean (µ) of 0 and variance (σ2) of 1. The prediction interval for any standard score corresponds numerically to (1-(1-Φµ,σ2(standard score))*2). For example, a standard score numerically being x = 1.96 gives Φµ,σ2(1.96)=0.9750 corresponding to a prediction interval of (1-(1-0.9750)*2) = 0.9500 = 95%.

As a perceptive user noticed [1], I had mixed up variance and standard deviation in the diagram at Standard_score#Percentile_ranks_and_prediction_intervals. The text may be fixed now, but it seems that a correction in the diagram itself is necessary as well. I know how to easily change the letters in the image in Inkscape, but before I do so, can we conclude exactly how it needs to be changed? Is it enough to remove that little 2 above the σ at the top? Mikael Häggström (talk) 11:51, 14 November 2011 (UTC)

I've made an entry at Wikipedia talk:WikiProject Statistics for this issue now. Mikael Häggström (talk) 05:38, 30 November 2011 (UTC)

and when it's multivariate?[edit]

how dow you standardize when it's multivariate? 134.106.106.34 (talk) 16:36, 19 March 2013 (UTC)