User talk:Pdbailey

From Wikipedia, the free encyclopedia

You have new messages (last change).

Contents

[edit] Welch article

Sorry for the mix-up, when I saw the article, all that was there was a stub tag. On another note, leave messages for people on their talk pages, not their main user pages (I moved what you put on my userpage to my talk page). On yet another note, it's generally not a good idea to blank your talk page, since it gives other users insight to your contributions. [[User:Lachatdelarue|Lachatdelarue (talk)]] 03:35, 7 Nov 2004 (UTC)

Do you mean when you clicked on the article title (from another page) it took you straight to "edit this page"? If that's the case, no, the article didn't already exist. If that's not what you meant, could you explain what happened better? [[User:Lachatdelarue|Lachatdelarue (talk)]] 21:50, 7 Nov 2004 (UTC)
No problem, I'm here to help. [[User:Lachatdelarue|Lachatdelarue (talk)]] 00:17, 8 Nov 2004 (UTC)

[edit] Nuclear magnetic resonance

This edit is a bit garbled, but I can't figure out what it means so I can't fix it myself. --Yath 03:14, 13 Nov 2004 (UTC)

Thanks for pointing that out, I hope that does it Pdbailey 04:20, 13 Nov 2004 (UTC)

[edit] Standard deviation

Finally I see what you meant when you said my mention of normality came too late in the article. It looks as if you meant an assumption of normality is used in showing that a certain statistic is unbiased for σ2. But it's not. As long as

\operatorname{var}\left[\begin{matrix}X_1 \\ \vdots \\ X_n\end{matrix}\right]=\sigma^2 I_n,\,

where In is the n×n identity matrix (in other words, all of the variances are &simga;2 and the n random variables are uncorrelated (not necessarily independent!) then

E\left({1 \over n-1}\sum_{i=1}^n (X_i-\overline{X})^2\right)=\sigma^2,

where

\overline{X}=(X_1+\cdots+X_n)/n.\,

Neither normality nor independence is needed (although uncorrelatedness is). Michael Hardy 01:59, 5 Dec 2004 (UTC)

[edit] Richard Durbin

Please remember to sign your comments on article talk pages. DS1953 01:44, 24 Jun 2005 (UTC)

[edit] Lee-Jon

regards for the clarification on the NMR page. Lee-Jon 13:37, 9 September 2005 (UTC)

[edit] degrees Celsius

See this section of NIST SP811, clearly showing that you have jumped to the wrong conclusion. Gene Nygaard 14:34, 27 September 2005 (UTC)

[edit] Rudy Giuliani

Over at Talk:Rudy Giuliani on 4 October 2005 I asked for verification of the crime statistics you recently added to the article. If they can't be verified, they will be removed from the article. patsw 03:38, 9 October 2005 (UTC)

[edit] Image:MaxwellBoltzmann.png up for deletion

Image:MaxwellBoltzmann.png has been listed for deletion, since it has been obsoleted by Image:MaxwellBoltzmann.gif. --℘yrop (talk) 03:35, 19 February 2006 (UTC)

[edit] Generalized linear model

Hi, I reverted the page because I thought your edits made the topic less, not more understandable (to me, at least). You removed a good deal of material, including clarifying examples. Something funny is going on right now, though, because the page isn't parsing correctly (but if you link to it through the history, it parses OK...don't know what is happening.) Bill Jefferys 16:44, 19 April 2006 (UTC)

Hi again, I think that your re-edited article is much better and satisfies (more than satisfies) my objections to your original edit. Separating the main idea from the examples does the trick. I think that we can proceed from here. Bill Jefferys 02:38, 20 April 2006 (UTC)
Thanks for clarifying the GLM article a bit. Some of it is still unclear, but we will work on that. There's one sentence I still don't understand though, please see my comments on the GLM talk page.  :) -shaile 22:26, 20 April 2006 (UTC)

[edit] contact

I thought I'd mention that User:Drummond has been inactive since July. If he doesn't respond to your message or his account doesn't have an email address, he may also be able to be reached via his website, www.adaptivity.org .--Nectar 17:58, 28 April 2006 (UTC)

[edit] Reply from JQ (also on my talk page)

My idea would be to have a hierarchy of articles, starting with a top-level article on the General Linear Model, which would include results valid for the GLM and a brief taxonomy, including discussion of discrete v continuous dependent variables. Each element of the taxonomy would be linked to a more detailed article. In the case of discrete regression, there would be a general discussion, and a taxonomy linking to logit, probit and so on. JQ 10:53, 7 June 2006 (UTC)

[edit] Grinnell College

My justification for reverting your edits:

First of all, the quotation from Newsweek is just plain false (nowhere in the article does it state that) and grossly misrepresents what was really meant. Read the article here. The main criterion for the "best all-around" (not "best overall") college was "buzz." Second, Newsweek does not put out rankings as US News does and with the same authority. Third, putting a quotation from an article in a paragraph that has to do with rankings and accolades is misleading, suggesting that the Chronicle of Higher Education is ranking the wealth of schools when in fact it took that information from NACUBO's study (which is out-of-date for the year 2006).

The overall effect of your edits is to replace precise information with euphemistic information, making the article less informative, less authoritative and bulkier. (unsigned comment by User:Exeunt)

[edit] Linear regression

Please leave a summary of the work you have done. Thanks! Chris53516 13:59, 26 September 2006 (UTC)

[edit] SmackBot

Smackbot recently edited Radiation hormesis and changed "fact" to "Fact." in all but one instance (where it added a date). Is it really necessary to change "fact" to "Fact?" would it have made an edit if this was the only change? Pdbailey 20:58, 4 March 2007 (UTC)

No, but it's neater, no it would be unlikely to edit an article without an undated template, although if the template was dated by a third party between being identified and edited, this could happen in theory. Rich Farmbrough, 10:12 5 March 2007 (GMT).

[edit] spam @ organic farming

Hi pdbaily - Apologies for not starting a section re: my reversions at organic farming. I made the reversions to remove the addition of further spam, by a user who'd already been spamming in another article. I did explain my motivations in these reversions in the edit history. Apologies again for not making it more clear. MidgleyDJ 22:27, 23 March 2007 (UTC)

[edit] Statistician help needed

The WikiProject Vandalism Studies (Wikipedia:WikiProject Vandalism studies) just finished its first study and I was hoping that you being a statistician could help us formalize our findings. You can find our draft conclusions here [1]. Here's an excerpt of what we found so far:

The current study analyzed a sample pool of 100 random articles. Within these 100 articles there were a total of 668 edits during the months of November 2004, 2005, and 2006. Of those 668 edits, 31 (or 4.64%) were a vandalism of some type. The study's salient findings suggest that in a given month approximately 5% of edits are vandalism and 97% of that vandalism is done by anonymous editors. Obvious vandalism is the vast majority of vandalism used. From the data gathered within this study it is also found that roughly 25% of vandalism reverting is done by anonymous editors and roughly 75% is done by wikipedians with user accounts. The mean average time vandalism reverting is 758.35 minutes (12.63 hours), a figure that may be skewed by outliers. The median time vandalism reverting is 14 minutes.


Thanks. Remember 02:26, 26 March 2007 (UTC)

[edit] CAFE graph

The SVG format has many advantages over raster format for images, such as graphs, that comprise lines and solid colors. Since the graph you uploaded is of above-average quality, conversion seems hardly warranted. (Some people like to upload little 100 x 200 graphs that scream for conversion.) Yet, there are still some advantages to conversion; most of them are outlined on Wikipedia:Collaboration to convert graphs to SVG. Please don't be offended by my tagging of your graph. If you really feel it is a problem, go ahead and remove the tag; after all, there are more desperate graphs to be converted. Thanks. MithrandirMageT 11:37, 30 March 2007 (UTC)

There is certainly nothing wrong with your image; however, you are right about SVG being the "preferred" format for line drawings. As for quality: though there is no official policy, the Collaboration to convert graphs to SVG recommends hand-drawing SVG graphs in a text-editor, since most vector image-making programs seem to produce somewhat inefficient or inaccurate graphs. It is certainly possible to make nice-looking graphs by hand -- just check out Image:Netscape-navigator-usage-data.svg for an example. Click the image name to view the full image in your browser, then view the page's source; this shows you the underlying SVG code used to make the image. As you can see, they've come up with some clever ways to include the actual data in the SVG file and transform it geometrically into data points on the graph. Thus, the data are not lost as it can be in other image-making processes. I hope this helps! MithrandirMageT 01:58, 2 April 2007 (UTC)

Retrieved from "http://en.wikipedia.org../../../p/d/b/User_talk%7EPdbailey_e99a.html"