Search String: Display: Description: Sort:

Results:

References: [ +from:korourke@lri.ca: 25 ]

Total 25 documents matching your query.

1. Statistician Position, Ottawa, Canada (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 4 Apr 2001 11:28:15 -0400
Two year position available for a statistician to provide data analysis support, consultation and collaboration to the Department of Surgery, University of Ottawa and the Ottawa Regional Cancer Centr
/archives/html/s-news/2001-04/msg00048.html (7,400 bytes)

2. Neutralizing effect of variables (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 15 Mar 2001 12:05:37 -0500
You might wish to start with Mosteller F, Tukey JW. (1977). Data analysis and regression. New York: Addison-Wesley. (there is a chapter on removing effects~ something like regression as exclusion) if
/archives/html/s-news/2001-03/msg00125.html (7,475 bytes)

3. Re: Confidence Interval on Quotient (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 1 Feb 2001 09:05:29 -0500
But IF you can avoid all this by taking logs, some think you are very well advised to. see page 194 Fisher RA The Design of Experiments 8th Edition 1971 (data on differences/ratios of heights of pla
/archives/html/s-news/2001-02/msg00012.html (7,328 bytes)

4. Re: "over-dispersed" gamma model (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 8 Feb 2001 11:58:18 -0500
The literature is extensive to say the least - a good start might be Y Lee and JA Nelder. Hierarchical generalized linear models. Journal of the Royal Statistical Society 58 (4):619-678, 1996. The c
/archives/html/s-news/2001-02/msg00087.html (8,289 bytes)

5. Re: t-test for binary data (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 2 Nov 2000 16:51:58 -0500
In fact apart from the rather finnicky Yates' correction the Finnicky!! Getting the size of the test correct and not a function of the nuisance parameter? (slightly different than being "exact") If t
/archives/html/s-news/2000-11/msg00023.html (7,520 bytes)

6. job re-posting (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Tue, 14 Nov 2000 15:03:34 -0500
Below is a job posting at my institute that I would like to ecourage interested parties to consider. It might be ideal for some one who would like to work in an applied area for a couple of years bef
/archives/html/s-news/2000-11/msg00152.html (8,778 bytes)

7. job re-posting (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 15 Nov 2000 10:12:47 -0500
/archives/html/s-news/2000-11/msg00161.html (9,564 bytes)

8. Re: [S] Sampling correlated vectors (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 7 Sep 2000 11:06:39 -0400
Shades of the "Problem of Moments" (circa 1800,s by Chebyshev) With a univariate sample it is perhaps easy to see that the standard bootstrap is equivalent to estimating a distribution function that
/archives/html/s-news/2000-09/msg00049.html (8,414 bytes)

9. [S] Simulating from a quasi likelihood GLM (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Mon, 11 Sep 2000 08:40:13 -0400
There is a sense in which it can't be done - Simulating from a quasi likelihood GLM - as quasi likelihood is a "dodge" from specifying a full (compound) probability model. But you can generate sample
/archives/html/s-news/2000-09/msg00073.html (8,077 bytes)

10. [S] critical history (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 27 Sep 2000 08:53:36 -0400
Just noticed on page 3 of the new "Introducing S-PLUS 6" brochure the following historically unsupported claim. "Note the apparent reversal of the years at the Morris site, a data error that went und
/archives/html/s-news/2000-09/msg00182.html (8,463 bytes)

11. RE: [S] Preventing overfitting artefacts in multiple linear regre ssion (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Tue, 8 Aug 2000 11:23:02 -0400
ant2assign2C, where=1, immediate=T) Greets Diego -- Diego Kuonen http://wap.kuonen.com http://stat.kuonen.com http://www.Statoo.ch Ya era Ud. statouado? ht
/archives/html/s-news/2000-08/msg00065.html (8,447 bytes)

12. [S] extracting what print methods "print" (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 9 Aug 2000 13:34:22 -0400
Samuel Buttrey and Frank E Harrell Jr for their advise on the use of the Leaps funtion. I haven't had time yet to work on their ideas, but I'll do it as so
/archives/html/s-news/2000-08/msg00081.html (7,419 bytes)

13. Summary RE: [S] extracting what print methods "print" (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 16 Aug 2000 16:13:11 -0400
d by the exact same need(!) I wrote a silly function that picks up the printout and add it to the output value as an attribute called "printout": snatch <-
/archives/html/s-news/2000-08/msg00158.html (12,011 bytes)

14. Re: [S] Interpreting stepwise results in Poisson regression (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 8 Jun 2000 10:56:23 -0400
cut EXPLANATION was starred but perhaps could been more boldly put as "description under the (unrealistic?) assumption that variation is (best thought of as) random" although subject to some interpr
/archives/html/s-news/2000-06/msg00055.html (9,763 bytes)

15. RE: [S] Interpreting stepwise results in Poisson regression (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Thu, 8 Jun 2000 16:36:34 -0400
as well as linearity issue was about selection, not given some selection and some conditions essential equivilance ... In fact, some would argue that very different selections will likely be made us
/archives/html/s-news/2000-06/msg00062.html (11,069 bytes)

16. [S] SUMMARY: (likelihoods &) weights in glm (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 28 Jun 2000 10:17:00 -0400
This relates to a more general concern (raised a year or so ago) about likelihoods being under-emphasized in SPlus. If you want to know/see what the "weights option does" in a glm or gam you need to
/archives/html/s-news/2000-06/msg00213.html (8,873 bytes)

17. [S] LAST ANOUNCEMENT: Cox & Wermuth Ottawa Workshop (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Tue, 2 May 2000 09:39:04 -0400
For those who may still be able to attend ... OTTAWA WORKSHOP ON APPLYING STATISTICS 2000 Monday & Tuesday, May 15&16 2000 9am to 5pm Ottawa, Ontario Canada Sponsored by Department of Epidemiology an
/archives/html/s-news/2000-05/msg00017.html (10,663 bytes)

18. [S] Non SPlus short course announcement (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Sun, 6 Feb 2000 16:46:08 -0500
-- This message was distributed by s-news@wubios.wustl.edu. To unsubscribe send e-mail to s-news-request@wubios.wustl.edu with the BODY of the message: unsubscribe s-news
/archives/html/s-news/2000-02/msg00046.html (8,642 bytes)

19. Posts Re: [S] Multivariate hypothesis tests (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Fri, 28 Jan 2000 14:51:03 -0500
cut cut -- and -- I found these two replies interesting particularily as they both implicitly involve the adoption of a Neyman null hypothesis (testing a main effect, say mean, in the presence of an
/archives/html/s-news/2000-01/msg00228.html (9,478 bytes)

20. [S] post-SUMMARY: data visualization references (score: 1)
Author: "O'Rourke, Keith" <korourke@lri.ca>
Date: Wed, 1 Dec 1999 11:58:12 -0500
Two more comments I. Sometimes it might be worthwhile to evaluate the visualization against random noise or permuted data for a possibly interesting example see [1999] All maps of parameter estimates
/archives/html/s-news/1999-12/msg00013.html (8,699 bytes)


This search system is powered by Namazu