sci.stat.math Google Group

Jordan Shoes for sale Jordan Fusion Wholesale Free shipping Paypal payment (www.vipchinatrade.com)

tradeyoy...@163.com (nicetoto) — Thu, 18 Mar 2010 18:54:36 UT

We offer Cheap Nike Shoes, Air Jordan Shoes are on sale now, our Air
Jordan shoes, Nike Shoes are exclusive and we offer original box of
jordan shoes.
Nike Air Jordan Shoes
Cheap Wholesale Jordan 1 (paypal payment)([link] )
Cheap Wholesale Jordan 1.5 (paypal payment)([link] )

Re: Outliers which are very large yet very infrequent

paige.mil...@kodak.com (Paige Miller) — Thu, 18 Mar 2010 18:50:45 UT

This is too vague for me to understand.
Let's start with what you mean by "This method does "blind"
estimation" — What method? What is blind estimation?

Outliers which are very large yet very infrequent

mtbrenne...@gmail.com (JunoExpress) — Thu, 18 Mar 2010 18:02:15 UT

Hi,
I have a problem where outliers occur very infrequently (less than
0.25% of the time), but when they do occur they are very large. So
large in fact that they have a small, but non-trivial effect on the
variance and to a lesser degree the mean. This method does "blind"
estimation, so I don't know that a value is an outlier when it occurs

2010 Hot sale Shox R3.R4, Shox TZ Nike Air Max 90 Air Max TN,Air Force One For Both men and women (www.vipchinatrade.com)

tradeyo...@163.com (disthrse) — Thu, 18 Mar 2010 17:44:08 UT

Cheap Wholesale Shox NZ <[link]> paypal payment
Cheap Wholesale Shox OZ <[link]> free shipping
Cheap Wholesale Shox R2 <[link]> paypal payment
Cheap Wholesale Shox R3 <[link]> paypal payment
Cheap Wholesale Shox R3+R4 <[link]> paypal payment

Re: Seeking Excel function for bivariate normal cdf

paulvonhip...@yahoo.com (Paul) — Thu, 18 Mar 2010 17:26:23 UT

Perfect, thank you.

Re: ID'ing Bias In Data ?

jazzg...@aol.com (Allen) — Thu, 18 Mar 2010 17:06:12 UT

Thanks Mr. Kendall. After further reflection I don't think the county
vs. State will be an issue, more of a thought I guess. I am
definitely going to try your suggestion. Thanks.

Re: ID'ing Bias In Data ?

a...@drkendall.org (Art Kendall) — Thu, 18 Mar 2010 16:52:14 UT

Why do you think that the form of government would influence the
numbers? Different methods of gathering the data?
Also what about States where counties are mostly for Census and a few
functions and localities are the bodies that serve most of the functions?
You might be able to get some additional variables from the City County

Re: ID'ing Bias In Data ?

jazzg...@aol.com (Allen) — Thu, 18 Mar 2010 16:07:03 UT

Thanks to all ! One thing, I do have percentages and in some cases
these percentages could be different from county to county due to the
State having county administered governments rather than a State
administrated government, to be safer, should I take this in to
account?

Re: ID'ing Bias In Data ?

a...@drkendall.org (Art Kendall) — Thu, 18 Mar 2010 15:43:31 UT

Reading between the lines it seems that you may have percents. Percents
on different bases are not additive. To get aggregate measures
(national, regional) you need to go back to numerators and denominators
as separate variables.
For example think of a state with 2 counties. County A has 1 homicide

Re: ID'ing Bias In Data ?

gordon.sa...@eastlink.ca (Gordon Sande) — Thu, 18 Mar 2010 14:59:04 UT

Paul Velleman has a nice talk in which he gives some county based statistics
and their "analysis". Marvelous story behind why the good ones are good and
localized with an equally marvelous story about the bad ones which
unfortunately
have the same localization. Both stories are bogus (but sound all too common!)

Re: ID'ing Bias In Data ?

paul_ru...@att.net (Paul) — Thu, 18 Mar 2010 14:58:13 UT

Are these scores attempting to measure something relating to an
individual resident (such as crime rate per 1,000 population, or
average income) rather than something relating to the county
administrative level (where a county has one government regardless of
its population)? If it's at the individual level, you could try

ID'ing Bias In Data ?

jazzg...@aol.com (Allen) — Thu, 18 Mar 2010 14:42:50 UT

I am drawing a blank on this one. I have individual county scores
(numeric values, ratio level, scores by each county can range from 0%
to 99% and I have over 2,500 county scores). I am guessing that
States with more counties could bias the results at a national level,
however, I am drawing a blank on how I can determine if the bias

Re: Seeking Excel function for bivariate normal cdf

gaj.vid...@mf.uni-lj.si (Gaj Vidmar) — Thu, 18 Mar 2010 10:07:32 UT

A little googling quickly finds a couple of readily and freely available
solutions:
1. [link]
=> workbook [link]
(enable macros, of course; just tested it, simple, works, VBA not protected
so you can get the function itself)

Re: Data Randomisation /mean of randomised data

koop...@sfu.ca (Ray Koopman) — Thu, 18 Mar 2010 09:23:41 UT

On Mar 18, 1:35 am, Konstanze Krueger <konstanze-krue...@usa.com>
wrote:
If all you want to do is shuffle the offdiagonals then there's no
need to actually do it. For each offdiagonal cell, the distribution
of the shuffled values will be exactly the same as the observed
distribution of offdiagonals.

Re: Data Randomisation /mean of randomised data

konstanze-krue...@usa.com (Konstanze Krueger) — Thu, 18 Mar 2010 08:35:58 UT

Ray Koopman schrieb:
Dear Ray,
answers are pasted below
for both questions: yes!
Yes, we would like to shuffle offdiaginale values.
Generate a new matrix randomly with the
Behaviour counts for pairs differ depending on the direction they are
displayed. For instance A display 3 threats towards B, and B displays 5