<?xml version="1.0" encoding="UTF-8"?><!-- generator="wordpress/2.3.3" -->
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	>
<channel>
	<title>Comments on: Statistics of Breath Testing</title>
	<link>http://blogs.gnome.org/jamesh/2006/05/16/statistics-of-breath-testing/</link>
	<description>Random stuff</description>
	<pubDate>Mon, 07 Jul 2008 01:02:15 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.3.3</generator>
		<item>
		<title>By: Johannes</title>
		<link>http://blogs.gnome.org/jamesh/2006/05/16/statistics-of-breath-testing/#comment-164</link>
		<dc:creator>Johannes</dc:creator>
		<pubDate>Tue, 30 Nov 1999 00:00:00 +0000</pubDate>
		<guid>http://blogs.gnome.org/jamesh/2006/05/16/statistics-of-breath-testing/#comment-164</guid>
		<description>Hi!&lt;p/&gt;Just for your notice. In Germany breath tests are only used for a fast investigation. You can only be judged if a blood test also gives a positive result. And of course, if people are slightly below the limit in the breath test, they do a blood test, too!&lt;p/&gt;Regards,&lt;br/&gt;Johannes</description>
		<content:encoded><![CDATA[<p>Hi!
<p />Just for your notice. In Germany breath tests are only used for a fast investigation. You can only be judged if a blood test also gives a positive result. And of course, if people are slightly below the limit in the breath test, they do a blood test, too!
<p />Regards,<br />Johannes</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Olaf</title>
		<link>http://blogs.gnome.org/jamesh/2006/05/16/statistics-of-breath-testing/#comment-165</link>
		<dc:creator>Olaf</dc:creator>
		<pubDate>Tue, 30 Nov 1999 00:00:00 +0000</pubDate>
		<guid>http://blogs.gnome.org/jamesh/2006/05/16/statistics-of-breath-testing/#comment-165</guid>
		<description>Hi James,&lt;p/&gt;assuming the device is unbiased (something that can easily be fixed if it where so by calibartion) you don't really need a more accurate method. All you need to to is average over several measurements. The mean of n measurements will again be normaly distributed, but with a variance of 1/n*sigma^2 if the original measurements had a variance of sigma^2. So by simply repeating the procedure you get a much lower margin of error. &lt;p/&gt;-- Olaf</description>
		<content:encoded><![CDATA[<p>Hi James,
<p />assuming the device is unbiased (something that can easily be fixed if it where so by calibartion) you don&#8217;t really need a more accurate method. All you need to to is average over several measurements. The mean of n measurements will again be normaly distributed, but with a variance of 1/n*sigma^2 if the original measurements had a variance of sigma^2. So by simply repeating the procedure you get a much lower margin of error.
<p />&#8211; Olaf</p>
]]></content:encoded>
	</item>
</channel>
</rss>
