<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Working with numbers</title>
	<atom:link href="http://lulupinney.co.uk/2011/10/working-with-numbers/feed/" rel="self" type="application/rss+xml" />
	<link>http://lulupinney.co.uk/2011/10/working-with-numbers/</link>
	<description>Done well, infographic design clarifies, explains, reveals or adds meaning</description>
	<lastBuildDate>Wed, 02 Jan 2013 11:12:03 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.5.1</generator>
	<item>
		<title>By: lupi</title>
		<link>http://lulupinney.co.uk/2011/10/working-with-numbers/#comment-1246</link>
		<dc:creator>lupi</dc:creator>
		<pubDate>Thu, 17 Nov 2011 15:09:29 +0000</pubDate>
		<guid isPermaLink="false">http://lulupinney.co.uk/?p=1920#comment-1246</guid>
		<description><![CDATA[I said I&#039;d welcome feedback, and I do - obviously I&#039;m not the only person who&#039;s sensitive to these issues! Thank you all for your comments.]]></description>
		<content:encoded><![CDATA[<p>I said I&#8217;d welcome feedback, and I do &#8211; obviously I&#8217;m not the only person who&#8217;s sensitive to these issues! Thank you all for your comments.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Mark</title>
		<link>http://lulupinney.co.uk/2011/10/working-with-numbers/#comment-1226</link>
		<dc:creator>Mark</dc:creator>
		<pubDate>Tue, 08 Nov 2011 17:59:56 +0000</pubDate>
		<guid isPermaLink="false">http://lulupinney.co.uk/?p=1920#comment-1226</guid>
		<description><![CDATA[I love the line, &quot;leave out as much detail as you can&quot;.

And I generally love this article.  My one dissent is not so much against what is said here, but a common misuse of the principle.    If you are providing estimates NOT in the text of an analysis, but in something like detailed tables,  rounding is not the ideal way to tell data users about the accuracy of your estimate.     Particularly if the data users are going to use your data for other analyses.    

An example:  If your reasonably unbiased estimates of average earnings for two groups are $12,395  and  $9,637, it is much more meaningful to take an income ratio of those two numbers rather than of $12,000 and $10,000, even if 2 standard errors &gt; $500.    This is even more so if the analysis being done by a data user is going to look at a number of such ratios over time or over different groups. 

It is always difficult to prevent misuse of data, but suppressing it is not a good answer.]]></description>
		<content:encoded><![CDATA[<p>I love the line, &#8220;leave out as much detail as you can&#8221;.</p>
<p>And I generally love this article.  My one dissent is not so much against what is said here, but a common misuse of the principle.    If you are providing estimates NOT in the text of an analysis, but in something like detailed tables,  rounding is not the ideal way to tell data users about the accuracy of your estimate.     Particularly if the data users are going to use your data for other analyses.    </p>
<p>An example:  If your reasonably unbiased estimates of average earnings for two groups are $12,395  and  $9,637, it is much more meaningful to take an income ratio of those two numbers rather than of $12,000 and $10,000, even if 2 standard errors &gt; $500.    This is even more so if the analysis being done by a data user is going to look at a number of such ratios over time or over different groups. </p>
<p>It is always difficult to prevent misuse of data, but suppressing it is not a good answer.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Brad Smith</title>
		<link>http://lulupinney.co.uk/2011/10/working-with-numbers/#comment-1225</link>
		<dc:creator>Brad Smith</dc:creator>
		<pubDate>Tue, 08 Nov 2011 16:03:21 +0000</pubDate>
		<guid isPermaLink="false">http://lulupinney.co.uk/?p=1920#comment-1225</guid>
		<description><![CDATA[Helpful article.  Thank you.  

In the first table the implied decimal points shift to the left when going from double to single digits.  All of the decimal points in a column, real or implied, should line up.]]></description>
		<content:encoded><![CDATA[<p>Helpful article.  Thank you.  </p>
<p>In the first table the implied decimal points shift to the left when going from double to single digits.  All of the decimal points in a column, real or implied, should line up.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Lavorare con i numeri: cifre significative e rilevanza &#171; Sbagliando s&#8217;impera</title>
		<link>http://lulupinney.co.uk/2011/10/working-with-numbers/#comment-1212</link>
		<dc:creator>Lavorare con i numeri: cifre significative e rilevanza &#171; Sbagliando s&#8217;impera</dc:creator>
		<pubDate>Sat, 05 Nov 2011 15:41:55 +0000</pubDate>
		<guid isPermaLink="false">http://lulupinney.co.uk/?p=1920#comment-1212</guid>
		<description><![CDATA[[...] Lavorare con i numeri [...]]]></description>
		<content:encoded><![CDATA[<p>[...] Lavorare con i numeri [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Scott Hendrickson</title>
		<link>http://lulupinney.co.uk/2011/10/working-with-numbers/#comment-1211</link>
		<dc:creator>Scott Hendrickson</dc:creator>
		<pubDate>Sat, 05 Nov 2011 15:12:43 +0000</pubDate>
		<guid isPermaLink="false">http://lulupinney.co.uk/?p=1920#comment-1211</guid>
		<description><![CDATA[Part of the problem is that common tools don&#039;t support paying attention to and displaying significance. MS Excel does a poor job at this as the easiest choice there is to give all of your numbers the same number of decimal places. The constant-decimal-places idea applies best to money, best I would argue that if you are talking about many millions of dollars in any real situation, you have lost track of counting individual cents and so they don&#039;t really mean anything there either.  And as you note, calculators display a lot of digits, leading users to think they mean something.]]></description>
		<content:encoded><![CDATA[<p>Part of the problem is that common tools don&#8217;t support paying attention to and displaying significance. MS Excel does a poor job at this as the easiest choice there is to give all of your numbers the same number of decimal places. The constant-decimal-places idea applies best to money, best I would argue that if you are talking about many millions of dollars in any real situation, you have lost track of counting individual cents and so they don&#8217;t really mean anything there either.  And as you note, calculators display a lot of digits, leading users to think they mean something.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
