<?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/"
	xmlns:creativeCommons="http://backend.userland.com/creativeCommonsRssModule"
xmlns:media="http://search.yahoo.com/mrss/" 

	>
<channel>
	<title>Comments on: What Does it Mean to &#8220;Control For&#8221; Something?</title>
	<atom:link href="http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/feed/" rel="self" type="application/rss+xml" />
	<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/</link>
	<description>Sociological Images encourages people to exercise and develop their sociological imaginations with discussions of compelling visuals that span the breadth of sociological inquiry.</description>
	<lastBuildDate>Fri, 10 Apr 2015 03:38:00 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	
	<item>
		<title>By: EB</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-568620</link>
		<dc:creator><![CDATA[EB]]></dc:creator>
		<pubDate>Fri, 01 Feb 2013 17:17:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-568620</guid>
		<description><![CDATA[ the whole point of the types of maps featured in the comic are to show population density.  So this analysis misses the point.]]></description>
		<content:encoded><![CDATA[<p> the whole point of the types of maps featured in the comic are to show population density.  So this analysis misses the point.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Lunad</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567533</link>
		<dc:creator><![CDATA[Lunad]]></dc:creator>
		<pubDate>Fri, 18 Jan 2013 20:09:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567533</guid>
		<description><![CDATA[whatever the method used to present data (be it GIS mapping or bar charts) failing to control for an important variable such as population density will skew your results.  Comparing countries, for instance, it is often useful to use per capita measures, to control for the fact that Norway has a lot fewer people than China.]]></description>
		<content:encoded><![CDATA[<p>whatever the method used to present data (be it GIS mapping or bar charts) failing to control for an important variable such as population density will skew your results.  Comparing countries, for instance, it is often useful to use per capita measures, to control for the fact that Norway has a lot fewer people than China.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: What Does it Mean to “Control For” Something? » Sociological Images &#124; digitalnews2000</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567487</link>
		<dc:creator><![CDATA[What Does it Mean to “Control For” Something? » Sociological Images &#124; digitalnews2000]]></dc:creator>
		<pubDate>Tue, 15 Jan 2013 22:48:49 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567487</guid>
		<description><![CDATA[[...] on thesocietypages.org Share this:TwitterFacebookLike this:LikeBe the first to like this.   This entry was posted in [...]]]></description>
		<content:encoded><![CDATA[<p>[&#8230;] on thesocietypages.org Share this:TwitterFacebookLike this:LikeBe the first to like this.   This entry was posted in [&#8230;]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: EB</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567484</link>
		<dc:creator><![CDATA[EB]]></dc:creator>
		<pubDate>Tue, 15 Jan 2013 21:20:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567484</guid>
		<description><![CDATA[ &quot;slight variations of census maps of the population density of the US&quot;

I a word]]></description>
		<content:encoded><![CDATA[<p> &#8220;slight variations of census maps of the population density of the US&#8221;</p>
<p>I a word</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: EB</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567483</link>
		<dc:creator><![CDATA[EB]]></dc:creator>
		<pubDate>Tue, 15 Jan 2013 21:18:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567483</guid>
		<description><![CDATA[This analysis misses the  actual point of the xkcd cartoon that is a joke for GIS analysts and census data users - the three maps are basically slight variations of census maps of the population density the US.

Because of this, yes, more furries, the website users, and martha steward living subscribers are probably located in areas of high population density.  

This is not about control variables at all, but rather, poking fun of the trend where reviewers/customers want you to GIS map where the users/consumers/respondents who do X are located.  A fair number of times you may find that more &lt;&gt; are located in heavily populated areas.  

As a result, the point is, when GIS mapping, be aware of what you are doing.]]></description>
		<content:encoded><![CDATA[<p>This analysis misses the  actual point of the xkcd cartoon that is a joke for GIS analysts and census data users &#8211; the three maps are basically slight variations of census maps of the population density the US.</p>
<p>Because of this, yes, more furries, the website users, and martha steward living subscribers are probably located in areas of high population density.  </p>
<p>This is not about control variables at all, but rather, poking fun of the trend where reviewers/customers want you to GIS map where the users/consumers/respondents who do X are located.  A fair number of times you may find that more &lt;&gt; are located in heavily populated areas.  </p>
<p>As a result, the point is, when GIS mapping, be aware of what you are doing.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Yrro Simyarin</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567450</link>
		<dc:creator><![CDATA[Yrro Simyarin]]></dc:creator>
		<pubDate>Mon, 14 Jan 2013 14:59:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567450</guid>
		<description><![CDATA[Fair enough. I wasn&#039;t necessarily meaning that you should send it off as a final step as in &quot;here&#039;s my data, please plug it into matlab for me.&quot; I was more speaking as a consultant or level of cooperation throughout the project. I expect a biologist to be an expert on biology, not on bayesian inference. I guess in the academic world that level of cooperation means cosigning the paper - in business it just means you pay someone who designs databases to design your database, then go make your application on top of it.]]></description>
		<content:encoded><![CDATA[<p>Fair enough. I wasn&#8217;t necessarily meaning that you should send it off as a final step as in &#8220;here&#8217;s my data, please plug it into matlab for me.&#8221; I was more speaking as a consultant or level of cooperation throughout the project. I expect a biologist to be an expert on biology, not on bayesian inference. I guess in the academic world that level of cooperation means cosigning the paper &#8211; in business it just means you pay someone who designs databases to design your database, then go make your application on top of it.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: lambdaphage</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567442</link>
		<dc:creator><![CDATA[lambdaphage]]></dc:creator>
		<pubDate>Sun, 13 Jan 2013 20:41:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567442</guid>
		<description><![CDATA[I dunno.  I agree that abuse of statistics is rife, but I don&#039;t know whether contracting out stats grad students is the best way to address it.  Many research universities have &#039;imaging cores&#039; or &#039;bioinformatics cores&#039; which usually consist of salaried scientists attached to the university&#039;s College of Science or whatever,  The idea is that a single lab could never justify its own SEM or Illumina sequencer, so the university buys one and funds some techs to run it.  Typically, though, getting your imaging done or your genome assembled is both (1) too technical to do yourself and (2) not central enough to the fundamental contribution of the paper to justify authorship.

With statistics, though, I would suspect that the questions are either utterly trivial (&quot;how do I test non-parametrically for equality of the distributions?&quot;) or so central to the project (&quot;how should I design my clinical trial?&quot;) that the statistician&#039;s name should really be on the paper anyway.  Ultimately, I think the attitude that statistics is something you can deal with as an afterthought, or even contract out, is a very bad sign.  

In the natural sciences some scientists may be statistically naive, but they&#039;re at least a little sheepish about that fact.  I&#039;ve never encountered anyone who, like, actively &lt;i&gt;denies&lt;/i&gt; the very concept of statistical inference until I started reading the comments on this site.
]]></description>
		<content:encoded><![CDATA[<p>I dunno.  I agree that abuse of statistics is rife, but I don&#8217;t know whether contracting out stats grad students is the best way to address it.  Many research universities have &#8216;imaging cores&#8217; or &#8216;bioinformatics cores&#8217; which usually consist of salaried scientists attached to the university&#8217;s College of Science or whatever,  The idea is that a single lab could never justify its own SEM or Illumina sequencer, so the university buys one and funds some techs to run it.  Typically, though, getting your imaging done or your genome assembled is both (1) too technical to do yourself and (2) not central enough to the fundamental contribution of the paper to justify authorship.</p>
<p>With statistics, though, I would suspect that the questions are either utterly trivial (&#8220;how do I test non-parametrically for equality of the distributions?&#8221;) or so central to the project (&#8220;how should I design my clinical trial?&#8221;) that the statistician&#8217;s name should really be on the paper anyway.  Ultimately, I think the attitude that statistics is something you can deal with as an afterthought, or even contract out, is a very bad sign.  </p>
<p>In the natural sciences some scientists may be statistically naive, but they&#8217;re at least a little sheepish about that fact.  I&#8217;ve never encountered anyone who, like, actively <i>denies</i> the very concept of statistical inference until I started reading the comments on this site.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Yrro Simyarin</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567439</link>
		<dc:creator><![CDATA[Yrro Simyarin]]></dc:creator>
		<pubDate>Sun, 13 Jan 2013 15:45:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567439</guid>
		<description><![CDATA[Pretty much. &quot;Researchers who cannot read a textbook&quot; seems to apply to an unsettling large number of studies I&#039;ve read.

I didn&#039;t say they should do it for free.]]></description>
		<content:encoded><![CDATA[<p>Pretty much. &#8220;Researchers who cannot read a textbook&#8221; seems to apply to an unsettling large number of studies I&#8217;ve read.</p>
<p>I didn&#8217;t say they should do it for free.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: lambdaphage</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567430</link>
		<dc:creator><![CDATA[lambdaphage]]></dc:creator>
		<pubDate>Sun, 13 Jan 2013 07:57:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567430</guid>
		<description><![CDATA[As in, to perform basic statistical analysis for researchers in other fields who cannot read a textbook, instead of pursuing their own research?  ]]></description>
		<content:encoded><![CDATA[<p>As in, to perform basic statistical analysis for researchers in other fields who cannot read a textbook, instead of pursuing their own research?  </p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Yrro Simyarin</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567428</link>
		<dc:creator><![CDATA[Yrro Simyarin]]></dc:creator>
		<pubDate>Sun, 13 Jan 2013 05:20:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567428</guid>
		<description><![CDATA[Do any colleges have their stats department farm out grad assistants to help with this stuff? I&#039;ve seen some pretty horrible abuse of statistics in pretty much all fields of research.]]></description>
		<content:encoded><![CDATA[<p>Do any colleges have their stats department farm out grad assistants to help with this stuff? I&#8217;ve seen some pretty horrible abuse of statistics in pretty much all fields of research.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: L</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567426</link>
		<dc:creator><![CDATA[L]]></dc:creator>
		<pubDate>Sun, 13 Jan 2013 00:05:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567426</guid>
		<description><![CDATA[That&#039;s why using choropleths to represent magnitude is wrong.  ]]></description>
		<content:encoded><![CDATA[<p>That&#8217;s why using choropleths to represent magnitude is wrong.  </p>
]]></content:encoded>
	</item>
	<item>
		<title>By: msilfan</title>
		<link>http://thesocietypages.org/socimages/2013/01/12/what-does-it-mean-to-control-for-something/comment-page-1/#comment-567423</link>
		<dc:creator><![CDATA[msilfan]]></dc:creator>
		<pubDate>Sat, 12 Jan 2013 20:31:00 +0000</pubDate>
		<guid isPermaLink="false">http://thesocietypages.org/socimages/?p=52764#comment-567423</guid>
		<description><![CDATA[Reminds me of the hysterically funny Twitter hashtag #overlyhonestmethods - I&#039;d love to see one of you do a post on that! ]]></description>
		<content:encoded><![CDATA[<p>Reminds me of the hysterically funny Twitter hashtag #overlyhonestmethods &#8211; I&#8217;d love to see one of you do a post on that! </p>
]]></content:encoded>
	</item>
</channel>
</rss>
