<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<meta name="generator" content="http://www.movabletype.org/" />

<title>Social Science Statistics Blog: September 2010 Archives</title>
<link rel="stylesheet" type="text/css" media="screen" href="http://iq.harvard.edu/files/iqss/old/Styles.css"/>
<link rel="stylesheet" href="http://www.iq.harvard.edu/blog/sss/styles-site.css" type="text/css" />
<link rel="start" href="http://www.iq.harvard.edu/blog/sss/" title="Home" />

<link rel="prev" href="http://www.iq.harvard.edu/blog/sss/archives/2010/08/" title="August 2010" />


<link rel="next" href="http://www.iq.harvard.edu/blog/sss/archives/2010/10/" title="October 2010" />



<link rel="shortcut icon" href="/favicon.ico" type="image/icon">
<script language="Javascript">
<!--
Sect = 0
Sub = 0
//-->
</script>
<script language="Javascript" src="/Javascripts/Main.js"></script>
</head>
<body id="Blog">
<a name="Top"></a>
<div id="wrapper">
	<div id="header">
		<div id="BlogBanner">
		<h1><a href="http://www.iq.harvard.edu/blog/sss/" id="BlogHome"><nobr>Social Science Statistics Blog</nobr></a></h1>
		</div>
		<div id="IQhome">
		<a href="http://www.iq.harvard.edu/" title="The Institute for Quantitative Social Science, Harvard University" id="IQhome"><img src="/files/iqss/old/images/IQSS_logo_small.jpg" alt="IQSS Home"></a>
		</div>
	</div>
	<div id="menu"></div>
        <div id="ShadowBottom"></div>
        <div id="LeftBox">
<div id="BB_Subs">
<div id="calendar">
                    <table summary="Monthly calendar with links to each day's posts">
                    <caption>September 2010</caption>
<tr>
<th abbr="Sunday">Sun</th>
<th abbr="Monday">Mon</th>
<th abbr="Tuesday">Tue</th>
<th abbr="Wednesday">Wed</th>
<th abbr="Thursday">Thu</th>
<th abbr="Friday">Fri</th>
<th abbr="Saturday">Sat</th>
</tr>

<tr>

<td>


&nbsp;</td>




<td>


&nbsp;</td>




<td>


&nbsp;</td>




<td>


1

</td>




<td>


2

</td>




<td>


3

</td>




<td>


4

</td>
</tr>

<tr>

<td>


5

</td>




<td>


6

</td>




<td>


7

</td>




<td>


8

</td>




<td>


9

</td>




<td>


10

</td>




<td>


11

</td>
</tr>

<tr>

<td>


12

</td>




<td>


13

</td>




<td>

   
        <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/14/">
            14
        </a>
    


</td>




<td>


15

</td>




<td>


16

</td>




<td>


17

</td>




<td>


18

</td>
</tr>

<tr>

<td>


19

</td>




<td>


20

</td>




<td>

   
        <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/21/">
            21
        </a>
    


</td>




<td>


22

</td>




<td>


23

</td>




<td>


24

</td>




<td>


25

</td>
</tr>

<tr>

<td>


26

</td>




<td>


27

</td>




<td>


28

</td>




<td>


29

</td>




<td>


30

</td>




<td>


&nbsp;</td>




<td>


&nbsp;</td>
</tr>

</table>
</div>

<!--<h3><strong><a href="http://gking.harvard.edu/blogdata/cttee.shtml">Authors' Committee</a></strong></h3>-->
<h3>Authors' Committee</h3>
<h4>Chair:</h4>
<a href="http://www.people.fas.harvard.edu/~blackwel/">Matt Blackwell (Gov)</a>
<p>

<h4>Members:</h4>

Martin Andersen (HealthPol)<br>
<a href="http://kevinbartz.com/">Kevin Bartz (Stats)</a> <br>
<a href="http://www.wjh.harvard.edu/soc/gs/Bloome_Deirdre/">Deirdre Bloome (Social Policy)</a> <br>
<a href="http://www.people.fas.harvard.edu/~aeggers">Andy Eggers (Gov)</a> <br>
<a href="http://www.people.fas.harvard.edu/~jagraves/">John Graves (HealthPol)</a> <br>
<a href="http://people.fas.harvard.edu/~rnielsen/">Rich Nielsen (Gov)</a><br>
<a href="http://www.gov.harvard.edu/people/maya-sen">Maya Sen (Gov)</a><br>
<a href="http://gking.harvard.edu">Gary King (Gov)</a><br>


<h3>Weekly Research Workshop Sponsors</h3>

<a href="http://ksghome.harvard.edu/~aabadie/">Alberto Abadie</a>, 
<a href="http://www.people.hbs.edu/lfleming/">Lee Fleming</a>, 
<a href="http://www.gov.harvard.edu/faculty/aglynn/">Adam Glynn</a>,
<a href="http://post.economics.harvard.edu/people/faculty.html"> Guido Imbens</a>,
<a href="http://gking.harvard.edu">Gary King</a>, 
<a href="http://http://www.people.fas.harvard.edu/~spirling/">Arthur Spirling</a>,
<a href="http://www.hsph.harvard.edu/facres/rbns.html">Jamie Robins</a>,
<a href="http://www.iq.harvard.edu/People/people.php?info=316&sub=4">Don Rubin</a>,
<a href="http://www.wjh.harvard.edu/soc/faculty/winship/">Chris Winship</a><p>


<div class="link-note">
<form method="get" action="https://blogs.hmdc.harvard.edu/mt/mt-search.cgi">
<input type="hidden" name="IncludeBlogs" value="2" />
<label for="search" accesskey="4">Search this site:</label><br />
<input id="search" name="search" size="15"/><br />
<input type="submit" value="Search" />
</form>
</div>


<h3>Recent Comments</h3>

<ul>


<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/10/stories_and_sta.shtml#18152">Stories and statistics</a><br/>
By: online gambling<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/10/workflow_agonis.shtml#18147">Workflow Agonistes</a><br/>
By: Bob Carpenter<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/10/workflow_agonis.shtml#18146">Workflow Agonistes</a><br/>
By: Tim Schwartz<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/you_are_not_so.shtml#18131">You are not so smart</a><br/>
By: Editor Studios<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/what_is_the_lik.shtml#18129">What is the likelihood function?</a><br/>
By: Increase<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/you_are_not_so.shtml#18112">You are not so smart</a><br/>
By: Salim S<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/08/rigor_and_model.shtml#18093">Rigor and modeling in economics</a><br/>
By: Ethan Fosse<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/what_is_the_lik.shtml#18076">What is the likelihood function?</a><br/>
By: Bob Carpenter<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/08/the_seven_deadl.shtml#18055">The Seven Deadly Sins of Contemporary Quantitative Analysis</a><br/>
By: sbutter<br/></li>



<li>Re: <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/08/the_seven_deadl.shtml#18049">The Seven Deadly Sins of Contemporary Quantitative Analysis</a><br/>
By: legitimate survey<br/></li>


</ul>

<h3>Recent Entries</h3>

<ul>
 
<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/what_is_the_lik.shtml">What is the likelihood function?</a></li>
 
<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/you_are_not_so.shtml">You are not so smart</a></li>

</ul>
                                                                               
<div id="categories">
<h3>Categories</h3>
<ul>
<!-- ADD NEW CATEGORY LINKS HERE -->
<li> <a href="http://www.iq.harvard.edu/blog/sss/archives//regular_post">Regular Post</a></li>
<!-- END NEW CATEGORY LINKS --> 
</ul>
</div>


<h3><strong>Blogroll</strong></h3>
<a href = "http://smrjournal.wordpress.com/">SMR Blog</a><br>
<a href = "http://www.j-bradford-delong.net/movable_type/">Brad DeLong</a><br>
<a href = "http://scienceblogs.com/cognitivedaily/">Cognitive Daily</a><br>
<a href = "http://www.iq.harvard.edu/blog/netgov/">Complexity & Social Networks</a><br>
<a href = "http://develintel.blogspot.com/">Developing Intelligence</a><br>
<a href = "http://econlog.econlib.org/">EconLog</a><br>
<a href = "http://www.educationwonk.blogspot.com/">The Education Wonks</a><br>
<a href = "http://www.elsblog.org/">Empirical Legal Studies</a><br>
<a href = "http://www.economist.com/debate/freeexchange/">Free Exchange</a><br>
<a href = "http://www.freakonomics.com/blog/">Freakonomics</a><br>
<a href = "http://healthcare-economist.com/">Health Care Economist</a><br>
<a href = "http://junkcharts.typepad.com/junk_charts/">Junk Charts</a><br>
<a href = "http://itre.cis.upenn.edu/~myl/languagelog/">Language Log</a><br>
<a href = "http://lawprofessors.typepad.com/law_econ/">Law & Econ Prof Blog</a><br>
<a href = "http://hunch.net/">Machine Learning (Theory)</a><br>
<a href = "http://www.marginalrevolution.com/">Marginal Revolution</a><br>
<a href = "http://scienceblogs.com/mixingmemory/">Mixing Memory</a><br>
<a href = "http://www.pollster.com/mystery_pollster/">Mystery Pollster</a><br>
<a href = "http://neweconomist.blogs.com/new_economist/">New Economist</a><br>
<a href = "http://politicalarithmetik.blogspot.com/">Political Arithmetik</a><br>
<a href = "http://poliscimethods.blogspot.com/">Political Science Methods</a><br>
<a href = "http://scienceblogs.com/purepedantry/">Pure Pedantry</a><br>
<a href = "http://lawprofessors.typepad.com/science_law/">Science & Law Blog</a><br>
<a href = "http://jackman.stanford.edu/blog/">Simon Jackman</a><br>
<a href = "http://socialscienceplusplus.blogspot.com/">Social Science++</a><br>
<a href = "http://www.stat.columbia.edu/%7Egelman/blog/">Statistical modeling, causal inference, and social science</a><br>


<h3>Archives</h3>

<ul>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/10/">October 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/">September 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/08/">August 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/05/">May 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/04/">April 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/03/">March 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/02/">February 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2010/01/">January 2010</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/12/">December 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/11/">November 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/10/">October 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/09/">September 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/08/">August 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/06/">June 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/05/">May 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/04/">April 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/03/">March 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/02/">February 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2009/01/">January 2009</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/12/">December 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/11/">November 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/10/">October 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/09/">September 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/08/">August 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/06/">June 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/05/">May 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/04/">April 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/03/">March 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/02/">February 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2008/01/">January 2008</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/12/">December 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/11/">November 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/10/">October 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/09/">September 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/06/">June 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/05/">May 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/04/">April 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/03/">March 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/02/">February 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2007/01/">January 2007</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/12/">December 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/11/">November 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/10/">October 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/09/">September 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/08/">August 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/07/">July 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/06/">June 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/05/">May 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/04/">April 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/03/">March 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/02/">February 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2006/01/">January 2006</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2005/12/">December 2005</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2005/11/">November 2005</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2005/10/">October 2005</a></li>

<li><a href="http://www.iq.harvard.edu/blog/sss/archives/2005/09/">September 2005</a></li>

</ul>

<h3>Notification</h3>
<div class="link-note">
Enter e-mail to [un]subscribe:
<form method="get" action="https://blogs.hmdc.harvard.edu/mt/mt.cgi">
                    <input type="hidden" name="__mode" value="verify_subs" />
                    <input type="hidden" name="blog_id" value="2" />
                    <input id="email" name="email" size="16" />
                    <input type="submit" class="button" value="Go" />
                </form>

</div>


<div class="powered">
Powered by<br /><a href="http://www.movabletype.org">Movable Type 4.24-en</a><br />
<br/>
<div class="syndicate">
<a href="http://www.iq.harvard.edu/blog/sss/index.xml">Syndicate this site (XML)</a>
</div>
<script type="text/javascript">
var gaJsHost = (("https:" == document.location.protocol) ? "https://ssl." : "http://www.");
document.write(unescape("%3Cscript src='" + gaJsHost + "google-analytics.com/ga.js' type='text/javascript'%3E%3C/script%3E"));
</script>
<script type="text/javascript">
var pageTracker = _gat._getTracker("UA-3922751-1");
pageTracker._initData();
pageTracker._trackPageview();
</script> 
</div>
</div>
        </div>
        <div id="RightBox">
                <div id="Content">

<div id="banner">
<small> This blog makes public the hallway conversations about
<em>social science statistical methods and analysis</em> from the <a
href="http://iq.harvard.edu">Institute for Quantitative Social
Science</a> and related research groups.  Expect to see posts on
trends in methodological thought, questions and comments, paper and
conference announcements, applied problems needing methodological
solutions, and methodological techniques seeking applied problems.
Also included are summaries of papers and comments from a popular <a
href="http://www.courses.fas.harvard.edu/colgsas/8142">weekly research
workshop</a> held here and billed as a tour of <a
href="http://www.harvard.edu">Harvard</a>'s statistical innovations
and applications with weekly stops in different disciplines. <a
href="http://www.iq.harvard.edu/blog/Social_Science_Statistics/archives/2005/09/welcome.shtml">[read
more...]</a> </small>

</div>
<hr>
<div class="BlogContent">

<p align="right" class="Grey">

<a href="http://www.iq.harvard.edu/blog/sss/archives/2010/08/">&laquo; August 2010</a> |

<a href="http://www.iq.harvard.edu/blog/sss/">Main</a>

| <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/10/">October 2010 &raquo;</a>

</p>




<p><b>21 September 2010</b></p>


<h3 id="a001353" class="Bigger">What is the likelihood function?</h3>

<p>An interesting 1992 paper by Bayarri and DeGroot entitled &#8220;Difficulties and Ambiguities in the Definition of a Likelihood Function&#8221; (<a href="http://www.springerlink.com/content/4322774u233m3gq0/" title="Gated Version">gated version</a>) grapples with the problem of defining the likelihood when auxiliary variables are at hand. Here is the abstract:</p>

<blockquote>
  <p>The likelihood function plays a very important role in the development of both the theory and practice of statistics. It is somewhat surprising to realize that no general rigorous definition of a likelihood function seem to ever have been given. Through a series of examples it is argued that no such definition is possible, illustrating the difficulties and ambiguities encountered specially in situations involving &#8220;random variables&#8221; and &#8220;parameters&#8221; which are not of primary interest. The fundamental role of such auxiliary quantities (unfairly called &#8220;nuisance&#8221;) is highlighted and a very simple function is argued to convey all the information provided by the observations.</p>
</blockquote>

<p>The example that resonates with me in on pages 4-6, where they describe the ambiguity of using defining the likelihood function when there is an observation <em>y</em> which is a measurement of <em>x</em> subject to (classical) error. There are several different ways of writing a likelihood in that case, depending on how you handle the latent, unobserved data <em>x</em>.  One can condition on it, marginalize across it, or include it in the joint distribution of the data. Each of these can lead to a different MLE. </p>

<p>Their point is that situations like this involve subjective choices (though, all modeling requires subjective choice) and the hermetic seal between the &#8220;model&#8221; and the &#8220;prior&#8221; is less airtight than we think. </p>




<p class="posted">
Posted by <a href="http://www.iq.harvard.edu/blog/sss/archives/author/matt-blackwell/">Matt Blackwell</a> at <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/what_is_the_lik.shtml"> 4:41 PM</a>

| <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/what_is_the_lik.shtml#comments">Comments (2)</a>


</p>





<p><b>14 September 2010</b></p>


<h3 id="a001352" class="Bigger">You are not so smart</h3>

<p><a href="http://youarenotsosmart.com/">You are not so smart</a> is a blog dedicated to explaining self-delusions. The most recent <a href="http://youarenotsosmart.com/2010/09/11/the-texas-sharpshooter-fallacy/#more-689">post</a> is on the <a href="http://en.wikipedia.org/wiki/Texas_sharpshooter_fallacy">Texas sharpshooter fallacy</a>:<br />
<blockquote><br />
<b>The Misconception:</b> You take randomness into account when determining cause and effect.<br />
<b>The Truth:</b> You tend to ignore random chance when the results seem meaningful or when you want a random event to have a meaningful cause.<br />
</blockquote></p>



<p class="posted">
Posted by <a href="http://www.iq.harvard.edu/blog/sss/archives/author/matt-blackwell/">Matt Blackwell</a> at <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/you_are_not_so.shtml"> 6:33 PM</a>

| <a href="http://www.iq.harvard.edu/blog/sss/archives/2010/09/you_are_not_so.shtml#comments">Comments (2)</a>


</p>



</div>
</div>
</div>
<div id="footer">
<script language="Javascript" src="/Javascripts/PageBottom.js"></script>
</div>
</div>
</body></html>

