<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
<p class="MsoPlainText" style="line-height: 200%;"><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">Can
anyone tell me whether the log likelihood calculation that follows
(from Rayson and Rayson and Garside) produces the exact same log
likelihood figures for a 2X2 table that the formula in Dunning (1993)
does?<br>
</span></p>
<p class="MsoPlainText" style="line-height: 200%;"><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">Rayson
and Garside: <br>
</span></p>
<p class="MsoPlainText" style="line-height: 200%;"><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">-2ln</span><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">λ</span><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">
= 2*(((a*ln(a/E1))
+ (b*ln(b/E2)))<o:p></o:p></span></p>
<span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">Dunning:<br>
<br>
-2ln</span><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">λ</span><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">
=2[logL(p1,k1,n1) + logL(p2,k2,n2) - logL(p,k1,n1) - logL(p,k2,n2)]<br>
<br>
I've tried both with Dunning's data in his 1993 paper and get very
slightly different results with each. However, as is obvious from this
email, I am neither a mathematician nor a statistician. <br>
<br>
Many, many thanks,<br>
<br>
Don<br>
<br>
<br>
Dunning, Ted. "Accurate Methods for the Statistics of Surprise and
Coincidence." <u>Computational Linguistics</u>, (1993): 61-74.<br>
</span>
<p class="MsoPlainText"
style="margin-left: 0.5in; text-indent: -0.5in; line-height: 200%;"><strong><span
style="font-size: 12pt; line-height: 200%; font-weight: normal;">Rayson,
Paul.</span></strong><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">
“Matrix: A Statistical
Method and Software Tool for Linguistic Analysis Through Corpus
Comparison.” <span style="">Ph.D. diss.,</span> </span><st1:place><st1:PlaceName><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">Lancaster</span></st1:PlaceName><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">
</span><st1:PlaceType><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">University</span></st1:PlaceType></st1:place><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";">,<b
style=""> </b>2003.</span><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";"
lang="EN-GB"><o:p></o:p></span></p>
<p class="MsoPlainText"
style="margin-left: 0.5in; text-indent: -0.5in; line-height: 200%;"><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";"
lang="EN-GB">Rayson,
Paul, and Roger Garside.<span style=""> </span>“Comparing
Corpora Using Frequency Profiling.”<span style=""> </span><u>Proceedings
of the Workshop on Comparing Corpora</u>, (2000): 1-6.</span><span
lang="EN-GB"> </span><span
style="font-size: 12pt; line-height: 200%; font-family: "Times New Roman";"
lang="EN-GB"><o:p></o:p></span></p>
<pre class="moz-signature" cols="72">
Donald E. Hardy
Associate Professor
Department of English
Colorado State University
Fort Collins, CO 80523
970-491-5349
<a class="moz-txt-link-abbreviated"
href="mailto:Don.Hardy@Colostate.edu">Don.Hardy@Colostate.edu</a>
<a class="moz-txt-link-freetext" href="http://textant.colostate.edu">http://textant.colostate.edu</a></pre>
</body>
</html>