<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
On 08/07/2012 06:28 AM, Yuri Tambovtsev wrote:
<blockquote
cite="mid:B15F31DA9E9242DFAE933A4D09625271@ngufa28a6c2639"
type="cite">
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
<meta name="GENERATOR" content="MSHTML 8.00.6001.19258">
<style></style>
<div><font face="Arial" size="2">Dear Corpora members, I wonder if
you could advise me which math. statistics criteria is more
reliable to state the difference between two samples, that is,
t-test or chi-square? I used t-test to see the difference in
the occurrence of all colours in the texts of 26 British and
American writers. Should I also use the chi-square criterion
to substantiate the difference between two samples or t-test
is reliable enough? Looking forward to hearing from you to <a
moz-do-not-send="true" href="mailto:yutamb@mail.ru">yutamb@mail.ru</a>
Remain yours sincerely Yuri Tambovtsev, Novosibirsk, Russia</font></div>
<br>
</blockquote>
<br>
Good question, Yuri! T-test and χ-square are only used for
measuring the reliability of samples If you're using the entire
collected works of these writers (or some well-defined subset), and
not a sample, then you don't need either of those.<br>
<br>
You're probably looking for some measure of effect size:<br>
<br>
<meta http-equiv="content-type" content="text/html;
charset=ISO-8859-1">
<a href="http://en.wikipedia.org/wiki/Effect_size">http://en.wikipedia.org/wiki/Effect_size</a><br>
<br>
<pre class="moz-signature" cols="72">--
Angus B. Grieve-Smith
<a class="moz-txt-link-abbreviated" href="mailto:grvsmth@panix.com">grvsmth@panix.com</a></pre>
</body>
</html>