<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: Contrast Statement in Proc Logistic. in SAS Health and Life Sciences</title>
    <link>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5606#M499</link>
    <description>Thanks!</description>
    <pubDate>Wed, 28 Nov 2007 03:56:18 GMT</pubDate>
    <dc:creator>Rucstat_huadli</dc:creator>
    <dc:date>2007-11-28T03:56:18Z</dc:date>
    <item>
      <title>Contrast Statement in Proc Logistic.</title>
      <link>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5604#M497</link>
      <description>In a logistic model, the explanatory variable treatment has four levels.I want to test the first versus the fourth level of treatment, we usually use this contrast statement below&lt;BR /&gt;
&lt;BR /&gt;
contrast '1 vs 4' treatment 1 0 0 -1;&lt;BR /&gt;
&lt;BR /&gt;
but,I read the SAS help about The LOGISTIC Procedure contrast statement,they use &lt;BR /&gt;
&lt;BR /&gt;
contrast '1 vs 4'  treatment 2 1 1;&lt;BR /&gt;
&lt;BR /&gt;
Both have different results,Why?</description>
      <pubDate>Thu, 22 Nov 2007 06:26:03 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5604#M497</guid>
      <dc:creator>Rucstat_huadli</dc:creator>
      <dc:date>2007-11-22T06:26:03Z</dc:date>
    </item>
    <item>
      <title>Re: Contrast Statement in Proc Logistic.</title>
      <link>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5605#M498</link>
      <description>I guess it depends on the way your variables are coded : check the PARAM option in the CLASS statement.&lt;BR /&gt;
If you typed CLASS treatment / PARAM = GLM ; then your coefficients 1 0 0 -1 are correct. If you simply used CLASS treatement ; without options, or with PARAM=EFFECT, then the coefficients would be 2 1 1 as the help says.&lt;BR /&gt;
&lt;BR /&gt;
Both lead to the same results :&lt;BR /&gt;
[pre]&lt;BR /&gt;
ODS SELECT contrastTest (PERSIST) ;&lt;BR /&gt;
PROC LOGISTIC DATA = sashelp.class ;&lt;BR /&gt;
	WHERE age IN (12,13,14,15) ;&lt;BR /&gt;
	CLASS age ;&lt;BR /&gt;
	MODEL sex = age ;&lt;BR /&gt;
	CONTRAST "test 12 vs 15" age 2 1 1 ;&lt;BR /&gt;
RUN ;&lt;BR /&gt;
PROC LOGISTIC DATA = sashelp.class ;&lt;BR /&gt;
	WHERE age IN (12,13,14,15) ;&lt;BR /&gt;
	CLASS age / PARAM = GLM ;&lt;BR /&gt;
	MODEL sex = age ;&lt;BR /&gt;
	CONTRAST "test 12 vs 15" age 1 0 0 -1 ;&lt;BR /&gt;
RUN ;&lt;BR /&gt;
ODS SELECT ALL ;&lt;BR /&gt;
[/pre]</description>
      <pubDate>Tue, 27 Nov 2007 10:42:02 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5605#M498</guid>
      <dc:creator>Olivier</dc:creator>
      <dc:date>2007-11-27T10:42:02Z</dc:date>
    </item>
    <item>
      <title>Re: Contrast Statement in Proc Logistic.</title>
      <link>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5606#M499</link>
      <description>Thanks!</description>
      <pubDate>Wed, 28 Nov 2007 03:56:18 GMT</pubDate>
      <guid>https://communities.sas.com/t5/SAS-Health-and-Life-Sciences/Contrast-Statement-in-Proc-Logistic/m-p/5606#M499</guid>
      <dc:creator>Rucstat_huadli</dc:creator>
      <dc:date>2007-11-28T03:56:18Z</dc:date>
    </item>
  </channel>
</rss>

