<?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: power/sample size for cochran mantel haenszel in Statistical Procedures</title>
    <link>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957648#M47960</link>
    <description>&lt;P&gt;I'm not aware of a way to do a power analysis specifically for the CMH test. However, since your response is binary, you could instead take a model-based approach to the same sort of test by using a logistic model with your stratifying variables as predictors in the model. For this approach you could get a power analysis using the LOGISTIC statement in PROC POWER. See the discussion and examples in the POWER documentation.&lt;/P&gt;</description>
    <pubDate>Wed, 29 Jan 2025 23:06:38 GMT</pubDate>
    <dc:creator>StatDave</dc:creator>
    <dc:date>2025-01-29T23:06:38Z</dc:date>
    <item>
      <title>power/sample size for cochran mantel haenszel</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957585#M47954</link>
      <description>&lt;P&gt;Can someone provide an example of how to calculate power/sample size for cochran mantel haenszel? The response variable is binary and there are two stratum (one has 3 levels and the other is binary).&lt;/P&gt;</description>
      <pubDate>Wed, 29 Jan 2025 15:04:00 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957585#M47954</guid>
      <dc:creator>msecic</dc:creator>
      <dc:date>2025-01-29T15:04:00Z</dc:date>
    </item>
    <item>
      <title>Re: power/sample size for cochran mantel haenszel</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957648#M47960</link>
      <description>&lt;P&gt;I'm not aware of a way to do a power analysis specifically for the CMH test. However, since your response is binary, you could instead take a model-based approach to the same sort of test by using a logistic model with your stratifying variables as predictors in the model. For this approach you could get a power analysis using the LOGISTIC statement in PROC POWER. See the discussion and examples in the POWER documentation.&lt;/P&gt;</description>
      <pubDate>Wed, 29 Jan 2025 23:06:38 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957648#M47960</guid>
      <dc:creator>StatDave</dc:creator>
      <dc:date>2025-01-29T23:06:38Z</dc:date>
    </item>
    <item>
      <title>Re: power/sample size for cochran mantel haenszel</title>
      <link>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957650#M47961</link>
      <description>You can perform any POWER analysis for any statistic test by data simulation.&lt;BR /&gt;Check &lt;a href="https://communities.sas.com/t5/user/viewprofilepage/user-id/13684"&gt;@Rick_SAS&lt;/a&gt; blogs:&lt;BR /&gt;&lt;A href="https://blogs.sas.com/content/iml/2020/08/12/simulation-estimate-power-of-test.html" target="_blank"&gt;https://blogs.sas.com/content/iml/2020/08/12/simulation-estimate-power-of-test.html&lt;/A&gt;&lt;BR /&gt;&lt;A href="https://blogs.sas.com/content/iml/2020/08/17/power-curve-parallel-sas-viya.html" target="_blank"&gt;https://blogs.sas.com/content/iml/2020/08/17/power-curve-parallel-sas-viya.html&lt;/A&gt;</description>
      <pubDate>Thu, 30 Jan 2025 02:07:58 GMT</pubDate>
      <guid>https://communities.sas.com/t5/Statistical-Procedures/power-sample-size-for-cochran-mantel-haenszel/m-p/957650#M47961</guid>
      <dc:creator>Ksharp</dc:creator>
      <dc:date>2025-01-30T02:07:58Z</dc:date>
    </item>
  </channel>
</rss>

