https://communities.sas.com/t5/SAS-Programming/Wilson-confidence-intervals-for-proportions/m-p/546579#M151343 <P>I need to report confidence intervals for proportions and proportion differences using the Wilson-score (also called Newcombe) method but I cannot use PROC FREQ because the proportions are output as predicted values from a logistic model. The algebraic algorithm could be incorporated into a data step but it's pretty messy - has anyone already created code to do this. Many thanks.</P> Wed, 27 Mar 2019 16:13:52 GMT lcmichael_unc 2019-03-27T16:13:52Z https://communities.sas.com/t5/SAS-Programming/Wilson-confidence-intervals-for-proportions/m-p/546579#M151343 <P>I need to report confidence intervals for proportions and proportion differences using the Wilson-score (also called Newcombe) method but I cannot use PROC FREQ because the proportions are output as predicted values from a logistic model. The algebraic algorithm could be incorporated into a data step but it's pretty messy - has anyone already created code to do this. Many thanks.</P> Wed, 27 Mar 2019 16:13:52 GMT https://communities.sas.com/t5/SAS-Programming/Wilson-confidence-intervals-for-proportions/m-p/546579#M151343 lcmichael_unc 2019-03-27T16:13:52Z https://communities.sas.com/t5/SAS-Programming/Wilson-confidence-intervals-for-proportions/m-p/546588#M151348 <P><A href="https://www.lexjansen.com/sesug/2015/103_Final_PDF.pdf" target="_blank">Paper: https://www.lexjansen.com/sesug/2015/103_Final_PDF.pdf</A></P> <P>&nbsp;</P> <P>Code:</P> <P><A href="https://raw.githubusercontent.com/Jiangtang/Programming-SAS/master/CI_Single_Proportion.sas" target="_blank">https://raw.githubusercontent.com/Jiangtang/Programming-SAS/master/CI_Single_Proportion.sas</A></P> <P>&nbsp;</P> <BLOCKQUOTE><HR /><a href="https://communities.sas.com/t5/user/viewprofilepage/user-id/268201">@lcmichael_unc</a>&nbsp;wrote:<BR /> <P>I need to report confidence intervals for proportions and proportion differences using the Wilson-score (also called Newcombe) method but I cannot use PROC FREQ because the proportions are output as predicted values from a logistic model. The algebraic algorithm could be incorporated into a data step but it's pretty messy - has anyone already created code to do this. Many thanks.</P> <HR /></BLOCKQUOTE> <P>&nbsp;</P> Wed, 27 Mar 2019 16:39:27 GMT https://communities.sas.com/t5/SAS-Programming/Wilson-confidence-intervals-for-proportions/m-p/546588#M151348 Reeza 2019-03-27T16:39:27Z