https://communities.sas.com/t5/SAS-Procedures/Proc-freq/m-p/64459#M18314 AFAIK Not by default because its not a standard statistical measure. I think if you want it you'll also need to calculate a formula for the 95% CI via the delta method.<BR /> <BR /> The odds ratio is more commonly used and for a 2x2 table that can be included in the output. <BR /> <BR /> You may also want to look up the binomial proportion estimates.<BR /> <BR /> HTH,<BR /> Reeza Tue, 10 May 2011 17:45:51 GMT Reeza 2011-05-10T17:45:51Z https://communities.sas.com/t5/SAS-Procedures/Proc-freq/m-p/64458#M18313 a 2X2 table below.<BR /> <BR /> N (row %)<BR /> <BR /> 70(89.7) 8(10.3)<BR /> 62(68.9) 28(31.1)<BR /> <BR /> <BR /> Question:<BR /> Is there any way to calculate the difference, e.g. 31.1%-10.3% and its 95%CI?<BR /> Thanks. Tue, 10 May 2011 17:38:36 GMT https://communities.sas.com/t5/SAS-Procedures/Proc-freq/m-p/64458#M18313 Ken_oy 2011-05-10T17:38:36Z https://communities.sas.com/t5/SAS-Procedures/Proc-freq/m-p/64459#M18314 AFAIK Not by default because its not a standard statistical measure. I think if you want it you'll also need to calculate a formula for the 95% CI via the delta method.<BR /> <BR /> The odds ratio is more commonly used and for a 2x2 table that can be included in the output. <BR /> <BR /> You may also want to look up the binomial proportion estimates.<BR /> <BR /> HTH,<BR /> Reeza Tue, 10 May 2011 17:45:51 GMT https://communities.sas.com/t5/SAS-Procedures/Proc-freq/m-p/64459#M18314 Reeza 2011-05-10T17:45:51Z