https://communities.sas.com/t5/Statistical-Procedures/CL-for-OR-contains-0-99/m-p/777883#M38122 <P>I have an odds ratio that is 0.88 (95% CL: 0.78,0.99). I'm keeping decimals to 2 places, so is it technically still fair to consider the OR estimate statistically significant (since it still doesn't contain 1.0 when using 2 decimal places)? Or for publication purposes, would something like this be downplayed?</P> <P>&nbsp;</P> Tue, 02 Nov 2021 11:22:45 GMT SAS93 2021-11-02T11:22:45Z https://communities.sas.com/t5/Statistical-Procedures/CL-for-OR-contains-0-99/m-p/777883#M38122 <P>I have an odds ratio that is 0.88 (95% CL: 0.78,0.99). I'm keeping decimals to 2 places, so is it technically still fair to consider the OR estimate statistically significant (since it still doesn't contain 1.0 when using 2 decimal places)? Or for publication purposes, would something like this be downplayed?</P> <P>&nbsp;</P> Tue, 02 Nov 2021 11:22:45 GMT https://communities.sas.com/t5/Statistical-Procedures/CL-for-OR-contains-0-99/m-p/777883#M38122 SAS93 2021-11-02T11:22:45Z https://communities.sas.com/t5/Statistical-Procedures/CL-for-OR-contains-0-99/m-p/777889#M38123 <P>Some journals address issues related to rounding in their author guidelines. Usually that guidance is related to p values, but I would recommend looking to see if the target journal makes reference to the presentation of statistics such as odds ratios.&nbsp;&nbsp;</P> Tue, 02 Nov 2021 12:07:40 GMT https://communities.sas.com/t5/Statistical-Procedures/CL-for-OR-contains-0-99/m-p/777889#M38123 svh 2021-11-02T12:07:40Z