https://communities.sas.com/t5/Statistical-Procedures/Should-I-concern-signs-from-the-model-result-using-Weight-of/m-p/913942#M45371 No worry. It is normal.<BR /><BR />Check <a href="https://communities.sas.com/t5/user/viewprofilepage/user-id/13684">@Rick_SAS</a> 's blog:<BR /><A href="https://blogs.sas.com/content/iml/2023/03/27/simpsons-paradox.html" target="_blank">https://blogs.sas.com/content/iml/2023/03/27/simpsons-paradox.html</A> Thu, 01 Feb 2024 09:40:49 GMT Ksharp 2024-02-01T09:40:49Z https://communities.sas.com/t5/Statistical-Procedures/Should-I-concern-signs-from-the-model-result-using-Weight-of/m-p/913908#M45369 <P>Hi,</P><P>I ran a logistic regression model with only 5 inputs using WOE and found all 5 inputs are significant and have positive signs . However, when I ran bivariate model, some inputs have negative relationship with target variable. Should I worry about the sign issue? Thank you.</P> Thu, 01 Feb 2024 01:50:55 GMT https://communities.sas.com/t5/Statistical-Procedures/Should-I-concern-signs-from-the-model-result-using-Weight-of/m-p/913908#M45369 lionking19063 2024-02-01T01:50:55Z https://communities.sas.com/t5/Statistical-Procedures/Should-I-concern-signs-from-the-model-result-using-Weight-of/m-p/913942#M45371 No worry. It is normal.<BR /><BR />Check <a href="https://communities.sas.com/t5/user/viewprofilepage/user-id/13684">@Rick_SAS</a> 's blog:<BR /><A href="https://blogs.sas.com/content/iml/2023/03/27/simpsons-paradox.html" target="_blank">https://blogs.sas.com/content/iml/2023/03/27/simpsons-paradox.html</A> Thu, 01 Feb 2024 09:40:49 GMT https://communities.sas.com/t5/Statistical-Procedures/Should-I-concern-signs-from-the-model-result-using-Weight-of/m-p/913942#M45371 Ksharp 2024-02-01T09:40:49Z