Turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- Multivariable Logistic Regression Interpretation

Topic Options

- Start Article
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Sep 18, 2017 2:08 PM
(1933 views)

Hello,

I am completing a logistic regression model with multiple independent variables. I am having a hard time determining how to interpret the estimate/tests produced by SAS.

I had read that the "Effect likelihood test" had the best approximation for my P values to determine significance (since my data set is not large enough for the Wald test) and was using the Odds Ratios to determine what that effect was. However, for several values the effect likelihood ratio test determined significance (Prob>ChiSq was less than 0.05) but the 95% CI for the Odds Ratio cross 1 (and therefore is not significant). How do I interpret these differences?

Additionally, I would like to make sure that the model looked at whether the predictor values were actually independent. How do you test for this and what do you do if the "independent" variables are not actually independent?

Happy to send screenshots if needed. Thank you in advance

1 ACCEPTED SOLUTION

Accepted Solutions

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

In the odds ratio table JMP reports the Wald Chi square p-values and the Wald confidence limits. However, in the Effect Tests table, the Likelihood Ratio tests are reported by default. So they may be different from Wald tests in the Odds Ratio table. Effect Wald tests can be requested in the red triangle.

There is a test called Hausman-Wu test for testing endogeneity of an explanatory variable.

http://support.sas.com/documentation/cdl/en/etsug/63939/HTML/default/viewer.htm#etsug_model_sect051..... It is not currently available in JMP.

1 REPLY 1

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

In the odds ratio table JMP reports the Wald Chi square p-values and the Wald confidence limits. However, in the Effect Tests table, the Likelihood Ratio tests are reported by default. So they may be different from Wald tests in the Odds Ratio table. Effect Wald tests can be requested in the red triangle.

There is a test called Hausman-Wu test for testing endogeneity of an explanatory variable.

http://support.sas.com/documentation/cdl/en/etsug/63939/HTML/default/viewer.htm#etsug_model_sect051..... It is not currently available in JMP.