Programming the statistical procedures from SAS

Questions about proc logistic options

Accepted Solution Solved
Reply
Frequent Contributor
Posts: 137
Accepted Solution

Questions about proc logistic options

Hi,

 

I have some questions about proc logistics procedure (SAS EG):

 

1) 'Effects' option: I can inclusd vars as 'cross', 'nest', 'factorial', what does it mean??, I ususally use 'main'.

2) How can I interpret 'Wald Chi-square', I thinks that measure the importance of the var in the model and the higher the value the better, am I wrong?

 

Any help will be greatly appreciated


Accepted Solutions
Solution
‎04-14-2018 01:59 AM
Respected Advisor
Posts: 3,066

Re: Questions about proc logistic options

Posted in reply to juanvg1972
  1. Crossed and nested factors are a fundamental concept in designing studies or experiments, and thus it affects the way you analyze the data. (There is no such thing as "factorial", both crossed and nested are "factors"). Your favorite search engine will find plenty to read on this topic, for example: https://www.theanalysisfactor.com/the-difference-between-crossed-and-nested-factors/
  2. The Chi-square does not indicate "importance". It indicates statistical significance. If the p-value (Pr>ChiSq) associated with the Wald Chi-Square is close to zero (the usual cutoff is <0.05) then the term is statistically significant in the model (in layman's terms, it is predicted a real effect and not predicting noise). If it is >0.05, then in layman's terms we can say that this model term is explaining noise (which is not a good thing) and so the term does not contribute to making the model more predictable.
--
Paige Miller

View solution in original post


All Replies
Solution
‎04-14-2018 01:59 AM
Respected Advisor
Posts: 3,066

Re: Questions about proc logistic options

Posted in reply to juanvg1972
  1. Crossed and nested factors are a fundamental concept in designing studies or experiments, and thus it affects the way you analyze the data. (There is no such thing as "factorial", both crossed and nested are "factors"). Your favorite search engine will find plenty to read on this topic, for example: https://www.theanalysisfactor.com/the-difference-between-crossed-and-nested-factors/
  2. The Chi-square does not indicate "importance". It indicates statistical significance. If the p-value (Pr>ChiSq) associated with the Wald Chi-Square is close to zero (the usual cutoff is <0.05) then the term is statistically significant in the model (in layman's terms, it is predicted a real effect and not predicting noise). If it is >0.05, then in layman's terms we can say that this model term is explaining noise (which is not a good thing) and so the term does not contribute to making the model more predictable.
--
Paige Miller
☑ This topic is solved.

Need further help from the community? Please ask a new question.

Discussion stats
  • 1 reply
  • 109 views
  • 0 likes
  • 2 in conversation