Building models with SAS Enterprise Miner, SAS Factory Miner, SAS Visual Data Mining and Machine Learning or just with programming

Interpretation of exponentiated coefficient for categorical variable in EMiner logistic regression?

Accepted Solution Solved
Reply
SAS Employee
Posts: 24
Accepted Solution

Interpretation of exponentiated coefficient for categorical variable in EMiner logistic regression?

What is the interpretation of the highlighted value in the image below?
 
I understand that in the odds ratio table (not pictured), the displayed value for this level of the categorical variable will be different because it will be compared to a reference level. But what does it mean here exactly - when not compared to a reference level?
 
temp.png

Accepted Solutions
Solution
‎02-19-2016 12:08 PM
SAS Employee
Posts: 24

Re: Interpretation of exponentiated coefficient for categorical variable in EMiner logistic regressi

I liked your suggestion, so I tried changing the input coding. Under Model Options -> Input Coding -> GLM.

 

With deviation coding the values are not the same:

 

Analysis of Maximum Likelihood Estimates
Parameter DF Estimate Error Chi-Square Pr > ChiSq Estimate Exp(Est)

M_DemAge 0 1 0.0741 0.0344 4.65 0.0311 1.077

 

Odds Ratio Estimates
Point Effect Estimate

M_DemAge 0 vs 1 1.160

 

With GLM coding, the values are the same. Thanks!!

 

Analysis of Maximum Likelihood Estimates
Parameter DF Estimate Error Chi-Square Pr > ChiSq Estimate Exp(Est)

M_DemAge             0      1      0.1389      0.0794          3.06        0.0802                       1.149

 

Odds Ratio Estimates
Point Effect Estimate

M_DemAge             0 vs 1         1.149

View solution in original post


All Replies
Super User
Posts: 17,837

Re: Interpretation of exponentiated coefficient for categorical variable in EMiner logistic regressi

It should be the same as your odds ratio table, and it should be for a specific level against reference table, assuming you've set the coding system to be the same.

 

In this case it looks like it's Level=0 against whatever your reference is.

If the variable is an indicator variable with 2 levels this makes sense.

 

If you're not getting the same as your odds ratio it's likely because you haven't specified the categorical variables to use the referential=PARAM coding system, but the GLM coding system.

 

http://support.sas.com/documentation/cdl/en/stathpug/68163/HTML/default/viewer.htm#stathpug_introcom...

Solution
‎02-19-2016 12:08 PM
SAS Employee
Posts: 24

Re: Interpretation of exponentiated coefficient for categorical variable in EMiner logistic regressi

I liked your suggestion, so I tried changing the input coding. Under Model Options -> Input Coding -> GLM.

 

With deviation coding the values are not the same:

 

Analysis of Maximum Likelihood Estimates
Parameter DF Estimate Error Chi-Square Pr > ChiSq Estimate Exp(Est)

M_DemAge 0 1 0.0741 0.0344 4.65 0.0311 1.077

 

Odds Ratio Estimates
Point Effect Estimate

M_DemAge 0 vs 1 1.160

 

With GLM coding, the values are the same. Thanks!!

 

Analysis of Maximum Likelihood Estimates
Parameter DF Estimate Error Chi-Square Pr > ChiSq Estimate Exp(Est)

M_DemAge             0      1      0.1389      0.0794          3.06        0.0802                       1.149

 

Odds Ratio Estimates
Point Effect Estimate

M_DemAge             0 vs 1         1.149

Super User
Posts: 17,837

Re: Interpretation of exponentiated coefficient for categorical variable in EMiner logistic regressi

Yeah, that's what I was trying to say in a roundabout way Smiley Happy

☑ This topic is SOLVED.

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

Discussion stats
  • 3 replies
  • 423 views
  • 3 likes
  • 2 in conversation