Programming the statistical procedures from SAS

Odds Ratio in Proc GLIMMIX

Reply
Contributor
Posts: 27

Odds Ratio in Proc GLIMMIX

Hello,

 

I am fitting a two-level generalized linear dichotomous random intercept and random slope model using PROC GLIMMIX. Code below

 

Proc glimmix data = Merge.mCPR method = LAPLACE;

Class region;

Model V313M (event = last) = V013 V106 V701 V136F V190 V218C V384

GR1 GR2 FA1 FA2 FR1 FR2 Distance / S DIST = binary LINK = logit DDFM = BW OR CL;

Random INT V013 V106 V701 V136F V190 V218C V384 / Sub = region S CL Type = VC;

Covtest ZEROG/ wald;

Run;

 

Note: GR1, GR2, FA1, FA2, FR1, FR2 and Distance are continuous variables.

 

I have used ODDS RATIO in the model statement and it gave me 14 ODDS RATIO estimate. The maximum label of categorical variable I have is 7. It also gave me one confidence interval for all variables in the same row. Why it gave me 14 and same confidence interval? I couldn’t figure out which estimate is for which label?

 

 

 

I just want to have the report like below table.

Predictors

n

Adjusted odds ratio (95% CI)

Level-1 predictor variables (individual and household characteristics)

 

Modern contraceptive use

Age in years

15 – 19

534

 

 

20 – 24  

1436

 

 

25 – 29

1876

 

 

30 – 34

1591

 

 

35 – 39

1412

 

 

40 – 44

953

 

 

45 – 49

671

 

Residence

Urban

2261

 

 

Rural

6212

 

Number of living children

<= 1

2198

 

 

2 – 4 

3644

 

 

>= 5

2631

 

Level-2 predictor variables

 

 

Distance

 

 

Service provision readiness score

 

 

 

 

Here attached what the code gave me

 

 

Thanks in advance

Ask a Question
Discussion stats
  • 0 replies
  • 139 views
  • 0 likes
  • 1 in conversation