As discussed in the previous task, specifying the survey design to analyze data using SUDAAN typically requires four key elements. However, when you wish to analyze subgroups of the data, a fifth element is required. This fifth element is the SUBPOPN statement, which allows you to define your subgroup of interest for the analysis, without removing respondents from your master dataset.
| Code | Element |
|---|---|
|
proc sort; by sdmvstra sdmvpsu ; run; |
Element 1 |
|
proc <SUDAAN procedure> design=WR ; |
Element 2 |
|
nest sdmvstra sdmvpsu ; |
Element 3 |
|
weight <appropriate sample weight variable>; |
Element 4 |
|
<more SUDAAN procedure syntax>; |
|
| subpopn (<logical expression defining subgroup>) | Element 5 |
|
run; |
|