Survey Logistic Regression: Some SAS® and SUDAAN® Comparisons

SUDAAN and SAS produced slightly different estimates (Table 6). Example of SUDAAN’s message: “There is a problem with nest variable VARSTR=3.000000 in record 434. It has only one VARPSU whose value is 2.000000. Standard fix-up is to use the square of ...

www.nesug.org