Thanks everyone for your responses. I re-ran the program adding "missing" to the proc tabulate statement as LaurieF suggested, but got the same exact table with the issues. mkeintz, the log is attached - there is only one dataset "dsn3" so I don't know how EG could be pointing to different datasets, and it finds all the variables and everything, but just shows values that cannot be accurate. Reeza, it's interesting that for d2: age, the sample sizes are correct, though, right? FYI, I have run this program hundreds of times on dozens of datasets and never had this issue, and it has always included that variable "x" that is always missing for all observations. FYI Cynthia_sas, I create the variable "x" and label it "<---->". I do this to create a blank line, because it ends up being a really long report and it helps me break out each variable into sections. Thanks again, all!
... View more