Hi all,
I have an SPSS longitudinal database in which one of the variables is monthnumber. The month ranges from 0 (baseline) through all of the months they were involved and then 51 for last month/study end. I'm trying to identify which cases have a 0 but not a 51. So it looks something like this:
ID MonthNumber
1 0
1 1
1 2
1 51
2 0
2 1
2 51
Maybe I'm just brain dead, but I'm stumped on how to do this! Please help.
Amy Bacharach, I'm pretty sure that what you want to do can be done within the Select Cases tab in SPSS. It takes a little bit of fiddling and careful thought, however.
Hi Amy, this is a little bit of a not as straight forward one, because your data is long. It is easier in SAS, which maybe because I am less familiar with the SPSS functions. But if the dataset is in SPSS, it is still possible, and there are many potential ways. One way could be if you are able to convert the data to wide format, then you could use the Select Cases function Robert suggested above easily. E.g.
1) convert to wide format;
2) make "0" into 10 000 by doing recode all of the month variables as old=0 new=10 000;
3) make "51" into 100 000 by doing recode all of the month variables as old=51 new=100 000;
4) sum up all the months by compute a new target summative variable by summing up all of the month variables
5) use the Select case function suggested by Robert above, by selecting cases if the summative variable is >=10 000 AND
I recommend the SPSS procedure: Data - Sort. Sorting can be performed simultaneously according to several variables each in the desired direction.
Best regards,
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS