SPSS Tips & Tricks #8 – delete missing
Hi, I got a question from a SPSS user who didn’t want to have the missing values within the Tree-analysis, (she use SPSS Decision Trees). If you don’t want cases with missing value in the tree, you must delete them before, and I will show you how. This differs from other analysis within SPSS like correlation, where you could choose to delete missing values pairwise or listwise within the dialog box (see example below):
Here below is a picture of the Tree before I deleted the missings in job_category. I know that 5 people doesn’t have any value on “Job_category” = sysmis :
As you see in the tree they have been merged with the “Team_leader” as they seems to have the same salaries. The tree analysis always merge groups together if they are similar to each other, statistically.
To delete these 5 cases do the command: Data – Select Cases and click on choice: “if condition is satisfied” and then click on “if”-button.
Then you get this dialog box, and you should write your condition into the white empty box. In this example I write “Job_category > 0”
(The variables can just be “clicked” in from the left variable list).
Click: “Continue” and click “OK”.
Now these 5 are filtered out (not deleted) and you can do the Tree analysis again. This time it looks like this:
Thanks for watching!