How do I turn outliers for a certain variable into missing values?

The ''recode into different variable'' doesn't quite work for me because the variable in question consists of 3 different survey versions which all have different cut-off scores when it comes to outliers. Therefore I can't just say ''all values higher than ...'' should become missing values as that value at which it should become missing differs per survey version. Also, some zeros are coded protest answers and some are not, and therefore I can't just simply say ''0 --> missing value''. I did create a dummy variable with what is a protest answer and what is not.

The ''select cases if'' does not give an option to turn excluded cases into missing values, only to exclude them from the dataset. I don't want the whole respondent to be removed from the dataset, just the one response.

I tried to use the ''select cases if'' option in the ''recode into different variable'' menu but I have not been successful so far.

I'm wondering if I can do this in a way that is not manually deleting all outliers. Aside from the fact that that would be a very cumbersome approach, my supervisor should be able to track what I have done and how.

Thanks in advance. Of course I can provide more information if needed.