In even the best run survey, there will be missing values in the data. People might refuse to answer a question, for example, or give an ambiguous answer. Rather than removing these cases completely from your data set, you can set SPSS to ignore these cases for the variable which has been missed.
You can set codes for missing values from the ‘Variable View’ tab on SPSS, selected from the bottom left of the screen.
The cells in the missing column of the variable view are active; click and the Missing Value dialog will open.
There are different choices in the dialog for how you want to assign missing vlaues.
For a numerical variable you can have up to three discrete missing values, or a range. Each missing value might have a different meaning, such as ‘refused to answer’ or ‘unknown’. In this case we have a single missing numerical value, 999.
If the variable was a string, we would set the missing value to a text string. In our case, for our ports of embarkation and disembarkation, we can use a space ‘ ‘ (without the quote marks).
Durham SPSS 