Recode single-factors or Spss variables with many coding values

From InterSciWiki

Lets say you have a variable you want to map or cross-tab, but like single-factor score variables, it has more than six values and is too complicated to map or cross-tab. If you are recoding in order to create a MAP of the factor variable, dont make too many categories (4, for example, is a good number).

Go to /Transform / Visual Binning - and then move a variable into the "Variables to bin" window. If you just created the factor, it will be at the bottom of your variable selection list. Then name the new variable in the blank window just below the name of the input variable (use a Name ending in ...Fact5Bin, where 5 is the number of bins). Now, in the scanned variable list (upper left), click the one you want to bin, i.e., to create discrete breaks into fewer categories. This will produce an image of the distribution frequenties for that variable in the previously empty Binned Variable window.

Then follow instructions to create breaks and create the binned variables. E.g., Click "Make Cutpoints" then provide some low number (4, 5, or 6) of categories (or use another option like means and standard deviations. When you click OK you will see the cutpoints in Blue. You may move them around and click/cut with the delete key to get rid of some cutpoints. Object is to get a good spread of cases within each interval. When you are done relocating cutpoints, press OK. DONT FORGET to enter the name of the new variable as above.