I am trying to compute a variable whereby I replace a value (9) in an ordinal variable with a randomly generated value drawn from an empirical distribution while keeping the other values (1-4) the same as the original variable.
I can accomplish this in SPSS if a theoretical distribution matched my data, such as:
DO IF Q10_D=9.
COMPUTE Q10D_RI = RND(RV.LNORMAL(2.683,.378)).
ELSE.
COMPUTE Q10D_RI = Q10_D.
END IF.
EXECUTE.
Any suggestions?
Bill
I can accomplish this in SPSS if a theoretical distribution matched my data, such as:
DO IF Q10_D=9.
COMPUTE Q10D_RI = RND(RV.LNORMAL(2.683,.378)).
ELSE.
COMPUTE Q10D_RI = Q10_D.
END IF.
EXECUTE.
Any suggestions?
Bill