|
|
A while back Jon Peck very kindly provided a script for weighting multiple
variables to their targets(RIM weighting) using Python running under SPSS
15.0.1.
My query/question is a related one. While I can weight to any number of
variables using this method, I am unsure about how to do the following:
If, say, I have a sample size of 1000 people (weighted to reflect the age,
gender, social class, regional distribution of a given population) how do
I assign values to individual cases which are 'more significant' than
others?
Everyone has been asked on a scale of 1 to 10 how likely it is that they
will vote at the next Election and I want to assign values to each case on
a sliding scale from value 1 for 10/10 likelihood to 0.9 for 9/10 down to
0.2 for 2/10.
Can I do this within the same syntax/script that I am using for the
purpose of weighting the sample using the age, sex, social class &al.
parameters?
If so how?
Thoughts and ideas would be very much appreciated.
Yours
Rick Nye
Director, Populus
|
Locked
