Hi there,

I'd to adjust my dataset by using SPSS Rake Weights, but I have missing a missing category in the population (the data was not measured). What do I enter in the "Categories and Control Tables or Fractions 1"

1 .15 2(?) 3. .35 4 .45

The unweighted data

Date Freq. Percentage

Group A 10%

Group B 5%

Group C 45%

Group D 40%

The population data is NOT complete - Group B was not measured.

Education Freq. Percentage

Group A 15%

Group B UNKNOWN (not measured)

Group C 40%

Group D 45%

Thank you in advance !

Without a control total for group B, you might be best off omitting that group from the weight calculation altogether. That would produce a missing value for the B weight, and those cases would automatically omitted from statistical procedures.

Otherwise, if you have some idea what the B proportion should be, you might try several reasonable values that span the possibilities and see whether your conclusions from the statistical procedures are reasonably robust to those choices.

But I am a bit confused in that your category identifiers at the top are numeric but the percentages are alphabetical. Maybe those are just labels.

Note that after you run RAKE, the calculated weight is turned on, so if you are going to try different possibilities, you should turn weighting off after each round.

Thank you so much Jon!