Dear community,
I have to weight cases in SPSS due to the calculation of norm values. Something mysterious happens when I compare frequency tables of the sample before and after weighting the cases. Imagine a sample of n = 1.000. If I create a frequency table before the weighting, there are in total n = 1.000 people. If I do the weighting and create a frequency table of the sample after doing that, only n = 997,48 people are in the total. If I have a look at the data, still n = 1.000 people are in there. Actually I have no clue, why this difference occurs. Might it be due to a rounding procedure in SPSS? Would be thankful for any advice, so that I can explain my frequency table to a reader.
Many thanks in advance and greetings from Germany,
Frank
What is the total (sum) of your weights column? so that the total is shrunk back In this case because it is not 1000. If each weight for all the 1000 was 1 you would get the same as the unweighted analysis.
You were right, the total of the weights colums is a little bit below 1 - that already solves my problem. :) This might be, because the weighting variable was calculated out of several other variables like age and sex and so on. It confused me a few days, but you hit the point. Thank you very much, I appreciate your help!
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS