I am very new to using SPSS' Complex Samples package. I'm currently analyzing data from the ECLS-K:2011 dataset available from the U.S. Dept of Education. As I'm learning it, I've been creating frequency tables. I've run them weighted and unweighted in standard SPSS (Analyze>Descriptives>Frequencies). The unweighted total is ~18,000. (I haven't found a way to get an unweighted count when using a weight variable.) However, when I run a frequency using Complex Samples (Analyze>Complex Samples>Frequencies), the unweighted count is ~10,000. I would think the counts should be the same? I'm afraid to move on to actually analyzing the data (it's for my dissertation), if I can't even figure out frequencies. I would greatly appreciate any insight! TIA!

Dear Support Member​ .

If the question is related to statistics, implementation, coding, general troubleshooting, please log your query on the following link:

And then please Ask our SPSS Statistics community and get answers from other users and experts worldwide.

This forum is mainly for question related to technical support question for licensing, errors, installation, fix packs, trial set-up.

Best regards.

Support Member​ 

I thought that was where I was-- Thank you for letting me know!

Complex Samples and the regular procedures treat weights quite differently. In CS procedures, weights are treated as probability of selection while in others they are frequency weights. And in CTABLES, they can be treated as effective base weights. I would expect different results from all of these, especially if the weights are fractional.