I am working with a stratified random sample with population weights and using Complex Samples for analysis. I am new to CS and have several questions.
1. I set up two complex sampling plans, one with the strata variable and one without (and both with the population weight) as a test to see how much adding the strata affected the standard errors and confidence intervals around population totals and proportions. I am getting the exact same results with each plan, as if the strata variable doesn't make a difference. I'm wondering why this is happening.
2. When I run complex samples crosstabs, I see there is a test for the overall independence of the variables, but I don't see a test for difference of proportions across column (independent variable) categories as there is in a regular crosstabs (Analyze > Descriptives > Crosstabs) or in Custom Tables. Is there a difference of proportions test in Complex Samples Crosstab?
3. The sample I am using did not come with a sample balance weight so I constructed one by dividing the population weight by its mean. If I were running any analysis not in complex samples (frequencies, crosstabs, regressions) I would use the sample balance weight for statistical testing, not the population weight. In complex samples, should I use the population weight or a sample balance weight?
Many thanks for your guidance.
-- Laurence Kotler-Berkowitz | Senior Director
(c) 646-872-9681
(0) 510-848-2502 x.185
CONFIDENTIALITY NOTICE - This message (including any attachments) contains information that may be confidential. Unless you are the intended recipient (or authorized to receive for the intended recipient), you may not read, print, retain, use, copy, distribute or disclose to anyone the message or any information contained in the message. If you have received the message in error, please advise the sender by reply e-mail, and destroy all copies of the original message (including any attachments).