DISCLAIMER: I profess no expertise in Complex Samples, never having used it.
You have 24 schools from 24 different countries. Therefore, country is redundant once you know which school it is. The results are as I would expect them to be. You would need more than one school per country (in at least one of the countries) for country to provide additional information.
oxymoron123 wrote
Hi all, I have a sample drawn from 4 countries and 24 schools (naturally, each school is "unique" to a country and there's no overlap). I created an analysis plan file with country and school as clusters and weight=1 for all subjects. I then ran regressions and noticed the below output:
---------
Sample Design Information
Stage 1 Strata 1
Units 24
Sampling Design Degrees of Freedom 23
-----------
It looks as if SPSS took into the account of school clusters (24) but not countries. I then ran the same analyses with another analysis plan in which the only cluster variable is school, and had the same results. I just wanted to check that this is okay or if there are other ways to tell SPSS about my country X school clusters?
Many thanks for your help in advance.
--
Bruce Weaver
bweaver@lakeheadu.ca
http://sites.google.com/a/lakeheadu.ca/bweaver/"When all else fails, RTFM."
PLEASE NOTE THE FOLLOWING:
1. My Hotmail account is not monitored regularly. To send me an e-mail, please use the address shown above.
2. The SPSSX Discussion forum on Nabble is no longer linked to the SPSSX-L listserv administered by UGA (
https://listserv.uga.edu/).
Locked
