We analyze data from a European survey (34 countries) for quality of life. The sampling procedure is described as a "multistage stratified and clustered sampling design". Our outcome variable is binary.
There are three levels of data to be taken into account (individuals, regions, countries). I have read a few things about these procedures, some suggesting that proc glimmix should be used ( since one can take into account the three levels of data ), but also others stating that one should apply proc survey logistic, since the parameters Estimates might otherwise be biased . However in this procedure, one cannot consider the different levels of data.
Can you make a recommendation, which procedure should be used?
I agree with Amir, multilevel modeling should only be used if you are explicitly trying to estimate the variance between clusters. If you are, you should still take into account the survey design in your multilevel analysis to get correct population estimates. This can be done by specifying the sample weights in glimmix. If you do not use Glimmix based on your research question I would suggest using GEE (with proc genmod in SAS, you can specify link=logit and dist=binomial for logistic regression models) to account for the clustering in your data. Again survey weights can be used in a weight statement to get correct population estimates.
Use surveylogistic, particularly if the survey has released design information or replicate weights. Check out this book and their companions site for the best current info. http://www.isr.umich.edu/src/smp/asda/
I include this book and other related topics on my list of survey data analysis tools/issues. It's not particularly organized, but it may be helpful.
https://www.diigo.com/list/mattjans/complex-sample-survey-analysis
Thank you very much for your answers. Is the inclusion of three levels of data not similar to the "corrections" that surveylogistic does?? And what about the three levels of data - I cannot incorporate that into suvrylogistic, can I?
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology