randomness of survey select samples

[wkossack]

I'm using proc survey select to randomly sample a population.

I've noticed some curious patterns in the resulting samples and I'm looking into the source.

Are there any problems with the randomness of survey select samples using the default seed? I've seen similar problems in random number generators only to discover that the period of the random numbers is smaller than I need

[Cynthia_sas]

Hi:
You might want to look at the documentation
http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#/documentation/cdl/...

and see whether the METHOD= option offers you a method closer to your liking (by my count, there are at least 10 or 11 methods to try).

As it says on the above page: "If you do not specify the METHOD= option, by default, PROC SURVEYSELECT uses simple random sampling (METHOD=SRS) " and the doc for the SRS method explains that SRS is simple random sampling, which is selection with equal probability and without replacement. And for more information about simple random sampling, you are further referred to:
http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#/documentation/cdl/...

cynthia