I should select a random sample of 'n' residents from the general list containing 'N' records. I need all the records from the general list to have the same probability to be chosen,
I tried to do that in Stata, but I haven't found a way:
- the command sample only provide tools for pseudorandom sampling, so the probability to be chosen from the general list becomes higher and higher after every extraction;
- "bsample" does a boostrap extraction-with-replacement for a specified number of extractions, but the point is that I need 'n' residents, not 'n' extractions!
Is there any other way to do what I'd like to do (either in Stata or in R)?
Many thanks for any suggestion!
The link below may possibly help
http://stackoverflow.com/questions/10240991/taking-data-sample-in-r
Thanks! I previously found another Q&A on StackOverflow, but it wasn't that clear if probabilities to be chosen were the same even when replace=F or not...
In the end, I found out that the package in R does pseudorandom sampling as well as (as bad as?) Stata's sample. Finally, I asked a workmate to help me and we implemented a user-written cycle in Stata to do a true random sampling.
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Extraction