somebody says if known population we should go by Random Population. somebody says if it is unknown population we should go non random population is it true
If you have a population which is unknown, then you have a problem. A nonrandom sample from it won't tell you about the population in general.
For a known population, you might have a random sample, and if you have covariate data you might use a model-assisted or model-based approach. Any attempt at inference from a nonprobability sample requires covariate data. You need to know your population to be able to infer to it from your sample.
Also, if by "unknown population" you mean one that is "hard to reach," "rare," or "hidden," you could look at 'snowball' sampling. (It might at least help you learn more about your population.)
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Writing