Abstract
Under "urnmodel-conditions", this means by using equal inclusion proba-bilities for all objects in the population and without consideration of the nonsampling-error, quota sampling with combined quotas is stratified random sampling with a multidimensional stratification-variable. Therefore it can claim the statistical attributes of stratified sampling, the unbiasedness of its estimators for the parameter under interest m and the gain in precision compared to simple random sampling. In general - even under these assumptions - quota sampling with marginal quotas produces biased estimators for m. The conditions for an unbiased estimation concerning the quota variables are determined and so proposals for the practical use of quota sampling are derived.
| Translated title of the contribution | Quota sampling under urn model conditions |
|---|---|
| Original language | German (Austria) |
| Number of pages | 19 |
| Journal | Österreichische Zeitschrift für Statistik |
| DOIs | |
| Publication status | Published - 1997 |
Fields of science
- 101 Mathematics
- 103 Physics, Astronomy
- 105 Geosciences
- 305 Other Human Medicine, Health Sciences
- 504 Sociology
- 106 Biology
- 502 Economics
- 509 Other Social Sciences