| title: | Multivariate estimation problem in sampling theory |
|---|---|
| reg no: | ETF5523 |
| project type: | Estonian Science Foundation research grant |
| subject: |
1.1-1.5. Exact Sciences |
| status: | accepted |
| institution: | TU Faculty of Mathematics |
| head of project: | Imbi Traat |
| duration: | 01.01.2003 - 31.12.2006 |
| description: | So far multivariate parameters are almost not considered in sampling theory. Therefore the aims and expected results of the project are: Study of the estimators of the finite population covariance matrix and its functions, and of other multivariate parameters in survey sampling context where sampling design affects behaviour of estimators. Derive in general case (i.e. for an arbitrary sampling design) distributional characteristics of the observed estimators (mean vector, covariance matrix, etc.), either exact or asymptotic. Derive special cases of interest. Study of the effect of the sampling design to results. A working hypothesis is that asymptotically effect of the design decreases and the results holding in sampling context tend to the classical results when sample and population sizes increase. The idea is to write a book (as a side result) along the approach developed by us through the years. |
| project group | ||||
|---|---|---|---|---|
| no | name | institution | position | |
| 1. | Kristiina Rajaleid | University of Tartu | doctoral student | |
| 2. | Kaja Sõstra | University of Tartu | doctoral student | |
| 3. | Imbi Traat | TU Faculty of Mathematics | Associate Professor | |