In this lecture, basic methods for the collecting of data using complex survey designs, methods for pre-processing the data as well as methods for the estimation of indicators from complex surveys are in main focus (~"from data to indicators")
For applications, data from official statistics, from geosciences, from clinical studies and data from the social and economic sciences in general are used. The methods are applied on these data with the help of the software environment R.
1. collection of data:
- CAPI, CATI, online, retrospective, ...
- addresses/registers/online resources/sampling frames/cost aspects
2. Complex Sampling Designs,
- stratified, multiple stage designs
- sample size issues
3. Pre-processing
- editing
- imputation (estimation of missing or erroneous information)
4. Calibration and Estimation
- non-resonse calibration
- weighted estimation
5. Evaluation:
- variance estimation, bias
- comparing methods through simulation
- resampling methods
6. Dealing with non-random samples:
- retrospecive studies
- response propensity models
7. Data fushion
- record linkage
- statistical matching
8. Practical appliations with R (+ brief introduction to R)