- Docente: Daniela Cocchi
- Credits: 10
- SSD: SECS-S/01
- Language: Italian
- Moduli: Daniela Cocchi (Modulo 1) Michele Scagliarini (Modulo 2)
- Teaching Mode: Traditional lectures (Modulo 1) Traditional lectures (Modulo 2)
- Campus: Bologna
- Corso: Second cycle degree programme (LM) in STATISTICS, ECONOMICS AND BUSINESS (cod. 8056)
Learning outcomes
This is an intermediate course on survey sampling, which illustrates the theory of complex sampling plans. The comparison between design based and model based inference is performed. The topics of the course aim at analyzing and summarizing the information obtained in problems of sampling from finite populations and to deal with non standard situations.
Course contents
The course is organized in 2 parts
Part 1 (Daniela Cocchi)
a) Design based inference
Elementary units and complex units in sampling from finite populations: cluster sampling. Cluster sampling with equal and unequal size of primary units. Cluster sampling with varying probability, ratio estimation in cluster sampling. Systematic sampling.
Two stage sampling with varying probability, simple random sampling in two stage sampling; simple random sampling in two stage sampling with equal size primary units.
b) Model based estimation
The most important models and comparison with design based inference.
c) Model assisted design based inference: GREG estimators.
d) Small area inference: design based indirect estimators. Area specific or unit specific superpopulation models.
e) Calibrated and balanced estimators
Part 2 (Michele Scagliarini)
The issue of multiphase sampling. Two phases independent and dependent sampling. Ratio estimator in two phase sampling.
Special problems in stratification. Two phase sampling when strata size is unknown. Post-stratification.
Sampling in multiple occasions. Composite estimators for repeated sampling with partial overlapping of units. Regression estimator in sampling in two occasions as a component of the composite estimator.
Variance estimation in non standard situations: non linear estimators and complex sampling plans.
Sequential sampling and hypothesis testing in finite populations. The sequential test of probability ratio: hypothesis testing on the parameter of a Bernoulli; hypothesis testing on the mean of a normal.
Readings/Bibliography
A downloadable syllabus (by Daniela Cocchi) is downloadable. It contains a summary of an elementary course.
Teaching methods
lectures
Assessment methods
Written and oral test .
Teaching tools
A file with the most important formulae can be downloaded
Office hours
See the website of Daniela Cocchi
See the website of Michele Scagliarini