MARKET ANALYSIS
The training aim for the student is to reach the following learning outcomes:
Knowledge and ability to understand
The course aims to provide methodological and application advances of business survey methods. We also want to encourage students to specialize in the use of the open source statistical package R to select random samples and produce survey estimates.
Ability to apply knowledge and understanding
At the end of the teaching course, the student will also be able to select random samples and produce survey estimates with Design-Based and Model-Assisted methods. The acquired knowledge will allow him to estimate economic and / or business aggregates.
Concepts, definitions and nomenclatures
Typologies of surveys
Frame Set up
The questionnaires
Population and sampling
The characteristics of some sampling designs
Design-Based Estimation techniques The use of auxiliary information in the sampling estimate
Data Editing
Data dissemination
1.1. Background
1.2. Concepts, definitions and nomenclatures
1.3. Survey Typologies
1.4. Frame Set-Up
1.5. The questionnaires
2. Population and sampling
2.1. Finite populations and superpolations
2.2. Samples with or without replacement
2.3. The sample space
2.4. Random samples and sample design
2.5. The inclusion probabilities
2.6. Sample selection algorithms
3. The characteristics of some sample designs
3.1. Introduction
3.2. The simple random sample
3.3. Samples with variable probabilities
3.4. The systematic sample
3.5. The stratified sample
3.6. Cluster sampling
3.7. The use of two or more sampling stages
3.8. Sampling of spatial units
3.9. The balanced sampling
4. Design-Based Estimation techniques 4.1. Introduction
4.2. Statistics and estimators
4.3. Estimate of the total
4.4. The estimators of Hansen-Hurwitz (HH) and Horvitz-Thompson (HT)
4.5. Variance of estimates and estimate of sampling error
4.6. Estimators and relative variances for some sampling plans
5. The use of ancillary information in the sample estimate
5.1. Introduction
5.2. The use of auxiliary variables
5.3. Model-assisted or model-based estimates
5.4. The estimators of post-stratification of the differences and the ratio estimator
5.5. The generalized regression estimator
5.6. Calibration estimators
5.7. Correction from the effects of total non-response
5.8. Estimate for small domains
6. Data Editing
6.1. Introduction
6.2. The set of rules
6.3. Identifications of errors
6.4. Selective and interactive corrections
6.5. Methods of imputation
7. Data Dissemination
7.1. Introduction
7.2. Protection of confidentiality
7.3. Dissemination of elementary data
7.4. Dissemination of estimates or aggregated data
Course slides.
Further readings:
Sarndal, C. E., Swensson, B. and Wretman, J. (1992). Model Assisted Survey Sampling. Springer, New York, NY.
Lectures.
R practice and exercises.
Knowledge and ability to understand
To verify the learning a test and an interview are scheduled. The test will consist of theoretical questions and empirical exercises on the whole program with particular attention to the use of the software R, simulating sample selection and estimation on real case studies. The final evaluation, expressed in thirtieths, takes into account both the test and the interview.
Ability to apply knowledge and understanding
During the test and the interview, students' ability to apply the knowledge of advanced survey methodologies is tested so as to be able to deal with specific case studies.
E-mail: benedett@unich.it
In the first semester the Professor sees students only by appointment (benedett@unich.it). In the second semester, the office hours for students is scheduled for Wednesday from 4 pm to 6 pm, studio DEC 2 ° floor, Viale della Pineta, 4.