SMALL AREA ESTIMATION, SPRING 2015
Type and Credits
Intermediate level course
Exam (6 cu) or exam plus (optional) practical/theoretical homework (2 cu), total 8 cu
Advanced level course
Exam (6 cu) plus (compulsory) practical/theoretical homework (2 cu), total 8 cu
Lectures 15 hours, PC classes 15 hours
The course covers topics in modern statistical methods for the estimation of parameters for population subgroups or domains and small areas (Small Area Estimation, SAE). Topics include sampling design for SAE, design-based model-assisted methods (generalized regression estimation, calibration techniques), model-based methods (synthetic, EBLUP and EBP estimators), variance and MSE estimation, SAS tools, R tools, and real-world applications (mainly in social and health sciences and official statistics). Case studies include applications in the estimation of poverty indicators (poverty rate, inequality indicators) for regional areas (small or large). The course is of applied type.
When completing the course, students are expected to be familiar with approaches, methods and computational tools in the estimation for regions and other population subgroups and becoming capable to apply the methods in typical real-world analysis situations. Basic knowledge in statistical modelling, sampling methods and statistical computation (R, SAS) would help successful participation.
The course is intended to fit for students majoring or graduating in statistics and for Master level and post-graduate (doctoral) students in quantitative studies in applied sciences incl. social and behavioral sciences and economics (e.g. REMS). As an applied type course, the course also would fit well for statisticians and researchers in research institutes and elsewhere.
Lecture sessions: Tuesday at 16-19 Exactum CK111, Kumpula campus (note time change)
PC training sessions: Thursday at 16-19 Exactum C128, Kumpula campus
Exam: Tuesday 24.2.2015 at 16-18 Exactum CK111
Textbooks and selected articles
Rao J.N.K. (2003). Small Area Estimation. New York: John Wiley & Sons.
Lehtonen R. and Veijanen A. (2009). Design-based methods of estimation for domains and small areas. In: C. R. Rao and D. Pfeffermann (eds.), Handbook of Statistics 29B. Sample Surveys: Inference and Analysis. Amsterdam: Elsevier. pp. 219-249.
Lehtonen R. and Veijanen A. Model-assisted methods to small area estimation of poverty indicators. In: Pratesi M. (Ed.) (2015). Analysis of Poverty Data by Small Area Estimation. Chichester: Wiley. (Forthcoming, to be distributed to participants)
Lehtonen R. and Djerf K. (2008). Survey sampling reference guidelines. Luxembourg: Eurostat Methodologies and Working papers.
VLISS-virtual laboratory in survey sampling http://vliss.helsinki.fi
Topic 1 Introduction to SAE
Topic 2 Basic concepts and approaches
Additional reference: Saei and Chambers (2003)
Small area estimation under linear and generalized linear mixed models with time and area effects.
University of Southampton: S3RI Methodology Working Paper M03/15.
PC training materials
PC class 1 (updated in PC class 22 Jan.)
PC class 4 on Thursday 12 Feb.
Guest speaker: Adj.Prof. Ari Veijanen
RDomest materials (zipped folder)
PC class 5 (updated 19 Feb.)
Population dataset (SAS-data to be downloaded)
Population dataset (pop.txt)
Small area estimation in SAS
Resources to help you learn and use SAS
(UCLA Statistical Consulting Group )
RDomest (Dr Ari Veijanen)
Homework assignment (for intermediate and advanced levels)
Did you forget to register? Please contact tilasto-info[at]helsinki.fi.