Topics in Social Statistics, fall 2010
Lecturers
Part 1: Seppo Laaksonen
Part 2: Risto Lehtonen
Scope
Part 1 is focused on survey data collection and data cleaning. The final output will the cleaned micro data. This is thus the file that is ready to use or analyse. It happens very often that survey micro data files are not very well cleaned and consequently the analyst is not happy with it. Hopefully this situation will be improved in future. There exists good examples, fortunately, such as the European Social Survey (ESS) data that is even freely available for everyone. This should also be the target. Go on to look at the ESS website.
The basic material for part 1 is attached:
http://wiki.helsinki.fi/download/attachments/59059560/Topics+Social+Statistics_Seppo.pdf
This covers basic concepts as well as principles of data collection modes, sampling designing and editing and something on imputation and weighting. Principles of cleaned data close the material. Comments welcome.
The next file is the description of the sampling file. In this presentation a very fine file is presented after a bad file. The sampling file is a micro file of the gross sample units that consists of key variables of sampling design, fieldwork and auxiliary information useful for cleaning data well.
https://wiki.helsinki.fi/download/attachments/59059560/Sampling_File_Presentation_Durban.pdf
Lectures 12 hours
PC classes 6 hours
4/6 cu (exam 4 cu + optional homework assignment of 2 cu)
Type
Intermediate/Advanced studies
Lectures
PART 1 (Prof. Seppo Laaksonen)
Lectures: Tuesday 7 Sept. at 14-18 and Thursday 9 Sept. at 14-16 (SSKH IT-sal)
PC training: Thursday 9 Sept. at 16-19 (SSKH IT-sal)
PART 2 (Prof. Risto Lehtonen)
Lectures: Monday 27 Sept. at 14-18 (NOTE time changed!) and Thursday 30 Sept. at 14-16 (SSKH IT-sal)
PC training: Thursday 30 Sept. at 16-19 (SSKH IT-sal)
Note: SSKH IT-sal, address Snellmaninkatu 12
Exams
Tuesday 12 Oct. at 14-17, Main Building (Fabianinkatu 33), Class 4 **NEW**
Description
Social statistics focuses on statistical methods for describing and analysing social phenomena and change. The methods are widely used in empirical research in many fields, including social and behavioural sciences and official statistics production. Important sub-areas of social statistics are survey sampling (sampling techniques, estimation), survey methodology (data collection methods, nonresponse treatment, measurement errors) and survey analysis (design-based and model-based multivariate methods). The course concentrates on topics in survey methodology (Part 1) and survey analysis (Part 2). The orientation is of practical type including general description of the methodology, worked examples and real-world case studies. Selected software is used in computer classes. The course can be included in methodologically oriented master studies, post-graduate methodological studies and in the REMS program, for example.
Textbook
Lehtonen R. and Pahkinen E. (2004). Practical Methods for Design and Analysis of Complex Surveys. Second Edition. Chichester: John Wiley & Sons. Chapter 8.
Web materials
VLISS-virtual laboratory in survey sampling http://mathstat.helsinki.fi/VLISS/
Lecture materials
Part 2: Analysis of Complex Survey Data
Computer classes
SAS demo SAS program code - SAS results for design-based logistic ANCOVA
SAS data set SAS-data OHC
SPSS data set SPSS-data OHC - SPSS Results for design-based logistic ANCOVA
Homework assignments
Homework (2 cu) using the OHC Survey data here **NEW**
Registration
Did you forget to register? What to do.
Updated 29 September 2010