top

Exemplar 1
1.1 Introduction
1.2 Getting started
1.3 Diferent anwsers for mean income and its standard error
1.4 Pifalls
1.5 Percentiles
1.6 Income for single parents
1.7 Data set and survey details

Exemplar 2
2.1 Background
2.2 Getting started
2.3 Results for simple tabulations
2.4 Chi squared tests
2.5 Results for subgroups
2.6 Results for logistic modelling
2.7 Data problems
2.8 Survey design
2.9 Data set details

Exemplar3
3.1 Background
3.2 Getting started
3.3 Smoking rates
3.4 Smoking rates by regions and health boards
3.5 Factors affecting smoking
3.6 Which packages
3.7 Details of survey and the data set

Exemplar4
4.1 Background
4.2 Sample design
4.3 Features of this exemplar
4.4 Getting started
4.5 Analyses
4.5.1 Weightedproportions
4.5.2 'equal opps' and other work place characteristics
4.5.3 The impact of weights on logistic regression
4.5.4 Finite population inference
4.6 Which packages can do these analyses
4.7 The dataset

Exemplar5
5.1 Background
5.2 Getting started
5.3 Carrying out thereweighting
5.4 Young women and drug use
5.5 Details of the survey and the dataset

Exemplar6
6.1 Background
6.2 Types of imputation
6.3 Getting started
6.4 Results with a small set of variables
6.5 Imputing with a large number of derived variables
6.6 Imputing with original questions
6.7 Comparisons and conclusions
6.8 Details of the survey and data sets

top

> THEORY

1. Stratification
Introduction
Proportionate stratitified sample
1.1 What it is
1.2 A simple example
1.3 Why proportionate stratification
1.4 Stratification estimates
1.5 Proportionate stratification
1.6 Calculating standard errors
1.7 Stratification for surveys with long fieldwork periods
Disporportionate sampling
1.8 What it is
1.9 A simple example
1.10 Why it is used
1.11 Obtaining unbiased estimates
1.12 The effect of disproportionate sampling on standard errors

2. Clustering
2.1 Introduction
2.2 Impact clustering has on standard errors
2.3 Selecting a proportionate clustered sample

3. Weighting
3.0 Introduction
3.1 Scaling of weights
3.2 Weighted means
3.3 Weighted tables
3.4 Weighted percentiles
3.5 Why weight?
3.6 Calculating weights
3.7 The effect of weighting on standard errors
3.8 Weighting in regression analyses
3.9 Different kinds of weights

4. Errors
4.1 What is a standard error
4.2 Design effects for surveys
4.3 Methods for survey standard errors
4.4 The practicalites of replication methods
4.5 How does the design affect standard errors?
4.6 Chi squared tests for survey data
4.7 Misspecification effects

5. Non response weighting
5.1 Why?
5.2 Post stratification
5.3 From sampling frame
5.4 Assumptions
5.5 Effect on standard errors

6. Imputation

6.0 Introduction
6.1 Imputation or weighting?
6.2 Types of imputation
6.3 Missing data concepts
6.4 Patterns of non-response
6.5 How to handle complicated missing value patterns
6.6 Software for imputation
6.7 Features and problems with imputation software

7. Analysing subgroups of surveys
7.1 Introduction
7.2 Stratification and subgroups
7.3 The effect of subgroups
7.4 Design effects for subgroups

8. Checking the survey design
8.1 Introduction
8.2 Split PSU's
8.3 Lonely PSU's

9. Finite populations
9.1 Introduction
9.2 Simple random sampling
9.3 FPC in stratified sampling
9.4 FPC in two-stage sampling

top

>SOFTWARE PACKAGES

A: the R project
A1 IIntroduction
A.2 Web LInks
A.3 Costs and Licencing
A.4 What kind of package
A.5 Methods for Complex Surveys
A.6 User written code

B: The SAS system
B1 Introduction
B2 Using SAS with P|E|A|S exemplars
B3 Web LInks
B4 Costs and Licencing
B5 What kind of package
B6 Methods for Complex Surveys
B7 User written code
B8 Imputation

C: The SPSS system
C1 Introduction
C2 Web LInks
C3 Costs and Licencing
C4 What kind of package
C5 Methods for Complex Surveys

D: The Stata system
D1 Introduction
D2 Web links
D3 Costs and Licencing
D4 What kind of package
D5 Methods for Complex surveys
D6 Contributed package for survey analysis by replication methods
Changes from Version 8 to Version 9

top