- Home
- Search
- Recent additions and modifications
- Survey Research Pantheon (the great and the good)
- Surveys in the news
- SPSS
- Survey Analysis Workshop
- SPSS Textbooks
- Useful links for SPSS
- Survey Research Methodology, Practice and Training
- Survey Methods Textbooks
- Useful links for Survey Research
- Survey Unit, Social Science Research Council (UK)
- Subjective Social Indicators (Quality of Life)
- Survey Research Unit, Polytechnic of North London
- Contact
- WEB Log
Buckingham & Saunders 2004
Alan Buckingham and Peter Saunders
The survey methods workbook: from design to analysis (Polity Press, 2004)
is an interesting book, much of which is viewable on-line. It covers the entire survey research process from a discussion of "positivism" through questionnaire design, sampling, data collection, data cleaning and data analysis from frequency counts to multivariate inferential statistics. There's a very detailed contents list with extensive explanations and examples of statistical techniques and the analysis is done using SPSS via the GUI There's also an accompanying website which is unique in having both initial "dirty" and eventual "clean" data sets for the same survey (on smoking) but their use of SPSS is unsophisticated and disappointing (they use the default varnames, integer variables are left with two decimal places and none of the variable labels use question numbers from the questionnaire). Work on the site seems to have finished in 2004, but if you click on Login it takes you to an annotated list of very useful appendices with links which still function.
The survey methods workbook: from design to analysis (Polity Press, 2004)
is an interesting book, much of which is viewable on-line. It covers the entire survey research process from a discussion of "positivism" through questionnaire design, sampling, data collection, data cleaning and data analysis from frequency counts to multivariate inferential statistics. There's a very detailed contents list with extensive explanations and examples of statistical techniques and the analysis is done using SPSS via the GUI There's also an accompanying website which is unique in having both initial "dirty" and eventual "clean" data sets for the same survey (on smoking) but their use of SPSS is unsophisticated and disappointing (they use the default varnames, integer variables are left with two decimal places and none of the variable labels use question numbers from the questionnaire). Work on the site seems to have finished in 2004, but if you click on Login it takes you to an annotated list of very useful appendices with links which still function.
