SPSS Statistics

SPSS Statistics is one of the most widely used and most capacious statistical programs. Its variety of possibilities is comparable to SAS (e.g. ability to analyse large data sets). SPSS Statistics is in particular recommendable for categorical data. Originally SPSS Statistics was developed for application in the field of social sciences, but today includes procedures for most areas of statistics. SPSS Statistics offers control using a graphical menu, which makes it rather simple to use. There is also the possibility to use command line based control using a syntax editor, which next to reproducible data analysis offers additional options not accessible by menu control. Therefore the latter method of control is strongly recommended for advanced users.

Availability at ETH Zurich

SPSS Statistics is available for all major platforms. ETH offers the current versions for Windows, Mac and Unix at a price of Sfr 50.- for a single workstation license at IDES. And also a free WAN version, which requires an Internet connection.

Consulting and courses

The Seminar for Statistics offers consulting for SPSS Statistics, as well as a single-day introductory course in German and English. Also the CompiCampus program offers evening courses for students regularly.

Program structure

SPSS Statistics has a modular structure, i.e. starting with the Base module, a system of modules, which are needed to perform the analysis, is combined. The following list gives a brief overview of the most important modules. Note, that there are differences across platforms.

• Base: As the name suggests the basic module of the SPSS Statistics module family. It contains comprehensive possibilities for file and data management, as well as basic procedures for graphics and statistics (descriptive statistics, simple ANOVA models, linear regression, non-parametric tests, contingency tables, factor analysis, cluster analysis, discriminant analysis, etc.)
• Advanced Models: Offers among other things methods for generalized linear models (including multivariate models, repeated measures, variance components, log-linear models, Cox regression and Kaplan-Meier-Estimators)
• Regression Models: Contains logistic and multinomial regression, probit analysis, nonlinear regression as well as weighted least squares.
• Classification Trees : Methods for data analysis using classification and decision trees.
• Tables : For efficient representation of data in tables, e.g. contingency tables.
• Exact Tests : For the analysis of very small data sets.
• Categories : Contains optimal scaling procedures for categorical data (correspondence analysis, nonlinear principal component analysis, multidimensional scaling, etc.)
• Trends : For time series analysis.
• Conjoint : Conjoint Analyse for market research.
• Missing Values : Diagnostics and estimation of missing values.
• Maps : Drawing maps for demographic analyses.
• Complex Samples : For experiment design.

Literature

The complete manual for SPSS Statistics can be viewed at the Seminar for Statistics, HG G15. The manuals are feasible and give good instruction. Interested persons can also contact the consulting team for further support.

There are also several good books dealing with SPSS Statistics. Some are available through the ETH library.

Springerlink
Several books about SPSS Statistics are available online on Springerlink .

Further information

Address of SPSS Switzerland
SPSS Schweiz AG, Schneckenmannstrasse 25, CH-8044 Zürich
Tel. 044 266 9030, email: info@spss.ch

• SPSS web site
• SPSS Switzerland