Frequently Asked Questions

Here we try to answer some of the questions about PSPP that arise from time to time.

What is PSPP?

PSPP is a program for statistical analysis of sampled data. It is a free replacement for the proprietary program, SPSS.

One goal of the PSPP project is compatibility with the SPSS language. It currently features:

High-quality output formatting.
An easy to use graphical user interface.
A command line interface to allow seasoned users to rapidly perform analysis.
Portability: PSPP is easy to port. It should be simple to port it to any POSIX.1 compliant operating system having an ANSI C compiler.

What can PSPP do?

PSPP provides many transformations and utilities. Procedures provide the ability to perform t-tests, anova, linear regression, reliability analysis, factor analysis, non parametric tests and other analyses. Future versions of PSPP will provide a greater variety of statistical analyses.

Refer to the manual for the complete list of supported commands.

Where can I get PSPP?

See this page.

How do I install PSPP?

There are several methods:

Find a copy of PSPP packaged for your operating system. This is often the easiest method, but might not get you the most up to date version. An incomplete list of packages for various operating systems.
Download a tarball and follow the instructions in the INSTALL file. This is the canonical method used for most GNU software.
However, if really want the very latest version, and are willing to put in extra effort to install it, and are prepared to accept that it may contain bugs, then compile from the Git snapshot. Look in the file README.GIT file and follow the instructions there.

How do I get started using PSPP?

After you've built and installed PSPP, try this:

Using your favourite editor, write a file (name it myfile.syn) containing the following:

DATA LIST LIST /name (a25) quantity (f8).
BEGIN DATA.
widgets 10345
oojars 2345
dubreys 98
thingumies 518
END DATA.
 
LIST.
 
DESCRIPTIVES /quantity
 /statistics ALL.

Enter the command pspp myfile.syn
Look at your results in pspp.list

PSPP is very versatile and there are lots of different types of statistical analysis possible. You need to become familiar with the documentation to unleash its full potential.

What documentation is available?

The PSPP manual, written in Texinfo format, is the primary reference for using PSPP. Besides the Texinfo format a few other formats are available like HTML and PDF. A few additional documentation files are included in the source distribution:

README: A quick introduction to PSPP.
INSTALL: Notes on building and installing.
ChangeLog: Details of the latest bleeding edge changes.

A manual for PSPP in French, written independently by Julie Séguéla, is also available in PDF and LaTeX formats.

PSPP is a clone of SPSS. So why should I use PSPP when SPSS has all the same features?

Firstly, PSPP is not a "clone". No part of SPSS was used to write PSPP and there is no code shared between the projects. If it was a clone, then there would indeed be no advantage using one program or the other. But instead, PSPP is an independent free software program (and one of its design goals was ease of use for people already familiar with SPSS).

Among other advantages, this means that if you find it doesn't completely fit your needs - perhaps there is a statistical test which you would like it to perform - then you have the means and the right to improve it to fit your requirements. Whereas if you attempted to do this with non-free software such as SPSS it would be both difficult and illegal.

Furthermore, because PSPP is free software, you can use it for whatever you like and for as long as you like (there's no "expiry") and you are welcome to give away (or sell) copies to others - for example to students undertaking an introductory statistics course. If you have made your own improvements, then you can include those changes in the copies you distribute.

Another important advantage: Since the source code is available to all, unlimited peer review is possible. Should PSPP's accuracy be called into question, it can easily be audited by a competent, independent authority. Conversely, since the source code of proprietary software is secret, nobody can check to see if it contains numerical instability, inappropriate algorithm implementations or other potential errors.

Some secondary reasons why people sometimes prefer PSPP include:

Reduced costs
Inter platform portability
Features not available in other software, such as importing from Postgres, Gnumeric etc.
It comes already bundled with popular operating systems, such as Debian GNU/Linux

How can I get more information on PSPP?

There are a few resources for PSPP information. The first is this FAQ and related documentation in the distribution and on the web site. The second is the PSPP source code itself, if you are programmatically inclined.

For discussion with other PSPP users our mailinglist pspp-users@gnu.org is available. You can subscribe to this list at http://lists.gnu.org/mailman/listinfo/pspp-users

Finally, we look forward to all comments and questions related to PSPP. The irc channel is #pspp, which is on the freenode network.

Why don't you merge PSPP with R, GRETL, Octave ... <my favourite program>?

This is unlikely to happen for several reasons:

PSPP has different goals to these other programs. PSPP is intended to be a Free version of SPSS, and fully compatible with it, whereas GRETL (for example) provides its own language and structure.
It's probably harder than you think. Merging two programs which have been developed independently is rarely trivial.
PSPP has several technical advantages over its counterparts which we don't wish to lose. Notably PSPP is very efficient at processing extremely large datasets.
Many of these other programs have numerous authors, each holding copyright on their respective parts. By contrast, the copyright in PSPP is help by one entity (the Free Software Foundation), which makes for easier copyright tracking and enforcement. Merging with another program would mean losing this advantage, or needing to ask scores of authors to re-assign their copyright. This would be neither practical nor polite.

However, we certainly want to provide features which will make PSPP easier to interact with other programs. If you want an import/export filter or some other feature to help PSPP complement your favourite program, then please talk to us about it.

PSPP lacks this feature which I really need. How can I get PSPP to support it?

There are several ways:

Download the source code and implement it yourself.
Hire somebody to implement it for you.
Send an email to the developers and ask for the feature. The developers are interested in hearing about what users want.

Any help you can give is appreciated. See the following question.

Can I help out with PSPP development?

Yes. Visit http://lists.gnu.org/mailman/listinfo/pspp-dev where you can subscribe to the mailing list or view the archives.

When you have fixed a bug or improved some parts of PSPP, you can submit your patches at http://savannah.gnu.org/patch/?group=pspp

What does PSPP stand for?

PSPP does not have any official acronymic expansion. But they're easy to come up with. For example:

Perfect Statistics Professionally Presented.
Probabilities Sometimes Prevent Problems.
People Should Prefer PSPP.

Send along your favorites!

Help! I'm stuck

If you need help using or installing PSPP, you can try one of the following:

Reading this FAQ again carefully.
The PSPP user's mailing list,
Channel #pspp on freenode.