SPSS is way too expensive for a student in Intro to Statistics (me). Most of the class remains a confusion to me despite my previous successes in math classes. Furthermore, I really don't want to have to use the public computer labs here on campus with the dirty keyboards and mice, as well as the small screen resolution, and of course, Windows (nobody in their right mind really wants to use Windows). In an effort to prevent having to take my work to the lab, I went seeking a solution. I could have illegally cracked the SPSS demo, but I really don't like to do such things, and don't encourage it...
After much googling, I found PSPP. PSPP is a nearly direct clone of SPSS. What's better is that it's multi-platform, uses the same file format as SPSS, and is totally open source and free. PSPP is a GNU project, and I have already been in contact with a developer who seems to be passionately at work to make the program even better.
The only caveat I ran into was that the output was only to two decimal places, and I needed three. My professor uses a program to check our answers against an answer set, and it requires three decimal places...after communicating with the developer, I found that if I enter “SET FORMAT=F8.3.” in the syntax editor and run it (see the video) that it will then output to three decimal places.
I managed to find all of the frequency statistics I needed, as well as the correlation coefficient. The output is all very nice, though text based, and I believe the developer told me he plans on improving it further.
That said, PSPP is very good and I highly recommend it. For the project information, source code, and building instructions, click here. Windows users can download prebuilt versions of the program here (just install and use).
thanks to your comments I downloaded PSPP. Since my spss licence had expired and I needed to work with data saved in that format, I needed something free right away.
PSPP looks like a good thing, but I am having trouble with computing frequencies. Eventhough I only have 340 cases, when computing frequencies PSPP gives me a very weird result 5436 cases!!!
Did you have any similar problems when opening exisitng Spss data files?
I was using a significantly smaller data set, but it should work.
I suggest you first look again within the next few days to see if the new release of PSPP is out...it should be. I know you probably don't want to compile it.
That said, if you can't get it working, I suggest you file a bug report...they'll want a copy of your dataset (possibly) but will really want to know what functions you're performing and such.
http://lists.gnu.org/mailman/listinfo/pspp- users
That's the mailing list...you don't have to sign up, but you can send the bug report to the email address pspp-users@gnu.org ... one of the developers is from Stanford, and he really helped me a lot when I first found the software.
Good luck!
(this is the beauty and curse both of opensource...)
The download links are still there, you'll have to scroll down. It would appear that site is having some wordpress issues. I'm sure they'll fix them eventually.
Try this one:
http://www.cecaps.ufmg.br/pspp/files/setup-ps pp-20090518.exe
