PSPP vs. SPSS

[memilanuk]

Hello all,

I was looking at some applied stats books to further my education beyond where my intro stats course left off... the problem is that many of these books nowadays assume you have some sort of specialized stats software packge available. The intro stats course I took had sections (at the end of each chapter) on doing those sorts of exercises in MiniTab, Excel, or a TI83/84 (and we were expected to do everything on a TI83/84 - yuck!). The 'higher level' texts all seem to expect MiniTab, SPSS, SAS, etc.

So... my question is, is the current version PSPP close enough in functionality to SPSS to be usable (more or less) straight across with one of these texts? I'd rather not end up with a $100 paperweight that is unusable without a $1000+ software package...

[gunksta]

At this time - PSPP is not a full blown replacement to SPSS.

Depending on what you are trying to learn - it may or may not be ready yet.

There are however, some other choices. On the R website, they list resources for learning R. Some of these are on-line guides. Others are $100 books. There are even books written for people learning stats, where all of the examples are prepped in R.

Unfortunately, the R website appears to be dead right now, so I'm having a hard time being more specific. But, I would say that there are lots of good books written for R. Go to Amazon and look around.

Another option which is also interesting is SOFA. I believe the author used to hang around here. It's a new and rapidly developing app. PSPP is an awesome idea, but it seems like they need more developers. They aren't moving very fast. SOFA is much younger and yet in some ways seems a bit more mature (but harder to install since it's not in the repos).

Downside of SOFA - because its so young, nobody has written a book for people learning statistics using SOFA. (This could be your big opportunity!)

Good Luck!

[memilanuk]

Hmmmm... I have a couple of the 'intro to statistics with R' books... according to the authors, they were intended to work along with a 'real' stats book, usually a specific one (i.e. the examples were tailored to suit). Most of the books that go beyond the basics on the R site are very discipline specific - mostly bio-informatics or econometrics. I was looking at something more general like this, but as I mentioned, most of the examples are tailored to one specific (expensive) software package.

Sofa does look pretty interesting. It'd probably cover at least 90-95% of what I'm interested in doing easily.