Books/Guides for learning SPSS

is there a reason you're looking to use SPSS over other statistical packages like STATA, SAS or R?

I like SPSS quite a bit because of how easy it is to use. However, if you expect to do this for a living, learn SAS. The drug industry has latched onto SAS for some reason, so lots of SAS programming jobs are out there.

R is free, so you can use that to learn the basics of stats if you want.

I actually found R to be harder to learn than either SPSS or STATA. I use STATA primarily and find it MUCH faster to use than SPSS. The learning curve is a little steeper, but once you learn the commands you can fly through things. SPSS is VERY easy to learn and would probably be good enough for beginning work.

Most people I know use R for more advanced statistical problems that the other programs aren't suited to handle.

Depends, lot's of social science work and market research people use SPSS.

I've used SPSS and STATA, spss is definitely easier to learn and it's a lot easier to make nice looking graphs in it compared to STATA. If your doing a lot of really basic stuff SPSS is probably best, but I do think STATA is easier to use if your doing a lot of advanced stats or if your working with really large data sets.

http://www.amazon.com/Discovering-St.../dp/1847879071

That's the best book for learning stats and SPSS in conjunction.

having used a number of different statistical packages, i just can't understand why people even use spss. it just doesn't do much, although it is the easiest to use; but maybe too easy to use. only advantage i can think of is that it is reasonably stronger in anova-related procedures than stata. but, spss doesn't have count, survival, or complex survey data procedures (at least when i last used it anyways, not sure if it has changed). imo, start with stata and be done with it, instead of learning spss and finding one day you'll have to learn stata. the learning curve for stata isn't that much steeper than spss. if you have the time, simultaneously learn sas which is best if you have complex data management needs. it's also the statistical package of choice in business and government. r is where the future is at and what most statisticians use these days. steepest learning curve. but if psych undergrads at my school are being taught it, there's no reason a person can't learn it quickly if he devotes the appropriate amount of time.

Honestly, save yourself some money and consult this... HEAVILY.

The majority of my grad school stats classes ripped material from here. It provides just about everything you'll need to know, especially if you already have some familiarity with stats.

Sorry... double post.

STATA is much better for larger datasets and complicated stats. The interface and output is much more intuitive and easier to read.

I guess I don't understand why STATA is better for large datasets than SPSS. When I was doing this professionally, SPSS could be run on mainframes and you were limited really by compute power, not by the program at all. I never had a problem with large datasets.

Is there a problem with newer SPSS versions in this regard? Also, the Python interpreter for the new SPSS seems to be a huge advantage. I hated coding SPSS scripts for everything. Python has to be much easier.

I've been out of this for a long time now. Getting paid as an hourly with no benefits to be a statistician/programmer was not worth it.

Incredibly powerful software .. in a 30 second free download. But fuck the learning curve. I got through all my undergraduate and most of my graduate econometrics material using R fine .. IE: Doing normal OLS in R is no harder then STATA. But try to do some of the more "cutting edge" stuff, and the software goes to shit. IE: we are trying to do state-space and cointegration modeling in R .. and what takes me a half hour to figure out and run in STATA takes me an afternoon to replicate in R. Which is the biggest downside to R. The learning curve is always there, as the software is open source and the consistency in programing for new packages is lacking. But alas, that is the only cost of the software, so ya. Not much to complain about.

I'm a firm supporter of SAS for industrial work. Learn some easy Base SAS from "The Little SAS Book", then start learning SAS STAT and SAS OR. Once you know something about SAS, start looking through this:
http://support.sas.com/onlinedoc/913/docMainpage.jsp

If you want to learn something you don't need to pay for, I'd suggest R. For most academic uses, R will suffice. It's seriously slow for large (1 GB+) data sets though.

Try SPSS Step by Step by Cole Davis:   http://www.coledavis.org/policypress.html