Looking for a stats book

[strudel]

I'm after recommendations for a decent stats book - what i'm looking for is not how to do a particular test, but the logic needed to produce correct statistics. I'm doing some work at the moment that has irregular numbers and i need to decide if they're true representations or issues caused by sampling. I'm after the sort of thinking that would be applied to produce freakonomics/fooled by randomness than stats for beginners (I'm mathematically competant!)

Any suggestions?

[Graham E]

I'd suggest the "sports almanac" featured in back to the future 2 - it has the result to every sporting event in a 10 year period, no matter how small or irrelevant the event was. As a bonus, all of this infomation can fit ino a magazine roughly the size of a copy of "razzle".

[Jem Thompson]

I have just done a level 1 stats course at university. I have a coursebook full of all the basics such as hypothesis testing, confidence intervals, normal distribution and so on. Its an interesting topic, I really enjoyed it. You should probably download SPSS if you are doing stats as well

[Podie]

I've got my old Uni stats text books in the loft - will have a look if you want the details.

[strudel]

I did notice the sps5 books when I had brief look on amazon, I'll investigate that course depending on how much future work is in the pipeline. Podie - don't put yourself out on my account, but any names are useful!

[Podie]

I'll most likely be in the loft on Friday - will have a nose.