Statistics is the biggest problem in my PhD study (first year Psychology), especially when dealing with multivariate statistics. What is (are) the best statistic book (s) you have used for both basic and more advanced statistic skills? I am struggling which book to buy for future reference as well. Any recommedations?
Andy field? Hair et al. ? Tabachnick and Fidell ?

Many thanks!

Andy Field in my opinion. If you would like an electronic copy, please send me a pm.

I think it depends whethr you've done them before. I'm new to statistics too and need them for my history thesis (yep, lol, ever thought I'd need that!) and I'm hopeless with numbers - add a few letters in too and I'm done! I was given statistics without tears by Derek Rowntree as a starter and am looking for more 'statistics for dummies' type books to get me in, particularly as I only need a fairly basicish grounding in it. I found the more complex books totally incomprehensible and enough to have me reaching for the vino! I've got to do an MA qualitative methods course in the spring term -dreading it - does your uni offer anything like that in any of the depts in your uni? The course I'll be doing isn't run by my own dept but I was advised to do it by my supervisor.

I know others love Field I just happen to have bought a different SPSS guide (Pallant - has been fine for me). Tabachnick and Fidell and similar are good as occaisional references but you might not get your money's worth from buying copies - wait and see if you need them that much. I have Cohen et al Applied multile regression/correlation for behavioural sciences which is very thorough and it saves me the trouble of trekking to the library as I work mainly from home. Depends what you are doing (how advanced is advanced?). I find if you have very specific isues you can usually find resources/papers online to help. The further you get into the math the less likely any one or even two texts are to deal with your issues as advanced stats is an area of math with many differences of opinion. I would try out other books from the library first as they are quite an expensive investment and I have bought a couple I really don't that find useful. I've found some really useful advice in this hefty text though: Handbook of Psychology volume 2 Research methods in psychology, Schinka and Velieer.

If you're looking for help on specific statistics issues or methods, I can recommend Sage's University Paper Series "quantitative applications in the social sciences". It's a series of little green books each concentrating on a specific issue/method and geared for applied researchers. The series covers more or less every imaginable topic relevant to applied statistics -- granted, with a focus on social sciences, but I would think that many statistical needs of psychologists would be covered there too. You get the booklets for a few pounds e.g. from bookdepository (online bookshop with free shipping).

It's been a very steep learning curve for me with the maths. I found the few lectures available almost unendurable! It only worked for me to be working with real data and figuring it out as I go along. Personally I think maths is the very worst subject to try and learn from a lecture format.

i can really recommend "statistics for business and economics" by mcclave, benson and sincich which i found very helpful even though i'm a linguist :-)