I am a new entry in statistical field with lots of interest in this area. I would like to know any statistical book which can enhance my statistical knowledge meaning; wish to know the name of the book which can help me in model prediction....for eg. detailed explanation and steps of regression (incl multiple, logistic, poisson...) , Anova, etc....

The book should contain each and every concept of stats so that I can practically handle projects in my company.

  • 1
    $\begingroup$ I'd suggest searching the site for answers to questions with the 'references' tag. This one & this one may be useful. It's more than a little optimistic to hope that one book will cover "each and every concept of stats" - a more realistic wish would be for an introductory general text plus an intro. to general linear models followed by an intro. to generalized linear models. $\endgroup$ Jun 11, 2014 at 8:59
  • $\begingroup$ I meant to put this one as well. If you don't find any useful recommendations here try to differentiate your question from the others by adding specific detail about what you're looking for with respect to topics covered, mathematical level, theoretical vs applied perspective, &c. $\endgroup$ Jun 11, 2014 at 9:23
  • 1
    $\begingroup$ I echo @Scortchi 's comment. Of course, different people have different training, but I took around 10 statistics related courses in graduate school (my PhD is in psychometrics). Each used at least one book. $\endgroup$
    – Peter Flom
    Jun 11, 2014 at 9:24
  • $\begingroup$ Basically, I have been given a task wherein I have to apply model prediction...but thinking which model to predict..so want to know various classifications when to apply which model...so that I can read through the book and make one chart for my reference for future so that if given a project I can just glance through my preparation and make any model anytime. Is this possible? with any book $\endgroup$
    – Arvinder
    Jun 11, 2014 at 9:57
  • 1
    $\begingroup$ Well, you could make a chart of the "default" models tried in different kinds of situation - Poisson regression for counts &c. - but once you've learnt the relevant techniques well enough to be able to use them successfully you'll be familiar enough with them not to need a chart any more. See e.g. these lecture notes to get an idea of how a textbook will cover GLMs. $\endgroup$ Jun 11, 2014 at 10:42

1 Answer 1


In brief (and accounting for my own preferences...), the posts linked by @Scortchi suggest Gelman & Hill, Data Analysis Using Regression and Multilevel/Hierarchical Models, which is a wonderful book. I love it, I've even contributed to translate its BUGS examples to Stan (here).

However, if you are interested in a "detailed explanation and steps of regression (incl multiple, logistic, poisson...) , Anova, etc...", a useful first book could better be Kutner et al., Applied Linear Statistical Models.

You'll find a thorough discussion about simple and multiple linear regression, including diagnostics, remedial measures, model selection and validation (Chapters 1-11), logistic and poisson regression (14), design of experimental and observational studies (15), ANOVA (16-21, 23-24), analysis of covariance (22), random and mixed effects models (25), nested designs (26), repeated measures (27) etc.

Gelman & Hill could be the best second step.


Not the answer you're looking for? Browse other questions tagged or ask your own question.